Intel Wants To Charge $50 To Unlock Your CPU's Full Capabilities

MBCook writes "Turnkey CPU upgrades aren't just for mainframes anymore. According to Engadget, OEMs (including Gateway) are selling computers with the Intel Pentium G6951, which can have extra cache and hyper-threading enabled through a $50 software unlock called Intel Upgrade Service."

Comments

I'm all for it

[Anonymous Coward]

Especially since it'll likely be pirated before the CPU ships.

Re:I'm all for it

[camperslo]

I suppose the "price above all else" PC vendors will like this. If they can load up machines with demo software, why not have demo hardware to match?

Don't expect Apple to go along with this though...

This may backfire in multiple ways. As this gets attention it'll make consumers more aware of the potential (which usually means likely) privacy implications of having serialized CPU chips.

Re:I'm all for it

[BrokenHalo]

Pirated or jailbroken, that is one CPU I will not buy. Intentionally holding a gun to the customer's head by crippling the device until you pay a ransom is not a way to get my business.

Re:I'm all for it

[DurendalMac]

Exactly. Free unlocking FTW!

Re:I'm all for it

[mysidia]

You know... Intel being the CPU manufacturer, could make this really robust. Each CPU already gets stamped with a unique serial number. They could stamp each one with a unique unlock code that goes with the serial number, as well.

Then the only way to 'unlock' the function would be to go through Intel, so they would look up your specific CPU's unlock code in the database.

That's impossible to pirate, because there's no way you can share the code. As long as they assure the unlock code is the only mechanism allowed to re-enable the capabilities, and there is no BIOS mechanism to override the lock.

Re:I'm all for it

[hardburn]

Happens all the time, actually, they usually just don't offer a way to unlock it. They make a run of all the chips of a given architecture, then put them through tests. The ones that pass clean are set to highest offered speed or full cache, while the not quite so good ones are brought down a notch. Also happens for GPUs, hard drive platters, and even resistor tolerances.

Sometimes people figure out tricks to unlock everything (with the caveat that the company sold it to you that way for a reason), but who knew Intel would sell their own tool hacker tool?

Re:I'm all for it

[derGoldstein]

I'd think Apple would be fine with this. Apple is already in the habit of charging small amounts for what would often be considered just "good will" on the side of the company (in terms of software updates), so Apple users wouldn't really be surprised by the request to "Pay to Unlock", especially if it's just $50 for what's effectively a hardware upgrade. Within the Apple ecosystem, this would seem more natural.
As for the privacy implications, I don't see why Apple would be *more* concerned than anyone else. They're not Google -- in the sense that privacy isn't a sore subject for them (maybe I'm just not up to date, but I don't recall any privacy slip-ups of that type from Apple).

Re:I'm all for it

[Runaway1956]

How long until the serial numbers are spoofed? This reminds me of the Pentiums with those identifiers being broadcast to the internet. It didn't take long for those to be disabled, and ultimately, Intel decided it was the wrong thing to do. http://www.wired.com/politics/law/news/2000/04/35950

Re:I'm all for it

[FatdogHaiku]

What, people won't line up for the assurance that their CPU is "Intel Genuine Advantage" approved?

Re:I'm all for it

[Keen Anthony]

You're not thinking about the marketing opportunity Apple has with this. This isn't like Apple charging customers extra for 802.11n. Apple had an acceptable reason. You sell me a chip that is intentionally handicapped, and then tell me that for $50 more, you will unlock it; as a consumer, it's fully reasonable for me to think you're taking advantage of me. It doesn't cost you anything to sell me that same product not handicapped, so why do this if it's not just to make more money? This is about perception, not logic and not facts. Why not just sell me a fully working product the first time? I can understand having to pay extra for MPEG-2 support in a piece of software or hardware, as there are licensing issues with a third party, but you're Intel. This is your chip.

Apple could bank on this by not using these chips and then running a new Mac vs PC ad wherein PC talks about all the great things you can do while going over a list of upgrades you have to unlock along the way: first to Windows so you can have all the features, then the CPU, and then toss in a joke maybe about GPU and RAM upgrades. Again, it's perception.

Re:I'm all for it

[JorDan Clock]

That's impossible to pirate

I think I've heard that line before.

Re:I'm all for it

[Genocidicbunny]

Someone goes and figures out the algorithm used, makes a piece of software that tell you your unlock code and bam, you have your unlock code, free.

Re:I'm all for it

[Dachannien]

This isn't the same thing, though - these are perfectly good chips that are crippled so that Intel doesn't have to manufacture chips at multiple price points. Maybe their reject rate has dropped enough that it's not a viable way to get lower-performance chips.

Re:I'm all for it

[Shadow of Eternity]

As long as someone has physical posession of the device there is no such thing as absolute security.

Re:I'm all for it

[sumdumass]

Your also forgetting that in some cases, they intentionally declock or degrade chips if inventory is particularly low on some budget offering. Like you said, they don't have separate runs for a 1 gig chip verses a 1.2 gig chip (simplified to keep the model numbers from creating confusion), it's all one given architecture then fixed to what is needed or what it will be the most reliable speeds.

Also the parent is a bit wrong on the ideology of this. It's not Intel or whoever selling a crippled part at the same price as a non-crippled part then asking for the $50. It's more like offering the part at a cheaper price then being able to upgrade it in place. You will pay the extra $50 whether it's in the original all there part of uncapping the crippled part you got at a discount.

Re:I'm all for it

[thePig]

Here is the algorithm
unlock_code = random.random()

Pretty difficult to break - eh!

Re:I'm all for it

[mysidia]

Which is why I said code in a database. As in, the serial number is not related to the unlock code by an algorithm, it is simply randomly generated every time and stored in a database, that is, there is simply an entry in some database on Intel's website that says "This unlock code goes with this serial number"

Without access to the database, you don't know any of the unlock codes.

This leaves the only form of piracy being the type that already happens with manufactured goods... a contracter secretly manufactures more goods than they are supposed to, or digs defective units out of a dumpster out back after hours, and sells them off on the black market, for profit.

Re:I'm all for it

[ImaLamer]

Agreed - there could be spoofing all around in an isolated environment.

Re:I'm all for it

[mysidia]

Piracy of crippleware, means users download a serial or crack from a third party, the 'pirate', who doesn't have possession of the software or hardware device being cracked.

If this comes down to physical manipulation of the device, and someone has to break open the case on their CPU or uncap it, to attempt to implement a physical attack, that will stop 99% of the population.

Because the CPU is easily damaged, and if it's damaged, the whole point of trying to upgrade for free is spoiled.

And the risk of damaging the CPU is so high, that it is unlikely to be a success.

At this point, one would call "unlocking" it by physically manipulating the device a "hack".

Not piracy. Piracy implies distributing things to the masses, such as a tool that can be used to easily turn on the restricted functionality.

Re:I'm all for it

[bertok]

As long as someone has physical posession of the device there is no such thing as absolute security.

But good enough security is quite doable. If you had to take your CPU apart and put it in under an electron microscope to read out a unique code that is essentially a one-time-pad, then it's not going to be worth it for anyone, particularly for a budget CPU.

Re:I'm all for it

[Cylix]

I can't find any documentation on the matter, but I'm willing to bet the PSN was stored in microcode.

I have great difficulty fathoming a process that can dynamically alter the design during fabrication. It's not like it's flipping a wheel on a serial number and punching some metal. (Well.. it's not too far from that). However, as much as I am interested in the process by which they used I am not dedicated enough to find some material on the matter.

In regards to the processor upgrade there are several methods they could invoke depending on how secure they want to make the process. However, given the price point isn't exactly the largest of numbers I'm going to wager the method will be something fairly inexpensive to implement.

Re:I'm all for it

[mwvdlee]

Indeed.

Turnkey CPU upgrades work in the mainframe world because IBM actually actively monitors and maintains those mainframe systems, so they could detect it.

That would be impossible in the desktop world without significant privacy issues.

Re:I'm all for it

[Cylix]

Reading over their document it looks like a fancy system for downloading a bios update.

Time will tell of course.

Re:I'm all for it

[kenj0418]

I don't see the problem. So you'd feel better if they instead spent more money on manufacturing, etc. to have two different CPUs. So you'd rather pay $250 or $200 for your choice of CPU than pay $175 for the single CPU with an option to upgrade it for $50?

* all figures generated via proctonumerogenesis

Re:I'm all for it

[slashtivus]

I presume mysidia was thinking of the concept of the 'One Time Pad' stored in a database:
http://en.wikipedia.org/wiki/One-time_pad/URL

Re:I'm all for it

[Decker-Mage]

It really depends on whether they use an algorithm to generate the unlock code or just a one-time pad for each cpu/serial number combination. With the former approach, it will depend on the strength of the algorithm and seed but given enough unlock codes the crackers will eventually break it as was proved recently with HDCP, which came from Intel! The latter approach is unbreakable with even a quantum computer if the one time pad is changed for each cpu/serial number combination (think each individual chip has its own public/private key combination). You'd have to have some incredible luck to get the key or some method to hack the silicon (which is also imaginable). However, then you'll see a shift of targets from the cracked algorithm approach to attacks on the database. I know of more than a few people that can pretty much carve their way into any protected system, and have seen them prove it, so then it comes down to how robust the protection and whether anyone that is capable is interested enough to do it.

I personally believe this is a stupid move on their part as its just waving a red flag in front of the black community. And, unless it's somehow hyped at time of sale, I can't see many people springing for extra 'features' they don't have a dim understanding of in the first place. Those in the real know will avoid these chips like the plague,; those not in the know what know what they are missing. I could be wrong on that call but I will be advising everyone, client or no, to avoid these chips. Since I design, build, and optimize systems of devices (not just desktops, for instance), people do listen to me. I think that is also true of people here (the technologically engaged ;-) and the places this is popping up on.

Re:I'm all for it

[mysidia]

How does the cpu know you provided the right unlock code if it cannot be calculated from the serial number?

Burn an 'unlock circuit' on the CPU that incorporates the knowledge of the code, and performs the unlock operation only when the input matches the code exactly.

Basically, a unique machine on each CPU that recognizes its code, but does not provide any access to the content of the code (without someone ripping open the CPU and analyzing the circuit with a $20000 logic analyzer).

Re:I'm all for it

[jackbird]

If only you had some CPU cycles nearby to brute-force it...

Re:I'm all for it

[plague911]

Not that I disagree with you. But your same argument can be made for any software anywhere. An additional user having a piece of software does not mean the original owner no longer has it.

Re:I'm all for it

[plague911]

Except that is what Intel and AMD have been doing for years and years. I do not know if you follow CPUs very closely, but both firms sell special chips with "unlocked" multiples.. However the only reason why they are locked in the first place is that Intel or AMD did extra work to lock it... Completely shifty on their parts.

On a side note I there should be some kind of federal law prohibiting companies from intentionally damaging their products to sell it to a different market.

Re:I'm all for it

[plague911]

What is to stop you/a program from brute forcing the code......? Its your cpu. You've got all the time you need.

Re:I'm all for it

[ZerothAngel]

The "unlock code" could be the public key of an asymmetric encryption algorithm.

Just having that wouldn't necessarily allow you to unlock the CPU since it would probably need to see the correct set of bits signed by the corresponding private key.

Re:I'm all for it

[MichaelSmith]

Using a very long code I suppose.

Re:I'm all for it

[mysidia]

I know of more than a few people that can pretty much carve their way into any protected system, and have seen them prove it, so then it comes down to how robust the protection and whether anyone that is capable is interested enough to do it.

Yeah, and I probably know people who could hijack truckloads of Intel's processors on the way to retailers, if they wanted too. Either way, it ceases to be mere piracy at that point, and more a matter of a break-in.

The interesting thing about a database like this, is by the time it goes into production, no person really needs access to it, it is so simple it can be made practically zero maintenance, and it can be chopped up into redundant pieces.

It's a perfect use case for a database which only supports two operations once built: "add serial number, code pair", and "lookup code by serial number"

This could be exposed over the intranet to the fulfillment server using a DNS service, with rate limiting and IPsec encrypted response; with new codes submitted over an out of band serial link shared by the database servers and manufacturing equipment that generated the randomized code assignment from a true random number generator.

No need for any complex database systems that might be susceptible to software-based compromise. No need for any remote administration services, at that point the question becomes one of physical security.

Re:I'm all for it

[mysidia]

If only you had some CPU cycles nearby to brute-force it...

The circuit could easily include a counter and disallow all further attempts after the 3rd try, until the CPU is power cycled.

I would think there would be a 'fixed portion' and a 'variable portion' of the circuit, with the variable portion set the same way as the clock multiplier (by burning certain solder points at the final stages of manufacture); and the fixed portion being buried out of reach without opening the CPU.

The fixed portion could contain the 'protect' fuse, to disconnect the circuit during times when a brute force attempt is detected, and (perhaps) re-enable it when power cycled, or if the manufacturer were evil, they could have 10 fuses representing the 10 attempts allowed, and burn a fuse every time an incorrect code is entered, so the CPU can no longer be upgraded after 10 failed attempts.

Re:I'm all for it

[muridae]

You know that you probably bought a handicapped chip already, right? Chances are a good portion of the economy cpus out there had a core or two disabled just to meet a shipping quota and price point, not because the core failed an integrity check. So, Intel spends some money making the Q/A test disable cores when it needs more chips in the economy bin and less in the high-end one. This just shifts the market a little. Now, instead of disabling a core by frying it completely, they just lock it in firmware. You, the end user, still get your economy priced chip. If you decide to upgrade you just buy the software to unlock it.

This is not some software that works the other way around, you know. The chip you buy isn't going to say "4 Cores and 32Mb cache" and then show up as 2 and 1 meg. The box might tell you, instead "2 core (upgradeable to 4)". The computer upgrade goes from being something geeks know how to do, to something any mom and pop and uncle bob can do. If they can get past the perception of it, no big deal. However, most of those people have no idea what a computer upgrade requires, and telling them that you can do it with software is something they have suspected anyways.

The real problem is that it becomes buy computer -> download spyware, games, p2p crap -> oh no's it is slow -> buy upgrade card -> more crapware -> "why can't I buy another upgrade card?"

Re:I'm all for it

[Keen Anthony]

Sure, I have. But I don't think about it much. People might react differently when seeing this product in stores. I'm generalising obviously. Perhaps the types of consumers who buy chips will also understand that they actually got what they paid for. As you said: a 4 Core with 32MB cache, and not a 2 and 1MB. But I don't think that's a safe assumption anymore.

Re:I'm all for it

[dargaud]

Pirated or jailbroken, that is one CPU I will not buy. Intentionally holding a gun to the customer's head by crippling the device until you pay a ransom is not a way to get my business.

I completely agree. This 'method' of doing business has been going on for a long time in the digital spectrometer world and mainframe world. I find it revolting and for the period when I had some decision power on what was being purchased I made it very clear to vendors that I would never consider their equipment for that very reason. Fortunately (for them), I'm back to lowly coder now.

Re:I'm all for it

[psergiu]

Unfortunatelly the Mac vs. PC ads are no more. Apple discontinued them :)

Re:I'm all for it

[doublebackslash]

Look up e fuses, that'd be how they do it. Have 128 efuses on the CPU for the "code" and one efuse to disable the commands to "blow" the efuse. That way you can't blow all the fuses in the field and then input all ones (or zeros, depending on how the CPU treats a blown fuse) as the unlock code. The serial number on the CPU is then effectively a nonce for looking up the proper code in their central database (that or fed into a Message Authentication Code [MAC] since it's easier to calculate from a secret than to maintain a large reliable database, but I digress). They've got the technology, and they seem to think it's worth the effort. Then again, they thought the same thing about the PIII serial numbers...

Re:I'm all for it

[koreaman]

I imagine it'd cost more to make the circuit on a CPU different for every single CPU coming off the line, as opposed to now where each chip (of a given model) is identical.

Re:I'm all for it

[CyprusBlue113]

Tell that to the DirectTV hackers. For high enough values of difficult, even all people will move on to easier targets.

Re:I'm all for it

[hairyfeet]

Uhhh...wrong. Those "unlocked" chips? Are the highest of the high performing and are sold to those that are gonna seriously push the chip. As for why they are locked? I can answer this quite easily, because in the old days of the Duron and Athlon first gens I saw many a dirty dealer OC the hell out of a chip and sell it as a higher performing PC, then the customer would wonder 6 months down the line why their PC became unstable so quickly. At least with only the higher end chips being speed unlocked you know what you are getting OOTB.

So I'd say it is quite different. With the former one can still try to unlock, at least with AMD, but you know by doing so you are simply rolling the dice as the CPU may be capable of higher speeds or more cores, and maybe not. Whereas in this case you aren't allowed to do squat unless you cut Intel a check, and BTW it only works on windows 7, unlike the former which works just fine on WinXP, BSD, Linux, anything I care to run. So I'd say it is a big deal.

Re:I'm all for it

[mysidia]

They're not identical. In the final stages of manufacturing, right before they cover it, they etch the chip with a laser to disconnect certain solder points in order to lock the clock speed, to the speed they are going to print on the box. And they also burn a serial number somewhere in the CPU's ROM/microcode.

It is probably nearly a trivial matter to add 32 or so more points to etch, in order for the 'unlock' code to be recorded.

It may be more expensive to implement the procedure on their industrial equipment, but it's basically a one-time cost per line, to secure their upgrade revenue...

Re:I'm all for it

[orkysoft]

The FAQ gives me the impression you can buy scratch-off cards with a unique code. It says nothing about the card being tied to a particular chip, just that it's only good for one upgrade. This makes me suspect the upgrade program reads the code, contacts Intel to check if it's valid and not used yet, and if it gets the okay from the Intel server, performs the upgrade. If true, that would make it easy to crack.

Re:I'm all for it

[orkysoft]

On second thought, maybe the "performs the upgrade" step does involve looking up the CPU's serial number, and querying the Intel server for its unique unlock code...

Re:I'm all for it

[CarpetShark]

The interesting part is that this is a chip company selling cheap, low-performance chips, but confirming that they can run at a higher spec if you're into overclocking and similar hacks. I've generally avoided overclocking as pushing a chip beyond its specifications, but if the company that makes it is happy to hack it, I am too.

Re:I'm all for it

[CarpetShark]

I can understand having to pay extra for MPEG-2 support in a piece of software or hardware, as there are licensing issues

You're too understanding. Cooking recipes cannot be copyrighted. Neither should video decoding recipes.

Re:I'm all for it

[CarpetShark]

One could argue, that Intels manufacturing yields and quality control, are so good, that they no longer have a significant quantity of chips that "fail" and are sold at a lower speed.

In other words, capitalist market forces (i.e., consumer demands) have worked to bring the price of goods down (i.e., supply).

Therefore, in order to extract as much $$$ out of the market as a whole, they are marking the high quality chips at lower speeds, and then marketing the unlock service to bring it up to full speed.

In other words, Intel have decided to bypass the central guiding principle of our economy, in order to make more profit.

Re:I'm all for it

[dotancohen]

Especially since it'll likely be pirated before the CPU ships.

That's the idea. People are already pirating software, so let them pirate _your_ hardware as well. At least, let them think that they can.

Why would Joe Average buy a $100 AMD chip when he can buy an equivalent $110 Intel chip then "pirate" it to be equal to a $200 Intel chip? This is genius.

Re:I'm all for it

[robinvanleeuwen]

"So you'd rather pay $250 or $200 for your choice of CPU than pay $175 for the single CPU with an option to upgrade it for $50?"

You'd rather pay $225 or $235 for a crippled one and $50 for an upgrade than $250 in the present situation. Because that is probably the way it's gonna play out. Not the 'companies have morals and business ethics'-way you sketched above...

Re:I'm all for it

[robinvanleeuwen]

Besides that. If i have $200 to spend i want to spend $200, and not $200 and than over the course
of the next year be harassed with the choice to spend another $50 to get the full-power i know
my computer is capable of... Since i know it's in there, and if I don't know i'm sure Intel is going
to make sure i'll be nagged with commercials, offers and unwanted advice from computer-savvy people
who 'do know what they are talking about' (read computer salesmen) etc. to make sure I do know...

There is exactly ONE reason why Intel wants this. That is to make more $$$ and those $$$ are coming out of OUR pockets. And we will be the ones that end up with less $$$ than in the present situation... So don't pretend it's a benefit to us...

 

Re:I'm all for it

[mysidia]

That would be secure... on the other hand, if they just contact the intel server and send the CPU's serial number and get back a generic command "OK, you can unlock the CPU", then that is hackable, by reverse-engineering the program, and determining what instructions to send to the BIOS to tell it to unlock the proc.

My suspicion is that at this stage, it is very insecure; I doubt Intel has designed this from the ground up to be secure, just for a little "trial" of the concept.

Most likely in reality it will be found that their current implementation is very hackable, if anyone bothers to look... at least that is my suspicion :)

Re:I'm all for it

[korean.ian]

Except Joe Average isn't going to be pirating the chip, as this will probably require a little more effort than downloading music or video.

Re:I'm all for it

[korean.ian]

And since Intel doesn't operate as a monopoly or as part of cartel (at least i don't think Intel and AMD are in collusion, lol) if the market works as it's supposed to, this will bite them in the ass until equilibrium is restored.

Re:I'm all for it

[xtracto]

Hehe, just what I thought...

My first thought after reading this was ... "sweet! now we will be able to buy $50 cheaper CPU's which we then can crack".

Really, after looking at the state of the most famous and famously cracked computer hardware (Wii, Xbox, Playstation) companies *still* continue to push these kind of locks...

Oh, and this also reminds me a story I heard about some hacker unlocking "hidden" cores of a multi-core CPU. IIRC the story was that when a core of a processor made by Intel/AMD does not pass quality checks, they only disable such core and sell the processor as a cheaper version (e.g., a core i7 with failed cores can become a i5)
.

Re:I'm all for it

[Dr_Barnowl]

Vendors will hate it. They'll get an increased rate of support calls about it, and none of the benefits, because the fifty bucks goes to Intel.

Re:I'm all for it

[robinvanleeuwen]

Those locked chips are locked to ensure they are not going to go haywire. When computer chips are made say 1800Mhz they are always going to perform reliable at speeds around that speed. Say 1790 and 1810. The ones that come off the production line that only work reliable at 1720 are sold as 1700Mhz chips. The ones that perform reliable at 1890 are sold as 1900Mhz chips.

Nowadays with multicore chips they have the option to bake a CPU with say 4 cores. If a core doesn't work reliable due to production issues, which always some chips do, they have the choice to either toss them away, or sell them with the flaky core locked out as tripe core CPU's. Some times the locked core works fine with regular home use, but not in stress tests at the production facility. If that is the case you can unlock the core at home and hope it doesn't go haywire with intensive use...

Has something to do with the way they are made. On a wafer. With dozens of CPU's at once. When producing a wafer always some CPU's are broken from the start, some work perfectly and others work perfectly but only at lower speeds... It's not deliberate crippling it's a production process with a margin of

That's why every CPU can be overclocked but always to different speeds. If you bought a 1800 cpu that performed reliable at 1825 but was still sold as an 1800Mhz cpu, you'll probably can overclock it with extra cooling to say 1895. Another which only worked reliable until 1801 you can probably only overclock reliable to 1870.

It's not knowingly crippling good CPU's it's a consequence of production process that not all chips are 100% of what they are intended to be. But that doesn't work in the store. They can't sell CPU's to customers and say. Well here's a CPU and say: "it works at 1800-ish MHZ". They have to ensure that if they SAY it's a 1800Mhz, it will work at minimal 1800Mhz.
   

Re:I'm all for it

[jo_ham]

Just because you believe that to be the case doesn't mean the current reality will reflect that. You can't just choose *not* not pay the licensing fee for a mpeg2 decoder from Apple (either rolled into the cost of the OS or in the Quicktime Pro fee) because they *do* respect it.

You can, of course, get the mpeg2 encoder from other sources but we're not talking about that in this instance.

Re:I'm all for it

[winwar]

"In other words, Intel have decided to bypass the central guiding principle of our economy, in order to make more profit."

The last time I checked, the central guiding principle of our economy was to make more profit.

The fact of the matter is that their slowest and cheapest chip of the current architecture is faster than most people actually need. But you have to pay for the research and development in some manner. And the expected profit. Someone is going to get gouged.

Re:I'm all for it

[Erikderzweite]

Storing hash, maybe?

Re:I'm all for it

[vadim_t]

Nobody said anything about not paying. What should be done is changing the law to make software unpatentable, and meanwhile developing alternatives like Theora and VP8.

Re:I'm all for it

[icebraining]

if they just contact the intel server and send the CPU's serial number and get back a generic command "OK, you can unlock the CPU", then that is hackable, by reverse-engineering the program, and determining what instructions to send to the BIOS to tell it to unlock the proc.

upgrades.intel.com 127.0.0.1

Re:I'm all for it

[lennier1]

You don't even have to go back that far.
Just remember the AMD chips where you only needed to add a line with a pencil to allow overclocking.

Re:I'm all for it

[Bert64]

There was never meant to be any way to unlock an iphone other than with the unique device specific code Apple use, and yet third party unlocks do exist.
There was never meant to be any way to unlock a ps3, xbox or wii to play homebrew or pirate games, and yet there are many ways to do this.
The HDCP master key just recently got leaked too, Intel designed that spec if i recall... Who's to say the processor unlock codes won't get leaked too?

Re:I'm all for it

[Bert64]

That assumes that easier options exist, which in the case of broadcast TV they do...
There are hundreds of different TV services around the world, all it needs is for one person to find the weakest service, rip the show and upload it somewhere.

Re:I'm all for it

[therealmorris]

Problem is, if they're selling these unlock cards then the chips can't just be disabled "when it needs more chips in the economy bin and less in the high-end one", they must be using better chips and artificially disabling every single one, or else how would anyone know if the card would work! That I think is the issue here.

Re:I'm all for it

[TheRaven64]

When computer chips are made say 1800Mhz they are always going to perform reliable at speeds around that speed. Say 1790 and 1810

You can say 1790 and 1810, but it's more likely to be between 1600 and 2000. The chips come of the QA line in that bucket and are then stamped with a clock speed depending on which part has the largest demand. This isn't even a new thing. Remember those 300MHz Celeron As? The ones that had a 66MHz FSB, but all of which ran stable at 450MHz with a 100MHz FSB? They were just sold at a lower speed because Intel didn't want them to compete with the 450MHz Pentium II (same FSB speed, same core, just a bit less cache, at more than twice the price).

Re:I'm all for it

[Joce640k]

these are perfectly good chips that are crippled

So are most of the other CPUs you've ever bought/used. The only difference is that the 'locks' on these aren't soldered in place.

Re:I'm all for it

[rogermcdodger]

It isn't just economy CPUs. All Xeon 5600 processors are 6-core parts. The 4-core models just have two disabled. I'm sure if enterprise users could just pay the difference later on to upgrade they wouldn't be complaining.

Re:I'm all for it

[Joce640k]

It really could be... eg. if you have to digitally sign the chip's serial number with a private key then the only way is to break into Intel's headquarters and do some safe-cracking. You're not going to be able to open up the chip and rewire it.

Re:I'm all for it

[Anonymous Coward]

Sure, you will probably just have to short circuit two parts of the chip to set the bit to unlock it. All you have to do is remove the cpu case (Carefully!) and connect line 259 and 2051 with your 40nm probes. Don't push to hard or you will damage the chip.

There is no such thing as absolute security. But this is not necessarily a case where the value you get by hacking the device ougtweights the cost of doing so.
Absolute security is never worth the effort. Enough security is probably what you want and for that even obscurity will do the trick sometimes.

Re:I'm all for it

[Joce640k]

Scratch my own post, that won't work...!

Re:I'm all for it

[roman_mir]

and if there is no check whether to 'skip' some part of the chip, but instead the entry inputs are scrambled and encrypted with a key, which is a 'one time pad' key and an input without that key will not yield the correct computational output values then you can't just 'skip a checkpoint', you actually need a one time decryption key as part of the input to generate execute the code correctly.

Re:I'm all for it

[Joce640k]

You're right of course ... but they could easily make it so the algorithm is different for every chip. Easch chip could have a serial number and a random AES key built into it. Only the serial number can be read out, the AES key is hidden.

To unlock the chip you have to provide a number which is the serial number encrypted with the AES key.

Re:I'm all for it

[captaincinders]

It has happened for many years, and is all part of a marketing strategy (and every time I hear the phrase 'marketing strategy' I immediatly think 'scam'. 20 years ago I bought an external hard drive for my BBC micro. Cost of a single sided drive was about £120, or I could upgrade to double sided for £150. Could not justify the extra price so went for the single sided. Being electronic students, the very first thing we did when it arrived was to take it apart. We found a disconnected wire, so we plugged it in to see what happened. Hey Presto - a double sided drive magically appeared. So £30 of the cost of a double sided drive was purely marketing hipe and setting different price points - and was also a con that put £30 straight onto their profit margin. Yes I know it is not illegal at all, but was the first time I realised that business had a set of morals that differed from mine. As to 'not doing business with them', they probably got a lot more business because of me because the word soon got around (with the help of a instructiion manual) of how to buy the very cheapest double sided drive around. Why did I send busness their way? Because I reasoned that I did not care what their business practice morality was, I got a bargan for my money, and so did everyone else who bought it. What do I get for my money is the only question.

Re:I'm all for it

[mysidia]

iPhones, PS3s, Xboxes, and Wiis all have something in common. They are programmable, and being "locked" or not is a software state. The security of those is in software, requiring cooperation of all programs on the system, and if even one software program has a bug, it can result in running arbitrary code, leading to the ability to run arbitrary software.

The attack surface in regards to those platforms is huge and massive compared to the security required to block access to something physical from software.

The 'attack surface' could be as small as one CPU instruction for submitting an unlock code to the 'unlock cores' circuit.

It doesn't matter what great microcode hack you have, if the extra cores are always held in a reset state or power off state by design, until a fusible link is broken by the blackbox security circuit, then there is no "circumventing the security circuit", without either providing the proper input, or by physically manipulating it (which is hard).

If there were a unique code for each CPU, the massive number of CPUs manufactured would make gaining access to, let alone 'leaking unlock codes' quite a feat.

This data could definitely be encrypted so no one team and no one server would even be able to extract a single unlock code, let alone all of them.

A typical setup is to have 3 types of servers... DB server, Transaction Processing server, and Notary server.

DB server has the code, encrypted.

Transaction processing server, queries the database server for an unlock code.

Transaction processing server gets the encrypted packet, sends with signed request to the notary server, gets back approval, and the other half of the key needed to decrypt the code.

Note any 2 of the 3 servers could be compromised, and it would not lead to any loss of unlock codes.

In such cases, each of 3 classes of server have their own independent security precautions, separate management team, and strong isolations and monitoring.

Note while leaking is in theory possible, it requires the collaboration of 3 different teams.

Re:I'm all for it

[Attila Dimedici]

There would either be a way to reverse engineer the algorithm used to generate the number or a way to read the number off of the chip (or both).

Re:I'm all for it

[stasike]

How long until some virus simply attempts to "unlock" the processor 11 times?
Just for the heck of it.

Re:I'm all for it

[CarpetShark]

The last time I checked, the central guiding principle of our economy was to make more profit.

Profit is a variable within the system of money. It cannot be the exterior goal for which the system was created. I think you're confusing how a buggy system is working out after many iterations, with its intended purpose.

Re:I'm all for it

[digitalhermit]

This is nothing new.

Oracle licensing is based on CPU performance, number of CPUs, etc.. IBM Power systems can selective enable/disable processors so as we need them, we turn them on. This gives us some flexibility in purchasing as we don't have to purchase until we need the extra resources.

Re:I'm all for it

[Joce640k]

Nope. The chip needs all the information in order to be able to verify the key, this means the private part of the key would need to be hidden in the chip and somebody would be able to figure it out eventually.

Re:I'm all for it

[itsdapead]

Apple is already in the habit of charging small amounts for what would often be considered just "good will" on the side of the company

ISTR with the iPod Touch update (the one that magically endowed the Touch with Bluetooth) a few years back, Apple claimed that "the accountant made us do it" (and I think it was around the time they were having trouble with the SEC). Sadly, I find that completely believable (the accountant didn't even have to be right).

Re:I'm all for it

[index0]

But now spyware and related software can now cripple your cpu without pegging your cpu at 100% usage.

Re:I'm all for it

[index0]

No way to share an algorithm? No way to run a key generator?

Re:I'm all for it

[itsdapead]

I suppose the "price above all else" PC vendors will like this. If they can load up machines with demo software, why not have demo hardware to match?

...and that's the real problem with this idea: its anti-consumer because it creates more opportunities for vendors to offer misleading sticker prices and cash in on "hidden extras".

Its rather different in the old mainframe world where the price was as much about service and support and the whole arrangement was probably closer to leasing than buying.

Re:I'm all for it

[Pharmboy]

Even in the 486 days, I have seen plenty of 25mhz chips running at 33mhz, which is a significant step up in speed, 50%, and sold as 33mhz for the extra money. There was a lot more hokey stuff going on back in the early 90s by local shops. For them, ignorance was bliss.

Re:I'm all for it

[Lord Byron II]

If you're referring to when Apple sells updates for a small amount ($5 or so), that's to satisfy accounting rules, which as a publicly traded company, they have to obey.

Basically, if they sell you a computer and then a year later give you something of value that improves that computer without charging for it, then it assumed that the upgrade was part of the price that you originally paid. In this case, they can't consider that part of the original purchase part as profit until you receive the upgrade. It gets really messy, real quick and it's a lot easier to just charge a token amount later on.

Re:I'm all for it

[flyneye]

Of course the same thing happens with automobile manufacturers, but you need a computer contraption to unlock the computer potential of the car in question. Then you can go drive that fully juiced Trans Am up a utility pole.

Re:I'm all for it

[Vectormatic]

using e-fuses, they could easily make mechanism which disables the upgrade entirely and irriversibly after say, 5 wrong attempts

Re:I'm all for it

[hedwards]

That's not equivalent. They couldn't sell those downclocked chips if they had to sell them at the full clock rate because the reliability would be crap. Some people are comfortable taking on more risk or replacing the chips more frequently and will overclock.

This is charging you for something that you've essentially already paid for. You've paid for the chip, and if you don't pay for the unlock, Intel doesn't get a dime. Consequently it follows that this is just a way of juicing the earnings at the consumers' expense.

Re:I'm all for it

[itsdapead]

But your same argument can be made for any software anywhere.

Software is licensed, not sold (yeah, I know, but sadly the law seems to allow this stupid end-run around the clear intent of copyright law) - this sort of thing is, in effect, extending the same licensing blight to hardware.

This gets interesting when the inevitable free alternative appears and Intel try to stop it. Then I guess it will come down to whether the "unlock" software is substantial enough to be copyrightable and/or whether the DMCA applies.

It was all rather clearer in the mainframe case where customers would have a proper contract with the vendor (rather than some take-it-or-leave-it EULA).

Re:I'm all for it

[turbidostato]

"You sell me a chip that is intentionally handicapped, and then tell me that for $50 more, you will unlock it; as a consumer, it's fully reasonable for me to think you're taking advantage of me."

Yes, but History shows that while reasonable, market doesn't tend to react that way.

I don't know if it's the case anymore, but both Windows NT 3.5 and 4.0 server's versions were just intentionally handicapped NT workstation versions to be unlocked by paying more, and it was plently known back in the day -why do you limit workstation edition to just 10 concurrent network connections? (I myself did the "trick" for testing purposes: some registry changes and a bit of hex editing was enough for a workstation version to be booted up as a server one, graphics work included). And still they were a great success.

Re:I'm all for it

[joshuao3]

You mean that Intel would have to maintain a database of a few hundred million unlock codes? NOBODY can store a database that huge! Sheesh.

Oh wait, I've got a copy of it on my thumbdrive.

Re:I'm all for it

[JamesP]

That's impossible to pirate, because there's no way you can share the code.
As long as they assure the unlock code is the only mechanism allowed to re-enable the capabilities, and there is no BIOS mechanism to override the lock.

Ha, they're probably doing this via Microcode update, which is not only commonplace (you only never heard of it), VERY piratable.

Re:I'm all for it

[MightyYar]

Yeah that was my first thought: "Well THAT isn't adhering to KISS..."

Re:I'm all for it

[Goaway]

So you basically know nothing at all about how Apple operates. They gain their popularity by not doing these kinds of incredibly annoying things other manufacturers do.

Re:I'm all for it

[jmv]

Of course. In this case, all you need to do to save $50 is to take the package off the CPU (without damaging the die), analyse the die with an electron microscope to find the key, and put back the package. Sounds like a simple way to save $50.

Re:I'm all for it

[betterunixthanunix]

More likely, Apple will sell an "upgrade" with "improved performance," and once again take advantage of their customers' general ignorance about computers. After all, a lot of people still do not think of the iPad as a "tablet," so why would they think anything is unusual with the "improvement?"

Re:I'm all for it

[tomhudson]

Except that one thing computers are good at is repetitive calculations - you can just let your "screen saver speed lock decrypter/decripplizer" try every possibility out there while you're not using it for anything else.

Re:I'm all for it

[commodore64_love]

>>>Apple could bank on this by not using these chips and then running a new Mac vs PC ad

Considering Apple Macs cost ~$1000 while my new Window 7 PC only cost $200 (sale price), damn straight Apple better not give me a handicapped CPU. Such things are okay for budget computers, but not for luxury computers that have 4-5 times markup. That would be like upgrading from a Honda to an Acura, and discovering I have to pay $50 extra to upgrade from FM to HD radio.

Re:I'm all for it

[nurb432]

Buying a chip that is underrated to pass QA is far different then selling you something with the intentional plan to charge you later to fully utilize what you bought. One is saving the scrap, the other is a scam.

Re:I'm all for it

[thygrrr]

This is actually how they did it for years - it costs the same to make a chip with more or fewer features, processes are so decent now they just deactivate parts in the chip's firmware and sell those for the high demand of low end chips.

They just made their firmware a little more flexible at what this so called "binning" can be made to do.

Re:I'm all for it

[ConfusedVorlon]

Vendors will _Love_ it.

Intel will cut the cost of the chip - knowing that they'll get the extra $50 in a good proportion of cases.

The vendor gets to produce a computer that is at a lower price point (customers are generally stupid and will buy an item with a lower initial cost and higher TCO which is why people buy phone contracts with 'free' phones).

My guess is this won't appear in the higher end systems, but it will be used a lot in low end systems where it is all about getting the total cost below some magic number like $300.

Plus, the vendor gets to say that the chip is 'Cache Upgradeable' (or some much more compelling marketing tag).

My guess is that the economics are something like this:

extra cost to make the more capable chip
$5
discount to vendor for using 'upgradeable chip'
$10
50% upgrade chip
$25 (average), which just about covers the admin cost (calls, processing, customer support) and the $15 additional cost

Intel breaks even approximately, but gains by selling more volume to delighted low-end manufacturers

Re:I'm all for it

[Registered Coward v2]

I presume mysidia was thinking of the concept of the 'One Time Pad' stored in a database: http://en.wikipedia.org/wiki/One-time_pad/URL

Actually, a one time pad would be no more effective than any other randomly generated set of keys. A one time pad is effective because you are coding a message that has no real clues to its content; with an unlock code you know you need to generate an x digit value that matches the one in the chip. Brute force would work; it's just a question of how long would it take. Intel could make it hard by varying the number of digits so that a brute force attack might take longer than the chip's useful life. By randomly assigning codes, rather than use an algorithm to generate them, is what makes it harder to pirate.Even if you knew what went into generating the codes and their makeup (length, use of special characters, etc.) it's the matching of the unique set that makes it hard.

Of course, in the end this is just a way to make more profits - companies sell machines at price points and feel that they make more money selling X machines at P1 and Y machines at P2 (where P2 has a faster chip) than they would selling X+Y at P1 with a faster chip. Intel could have simply delivered the chips and not offer an upgrade at retail; but they're trying to see if some group of Z customers will decide to upgrade and if it's worth the cost of selling upgrades. The danger is customers may decide to buy a lower priced machine and upgrade rather than the more expensive machine (assuming equal features in both otherwise) making the upgrade price the max price differential you can charge. You may see companies simply sell a base machine plus upgrade codes in the future - lowering their costs because they only need one base machine MB+chip for a wider range of performance.

Re:I'm all for it

[jedidiah]

> I can understand having to pay extra for MPEG-2 support in a piece of software or hardware

I was wondering when this would come up and when some fanboy would try to make excuses for this absurdity.

It's just bogus. MacOS can play a DVD. So obviously it has all necessary rights sorted out to deal with
MPEG2 and AC3 and ANYTHING else related to the fact that it is able to play a DVD for you "out of the box".

The idea that you have to pay extra to add any of these capabilities to Quicktime is pure nonsense.

It just shows that Apple thinks it can f*ck around with it's customers. ...and they're right too.

Re:I'm all for it

[Raenex]

Try reading the comment you are replying to: "Chances are a good portion of the economy cpus out there had a core or two disabled just to meet a shipping quota and price point, not because the core failed an integrity check."

Re:I'm all for it

[insufflate10mg]

I personally believe this is a stupid move on their part as its just waving a red flag in front of the black community.

Emphasis added. What's that supposed to mean?

Re:I'm all for it

[jedidiah]

Yes.

I would rather have simpler parts that aren't dependent on an upgrade procedure that could go wrong or could be seen as too complex by the average consumer.

Traditionally hardware has avoided a lot of the nonsense and shenanigans associated with software. Electrical engineering is a much more mature discipline and is something that you can reasonably call engineering (versus software development). Although now the nonsense associated with shoddy software seems to be seeping into the pure hardware side of things too.

If I want a faster part, I would rather pay for it upfront.

Re:I'm all for it

[jedidiah]

> using e-fuses, they could easily make mechanism which disables the upgrade entirely and irriversibly after say, 5 wrong attempts

That opens up some interesting possibilities for malware then...

Re:I'm all for it

[jedidiah]

> But most of the customers are not going to put their CPUs under an electron
> microscope and pay a reverse engineering specialist to extract an hardwired
> unlock code. It's not worth 50$, simply. ...except there will be someone on the planet that will have access to the resources and will do it for fun.

Re:I'm all for it

[Jeff DeMaagd]

It seems to me to be a means to unlock potential that was previously locked because of marketing demands. Chips aren't made for specific speeds, they make a batch and "bin" them based on testing.

Before, when you bought a 2.2 GHz chip, it was in a batch of chips that also happens to include chips that made the 3GHz bin, but they clock locked them to multipliers specific to 2.2 GHz. Sometimes those 2.2 chips were marked such because really weren't reliable at 3 GHz, but sometimes people found they ran perfectly fine at 3.0GHz, it's the luck of the draw.

I find this system preferable to having to replace the chip or computer in order to get a faster one.

Re:I'm all for it

[squiggleslash]

You don't have to go that far back. Windows 7 is entirely driven by the principle. That computer you bought comes with Home Premium and you need Professional? Just select the option from Add/Remove Programs (or whatever it's called these days), buy the updated key, and all those features that are "only in the Professional edition" suddenly become available. I believe the same applies to Office these days, the CD contains the entire version, the key defines what gets installed and what you're allowed to run.

Re:I'm all for it

[douglips]

Cooking recipes can be patented. Copyright != patent.

Re:I'm all for it

[zill]

Weird, I just cracked my chip on the first try. What are the odds of that.

Re:I'm all for it

[Smauler]

So when I'm sold a game, then the publishers release a patch, I've got to pay for it? Screw accounting rules.

Re:I'm all for it

[I(rispee_I(reme]

Hm, I must not have been paying attention. I thought they were popular because they candy-coated said annoying things.

Re:I'm all for it

[IICV]

One thing nobody's talking about is why Intel is doing this.

The only reason I can imagine is that they're sitting on some technology that will greatly reduce fabrication flaws, which means that far more chips will be coming out of their factories that are capable of running at full specification than the market wants.

That, or they're already outputting a high percentage of chips that are capable of running at higher rates, and disabling them - a much higher percentage than they used to be able to manage, if it makes sense to actually market these chips as upgradeable.

Re:I'm all for it

[Predius]

And yet, Apple is the only one doing that I've come across... weird.

Re:I'm all for it

[marcello_dl]

Posting to undo moderation. I didn't know it was possible to mod and post anon.

Re:I'm all for it

[Anonymous Coward]

This isn't the same thing, though - these are perfectly good chips that are crippled so that Intel doesn't have to manufacture chips at multiple price points.

Excuse me, but I fail to see the loser here.

Intel wins: it can focus resources on manufacturing a single processor for multiple price-points
Customer wins: for a fee, they can upgrade their processor without having to touch a single piece of hardware

Why the outrage? If you assume Intel doesn't have enough defects to bin into lower priced models, compare this to their alternative. They manufacture and sell additional processor lines in order to cover all their price segments. Consequently, fewer engineering resources per line, less optimization and likelihood of "free" clock-speed bumps, and higher manufacturing costs per processor (and therefore less downward pricing flexibility).

"But Intel isn't giving me all I bought!"
Bullshit. You *bought* the cheaper model. If you wanted the more expensive one, then you pay for it. If you don't like Intel's pricing, then you buy AMD.

Re:I'm all for it

[Kjella]

Well, I think there's a fundamental difference between "we're crippling your hardware" and "we're putting in extra hardware that you can optionally purchase". Putting in extra CPUs, extra sticks of RAM, extra disk drives that you won't get access to until you pay more is fine, they didn't have to deliver that. I think an extension of that is putting in more powerful hardware or bigger drives than you've spec'ed, because it makes economic sense. You wanted 500 GB, they figured it was better to put in a 1 TB disk and give you half. If you flip it around, banning that practice would mean they can only sell it to you in big increments, all or nothing. The customers don't really want that because there's no such thing as free hardware. They'd like to pay for only what they use and no more. Consider it "gouging" if you will but there's nothing stopping you from buying whole CPUs and RAM sticks and drives. Just expect fhat for 100% of the hardware you'll pay 100% of the price even if you only use 30%...

Re:I'm all for it

[arivanov]

So, what makes Microcode not software?

There is lost of precedent behind what Intel is doing. While Intel has cross-licensed most of their IPR I suspect there is at least some IPR which they are paying per core/feature as well so there may even be some financial basis.

While me or you may not like it (I definitely do not) there is nothing particularly vile or invalid businesswise in this.

Re:I'm all for it

[dotancohen]

That was the case until now. However, to appeal to the pirates market Intel invented this "upgrade". Exactly my point.

Re:I'm all for it

[vadim_t]

So far I've seen very little evidence of that patents do any good, so I'm in favour of eliminating them all completely. If I have to compromise I'd start with anything relating to software, math and business methods first though.

Re:I'm all for it

[hughbar]

This goes back even to the 1970s and 1960s. ICL, the British computer manufacturer would rent or sell a printer that was made artificially slow for one price, remove a couple of resistors and then the 'customer' had a faster printer at a new higher price.

Personally, I think that shows that the 'invisible hand' doesn't deliver more and more prosperity to the human race, something that the more enlightened economists such as Stiglitz are beginning to say aloud. This kind of activity shows that in many areas the 'market' is failed religion or illusion with lots of followers still clinging on.

Re:I'm all for it

[Reziac]

True enough, but I think it's going to backfire on them -- by driving down the price of those chips by the $50 everyone knows they'll have to spend to get what they figure they really paid for.

I think they'd do their market a lot better by releasing a free tool that would helpfully upclock CPUs by as much as the chip can handle (at your own risk, of course). Then people would feel like they got more than they paid for (instead of feeling ripped off), and that always results in good word-of-mouth.

People generally prefer a gamble to a forced payout.

Re:I'm all for it

[Deathlizard]

While I can somewhat understand the cost logistics of having an upgradable processor via software. what no one seems to be talking about here is the downgrading of processors.

For example, What stops someone from making a Ransomware that disables the full functionality of your high end processor that you have paid for and won't reverse it until you pay them $50? Now that I think about it, I'm amazed that this hasn't happened already, since hyper-threading and cores can be disabled at the BIOS level, but at least in those cases you clean the virus, reset the bios and your back to normal. A CPU lock or a BIOS lock that you can't adjust would make this permanent.

Then there's board replacement. Lets say you buy one of these systems and upgrade it, then the board fails. Since most likely this is a Bios lock, swapping the board would swap the bios and disable the chip again. does that mean another $50 bucks to Intel? I doubt that the program that unlocks processors could be used more than once.

It's things like this that makes me glad I bought AMD

Re:I'm all for it

[Haeleth]

This isn't the same thing, though - these are perfectly good chips that are crippled so that Intel doesn't have to manufacture chips at multiple price points.

Yes, like they have been doing for years. If demand for cheap chips exceeds demand for expensive chips, they cripple some expensive chips and sell them cheaply. This increases their profits and, by decreasing the complexity of the manufacturing process, also reduces the price of the expensive chips. It's a good thing.

The only new thing here is that they are now also providing a simple way for people who got one of the crippled chips to uncripple it. Which is also a good thing.

I was outraged when I saw the Slashdot headline too. Then I read TFA. Then I spent a few seconds thinking about the pros and cons of this. And suddenly I'm not outraged any more. I put it to Slashdot that this concept of "thinking" is a useful tool that ought to be applied more widely.

Re:I'm all for it

[commodore64_love]

>>>the markup is negligible.

No what you don't understand..... what you seem to not realize..... it that I'm a cheapass. I pay $0 for TV, $0/month for my pay-as-you-use cellphone, and only $15 for my internet. Why? Because I'm cheap.

For me the difference between a PC and a Mac IS significant, because I'm not going to spend $700 for any computer. I'll buy the lowest price I can find (which was $200 in this case). Also claims that Macs are worth $700 assume that I want the software they came with. I don't. All software I use is either open-source or free, so when I have the same capabilities as a Mac but still only spent $200.

Your arguments are reminiscent of how people claim the Civic Hybrid was worth the $22,000 pricetag. That's only true if you want the luxury. If you don't want the luxury, a standard 40mpg Civic can be had for $13,000... a significant savings.

Re:I'm all for it

[arth1]

Vendors will _Love_ it.

Intel will cut the cost of the chip - knowing that they'll get the extra $50 in a good proportion of cases.

If you really think this will make the not-unlocked version $50 cheaper, I have a bridge to sell you.

In reality, what you're looking at is:
The factory locked version becomes more expensive, because the extra costs of having an unlock system has to be paid for.
The factory locked + unlock version is $50 more than that.
The factory unlocked version also goes up in price, to a point just below factory locked + unlock.

This isn't going to fly -- it may float, like many unpalatable things do.

Re:I'm all for it

[I(rispee_I(reme]

I believe he neglected to append "hat" to "black", which makes for a demographic that can be slandered without fear of consequences.

Re:I'm all for it

[joeyblades]

Sorry to burst your bubble, but I've been in the semiconductor business for over 30 years and we've always been doing this. Manufacture one device with certain performance or features that can be disabled and sell 2 or more different versions. It happens with other kinds of technologies, as well.

You're confused about the cost part of the equation, though. Usually what happens is that the marketing gurus figure out how much market share you're going to get with each flavor and set the price accordingly. Often, if you only sold the low-end version you would not make enough profit. You count on some sales of the high-end version to offset the actual cost. It's a gamble and if your marketing team guesses wrong, and no one wants the high-end flavor, you lose money. If they guess wrong and everyone only wants the high-end flavor, then everyone gets a nice bonus...

I've seen it go both ways.

If the upgrade capability was zero cost, Intel would do exactly what you say. Just make that the standard; it's free marketshare. The fact that they had to get creative is a sure sign that it's not zero cost.

Re:I'm all for it

[GWRedDragon]

Just because markets produce some undesirable effects doesn't mean that they are not far outweighed by the desirable effects. Without the free market, you probably wouldn't even have a home computer in the first place.

Re:I'm all for it

[ConfusedVorlon]

did you read my post?

I suggested that the vendor might save $10.

They keep some of that, and after markup perhaps $10 is reduced from the end price allowing the product to edge below some important price point.

Of course I made the numbers up - but I never suggested for a moment that the non-unlocked version would be $50 cheaper.

Re:I'm all for it

[kc8apf]

Only if that patch includes new functionality or content and the company is publicly traded.

Re:I'm all for it

[BitZtream]

Chances are a good portion of the economy cpus out there had a core or two disabled just to meet a shipping quota and price point, not because the core failed an integrity check.

And most consumers have no clue about that, also, the extra core(s) is almost always broken in multiple ways, so you really can't use them reliably.

However, when they realize they have the full spec of hardware in working condition and Intel is just, basically, charging them a few to get access to the extra cylinders already in the car?

Thats probably going to raise a few more eye brows.

Re:I'm all for it

[BitZtream]

So you're comparing a chip thats broken and incapable of actually performing at the specified level ... to a chip that isn't broken and can perform at the specified level except they just told it not to, so they could get more money ...

Two entirely different things.

Re:I'm all for it

[martyros]

As long as someone has physical posession of the device there is no such thing as absolute security.

Yes, but they don't need absolute security. They just need it to be enough that the vast majority of their customers won't. If it's a choice between paying $50 and actually opening up the processor and doing some kind of surgery, the vast majority of people will just pay the $50. If a couple of hundred people get a free upgrade because they are willing to do it, who cares? You probably made more from them by just selling them the chip in the first place than you would have made otherwise.

It's similar to internet filtering in China. Sure, you can get around it if you know what you're doing. But China doesn't need to prevent every single person in the country from being able to read about some news story. Their goal isn't 1984-style "everyone will love Big Brother or else". Their goal is to generally keep social order. So as reading about riots in Urumqi or the massacre at Tienanmen Square are difficult enough to find that 98% of the population will never hear about them, that's good enough.

Re:I'm all for it

[linhares]

https://retailupgrades.intel.com/Page.aspx?Name=UpgradeBored? nothing to do? Call (408) 765-8080 and tell them you're trying to upgrade from your obscure *nix distro. And remember to check out the page to report problems http://www.intel.com/intel/report.htm?iid=intelfb+body_reportprob.

Re:I'm all for it

[Tacvek]

Happens all the time, actually, they usually just don't offer a way to unlock it. They make a run of all the chips of a given architecture, then put them through tests. The ones that pass clean are set to highest offered speed or full cache, while the not quite so good ones are brought down a notch. Also happens for GPUs, hard drive platters, and even resistor tolerances.

Sometimes people figure out tricks to unlock everything (with the caveat that the company sold it to you that way for a reason), but who knew Intel would sell their own tool hacker tool?

Indeed. That is true. That said, when demand for one of those lower performance chips is high enough the companies will sometimes sell chips without the defects. They often artificially cripple the chip when doing so, but on rare occasion they sell the better chips non-crippled as though they were the cheaper chips hopping nobody will complain. (That said, if you had a non-crippled better chip in your test system, that could be a problem, since you would get better performance than you could actually count on.

This is the first time I have heard of selling artificially crippled CPU's in their own tier, where none have defects, allowing them to be unlocked via software.

Re:I'm all for it

[Tacvek]

The downside here is that they cannot sell the chips with minor defects that necessitate crippling in with these artificially crippled chips, since they are offering the ability to unlock the full performance. That means they must be selling the chips with minor defects at an ever lower tier, along with more artificially crippled chips that they are not allowing unlocking for, or they are throwing away the chips with only minor defects.

Re:I'm all for it

[marcello_dl]

It's different because what you pay for in the software is the work of the programmer. In the chip, while you pay for the big development costs, and for the PHBs that pollu er manage the corporation, you pay also for the physical item itself.

So the message Intel sends is a problem for two reasons:
  - it tells the client that it is not striving to give you the best price. If it was the best possible price they could honestly make, they would be at loss each time you buy a crippled chip and you don't upgrade it.
  - it make people acquainted with the fact that you buy something over which the producer still have control.

In general, it is true that they likely disable chips. But that doesn't make this behaviour acceptable. It still is a problem for the first reason.

Re:I'm all for it

[mysidia]

If the unlock sequence is done from BIOS only, the BIOS could be required to send a DISABLE_UNLOCK before booting an OS, causing the 'unlock' capability to be disabled until CPU reset.

Intel already does this for the VT function. Once the BIOS has sent a VT LOCK instruction, the feature is incapable of being turned on until the reset line on the CPU has been activated.

The BIOS is then a security barrier, and the standard virus protections apply. If a virus manages to flash the BIOS, the user is screwed, anyways.

Re:I'm all for it

[Your.Master]

Intel knows precisely how fast their CPU is, especially when still locked. They can make their code almost arbitrarily long. They could easily stamp a code that would take more than the lifetime of the universe to crack with that CPU, or even crank it down to something like 20 years mean so they won't give a shit anymore that it's cracked. Or they could even just make the machine such that it ceases to function after 100000 failed tries: some high enough number that it basically guarantees you were trying to brute-force it.

Re:I'm all for it

[Your.Master]

So Intel loses $50 on that guy. Big deal. Unless that guy is going to do that for a statistically significant portion of Intel's customer base, for less than $50, then it's basically irrelevant. Intel probably loses more money to errant sneezes.

Re:I'm all for it

[mysidia]

Assuming the manufacturing equipment pulls an unlock code from a true entropy source, and burns some combination of 128 solder bridges to encode a MD5 message digest of a 20-character alphanumeric representation of the code.

(1) There is no 'algorithm' to discover

and

(2) There is no way to simply read the code off the top of the CPU.

Re:I'm all for it

[Your.Master]

There's no reason they have to be created by an algorithm. They could just take line-in noise from a lava lamp or gamma ray readings from a nearby coal mine.

Anyway, reverse-engineering an algorithm from knowledge of a limited number of unlock codes in an unknown sequence and with holes of unknown size in the sequence, is pretty ridiculously difficult in the first place.

Re:I'm all for it

[DadLeopard]

Only Copyright lasts a whole lot longer than a patent! Used to be 17 years, now twenty years! Copyrights last a lot longer than that!

Re:I'm all for it

[plague911]

Intel / AMD could sell them at their stock clocks. Overclocking already voids their warantee so presumably this should not effect Intel or AMD at all.

So the question becomes "why do they do it all if they are no longer responsible and only customers who are very techologicaly oriented will do it any way?"

The answer is they want to pry more money out of our cold dead hands.. There is no reason for this.

What you are referring to latter is something called binning. Yes there is reason to sell a quad core with one broken core as a three core.. This is not the same issue with multipliers.

Re:I'm all for it

[petermgreen]

How does the cpu know you provided the right unlock code if it cannot be calculated from the serial number?

There is just one way, store it too on the cpu.
Just put it in a special section of memory that can't be read out through software and lock out further attempts at entering a code for an hour or so after a small number of wrong codes.

With say 4 billion possible codes (and they could make the code much longer than that) and a chip that will only let you try 3 an hour you are going to be there for a very long time trying to guess the code.

Yes someone probablly could decap the chip, remove any protective layers and read out the code with a microscope but by the time they have done that it would probablly have been cheaper to just buy the damn code.

Re:I'm all for it

[Keen Anthony]

I remember that. I think SQL Anywhere was the same way, as was Novell NetWare. But it seemed to me that people were more forgiving when it's software due to licensing. "Licensing" is like some magical word that one can use to force a consumer into a corner. With hardware, I've only remembered hearing complaints when consumers knew that they were getting hobbled hardware. As one person pointed out to me earlier, how many crippled chips have I run that I've been aware of -- who's knows.

Re:I'm all for it

[Keen Anthony]

Less understanding, and more conceding defeat when it comes to proprietary video codecs.

Re:I'm all for it

[Keen Anthony]

Unless you own any patents on which you're earning revenue from either producing something or licensing out the technology you created, then you wouldn't see any evidence that patents do good. Besides, patents only serve the purpose of rewarding innovators with exclusivity so they can maximize profits and get some early benefit before a copycat undercuts the innovator on price. Patents were never meant to benefit anyone but the patent owner.

Re:I'm all for it

[Keen Anthony]

As a happy Apple owner, I do have to concede that I by not using iPhoto, and the rest of the iLife suite, I am not realizing the full worth of what I paid for when I bought my Mac, but I wouldn't buy $700 computer. I just like all my components to be discrete and not on the motherboard or a shared resource, unless I'm dealing with a laptop. I'm probably being unfair. I just remember back when I computers came with real Hayes modems. Hehe.

You probably wouldn't pay for a new PC that costs the same as a Mac, would you?

Re:I'm all for it

[Keen Anthony]

You calling me a fanboy? Of what exactly???

I didn't pay for Apple's MPEG2 decoder. But I can understand that the MPEG people want their cut for licensing out their technology, and by understand, I am saying, I concede that the issue has been argued to death and the result is people are expected to pay for that MPEG authorized decoder.

Of course Apple can fuck around with its customers. Microsoft can too, and if Red Hat could, it would find a way too.

Can't Windows play DVDs now too? Can Windows play MPEG2 straight out of Media Player without downloading anything? Even the less "evil" computer companies out there like Amiga and Atari have wronged me from time to time. I rely on open source software for my MPEG2 playback, and I tell other Mac users to do the same. But I'm smart enough to know my alternatives. I drive a Lexus. Lexus expects me to keep paying to update that damn nav computer, otherwise it's useless. I'm using my Android instead. There are ways around this stuff. Consumers aren't prisoner to things like licenses as they were before, but if you still want that official MPEG2 encoder/decoder software, you're going to pay for it.

Re:I'm all for it

[srmalloy]

Whats worse is that if this works they'll eventually change a $50 one off fee into $x per month to keep
it at the higher speed. Welcome to the era of rented hardware DLC.

I don't think so... If the 'upgrade' could be turned off, imagine how many gigatonnes of sewage Intel's going to take in the face when (not if) the first piece of malware shows up that derates your CPU again, when by making it a one-way transition they could have prevented it.

Re:I'm all for it

[Altrag]

You obviously failed the class on modern business practices. If they decided that something is worth $50, no matter how arbitrary or inane, then its obviously "costing" them $50 if they turn around and give it to you for free. And you should tell your customers that they're actually SAVING $50 if they decide not to purchase the extra power, so that they feel good about you and their purchase.

People who pirate the program will be dealt with in modern business practices II, taught by representatives from the RIAA. There you will learn how each pirated copy of the unlock software is actually costing Intel $50 _per transistor_!

Re:I'm all for it

[Keen Anthony]

Okay, I can accept that. So, out of curiosity, is AMD doing this? Has AMD done this anytime in the past? Do you think AMD will benefit from this move? I know some people on other forums have said their next CPU will be AMD. But I figure, if you're shopping for an Intel chip, it's not like you have the option to use an AMD anyway right? I don't know that AMDs can be dropped into MBs that call for Intel chips.

Re:I'm all for it

[vadim_t]

Unless you own any patents on which you're earning revenue from either producing something or licensing out the technology you created, then you wouldn't see any evidence that patents do good.

Paraphrased: Unless you own any slaves on which you're earning revenue, you wouldn't see any evidence that slavery does good.

Just because somebody benefits from something doesn't automatically make that a good thing.

Patents were never meant to benefit anyone but the patent owner.

Then I disagree with them on that basis. Society shouldn't grant a monopoly if it's not going to get anything in exchange.

You'll have to be more persuasive. Picture this: I will vote against patents if such an opportunity comes up. You have to convince me that it's actually in my interest to have them. That you happen to benefit from it is good for you, but does absolutely nothing to convince me. So you'll have to explain how you having a patent is going to benefit me. And really, you having a monopoly isn't in my interest. What's in my interest is to have the maximum amount of competition. So try again?

Re:I'm all for it

[Keen Anthony]

You're not understanding me. I'm not saying that I think software patents are good. I'm responding to your comment that you have seen little evidence that patents do any good, by saying that patents aren't meant to do any good. They are meant purely to benefit an innovator (presumptively) by giving them a chance to maximize returns early on for their innovation before competitors can copy the innovation and offer it cheaper. It's decidedly anti-consumer. I wasn't trying to convince you of anything except that patents do not address a public good. Sure, someone might say patents inspire people to be innovative since it gives them an economic incentive, but we all know that's not true.

Again, having a temporary monopoly is not supposed to be in your interest, it's supposed to me in MY interest if I'm a patent holder.

Re:I'm all for it

[TENTH SHOW JAM]

Nope, because you are given both the key and the lock. The lock being an algorithm that turns a magical code from Intel into a simple boolean value. (upgrade/not upgrade). Some clever clogs gets some chips and examines the ins and outs of the system. Manages to reverse engineer the lock. Then provides another website/executable etc... "Smear in your product ID here". "Your unlock code is ......" It would be hacked as quickly as games with a key code. on the back of the manual.

If they want to make the system secure there has to be a phone home component with a rolling code. And we all love a CPU that needs to be connected to the Internet to perform well.

Re:I'm all for it

[mysidia]

I don't doubt someone could physically defeat the control on their own CPU, but it's possible to make physical access required. I don't think you understand what i'm suggesting they could do.

Each CPU would have its own unique code passphrase chosen at time of manufacture from a true Entropy source; meaning the bits are random, and the only rule relating a code with the CPU serial number is "lookup the entry in the database".

A cryptographically strong HMAC) would be computed from the code and the features the CPU is being shipped with, and the corresponding solder bridges would be burned by the laser, to physically incorporate the CPU's individual code.

To lock a core, a fusible link is inserted between each core's Vcc+, and ground, making the core unable to be powered on unless the physical link is broken by transmitting a high voltage.

The fusible link is controlled by a relay attached to a dedicated security circuit, that will only activate the link, once a code has been input.

The security circuit will only allow 10 attempts, with a maximum of 1 attempt every 10 seconds. After 10 attempts, it will throw a relay that disconnects itself, and charge a capacitor to maintain that state, until the CPU is power loses power for at least 30 seconds, then 10 more attempts can be made.

Re:I'm all for it

[vadim_t]

In that case, I take what you said as an argument against patents. It's clearly a system to benefit the very few (since most people don't hold any patents), and what's more, the rest of the people lose from it. You yourself said it is anti-consumer, which makes it anti-99% of the population.

With such a state of affairs, things would be better if we eliminated them as soon as possible.

Re:I'm all for it

[MBGMorden]

You know that you probably bought a handicapped chip already, right?

Of course, but that's not the point. As the OP said, it's about PERCEPTION. Most people don't really think about (or even know about for many consumers) the fact that their CPU may be a downscaled version of another one. If you offer to "unlock" those disabled features for a price though, then it's a firm slap in the face reminding you that this chip most assuredly is just arbitrarily limited.

Consider it like a John visiting a prostitute. Most of them know that the prostitute is only sleeping with him for money, and they're ok with that in the back of their mind, but they certainly don't want to be reminded of that fact the whole time.

Or consider when World of Warcraft came out: in an effort to reduce excessive playtime, developers built in a "fatigued" mechanic. As long as you rested at an inn your character would return to normal, but eventually if you gained too much XP without letting your fatigue wear off you started gaining XP at half the normal rate. There was public outcry. So, rather than changing the mechanic AT ALL, Blizzard changed the wording so that instead of the extended play time XP being "halved", IT was renamed to the normal rate and the other state was renamed from normal to "rested" and the claim was that you get DOUBLE the XP in that state. Numerically identical, but the community accepted this version pretty well because the prospect of a reward that builds over time was more palatable than a penalty that decreases with time, even if they might work out the same.

Presentation of the situation is as important as the details of the situation.

Re:I'm all for it

[slashtivus]

I was really thinking of a 'Phone Home' system. I've been stuck with the task of licensing software before, and I had to be honest with management (showing them how many DRM systems have been broken before) that all licensing systems rely on security through obscurity in the end... since you have to give them the software/content to begin with. That was rather uncomfortable for me, but it was the honest truth. I didn't really have a comment regarding chip producers, so maybe I'm off-topic on that idea.

Re:I'm all for it

[muridae]

So, now there are two chips in the economy bin, a 2x2.6Ghz and a 2x2.6GhzXT. One costs $50 and failed a test, the other has the same features, costs the same price, and can be upgraded later. The OEMs buy the latter and put them in beige boxes with certain serial numbers, and the rest go to everyone else.

Re:I'm all for it

[muridae]

The same thing that stops them from writing a BIOS virus that undervolts the CPU, clocks down the FSB, and turns on PIO for your old drives. The same thing that keeps them from writing a virus that edits the microcode on the CPU, allowing a rather complete rootkit. The fact that each batch is going to be different from the others, and that the BIOS/CMOS on the motherboard will have to support these features.

Board replacement is an interesting question. I have no clue how that would work. Since it is a Gateway, I really doubt the market is the type of person to perform a board replacement. I suspect that the 'licensed repair facilities' would have access to that, and could perform the update on any warrantied repairs.

Re:I'm all for it

[muridae]

These are the same consumers who gladly pay for extended warranties, who do not know the difference between a duel core cpu and a two socket motherboard, and who ask "My nephew told me I need more of those giga bites to make my computer better, so do I need this drive with 500 of them, or this stick with 4? Wouldn't 500 be better?"

Phrase it this way instead "Sale, 50 dollars off!!!"

Re:I'm all for it

[toddestan]

Presumably, it wouldn't be permanent as they would want make it so the faster speed could be re-enabled again if you wanted to start giving them money again.

Besides, a virus could, in theory, derate your CPU today by mucking around in your BIOS.

Re:I'm all for it

[toddestan]

Exactly, just like how the Windows users had to pay Microsoft for SP2 on Windows XP...

Hey, wait a minute...

Re:I'm all for it

[Travoltus]

Yeah, it's not like those secret unlock codes haven't been broken before. (HD DVD/Blu-Ray, anyone?)

Re:I'm all for it

[md65536]

I definitely like the idea of pirated hardware.

Maybe in the future you'll be able to download a Ferrari off TPB, and burn it onto a blank Hyundai.

Re:I'm all for it

[Joce640k]

I understand it as well as you do. You're not thinking hard enough. If 'n' isn't different for every chip then it's broken. ...and if 'n' can be different for every chip then symmetric crypto will work just fine (and be much simpler to implement).

Re:I'm all for it

[dargaud]

because it makes economic sense

Nope, because if the extra memory (or cores or speed) is already inside your computer, then it means you have ALREADY paid for it. Paying extra later to unlock it is just a ripoff.

Re:I'm all for it

[Jesus_666]

But what would be the chances of that happening? I think that storing millions of unlock codes for years when only a fraction of them will ever be used sounds like it's unneccessarily expensive. And if Intel can't recreate an unlock key I'd expect them to get sued when someone tries to unlock their ten year old CPU only to get told that Intel doesn't have the unlock code anymore.

However, if Intel can create new unlock codes that means the unlock codes are most likely the output of an algorithm which takes the serial number (plus pehaps other distinguishing features) as its input. Which means that reverse-engineering the algorithm should be possible.

Plus, what's to stop you from brute-forcing the unlock procedure? Will the CPU permanently disable itself after a few tries? Is the key space so large the a modern CPU can't be expected to get through it withing reasonable time? (This seems fairly unsafe as a partially reverse-engineered algorithm could drastically reduce it.) Will Intel require you to boot their special live-CD with encryption and trusted computing up the wazoo so the no one can look at their tool, which uses a proprietary but secure cypher to communicate with Intel?


I expect either a crack or a lawsuit.

Re:I'm all for it

[mrchaotica]

...patents aren't meant to do any good.

Like Hell they're not! The express purpose of patents are "to promote the progress of science and the useful arts." Not to "benefit an innovator" -- that's merely a side effect; a means to an end.

...patents do not address a public good. Sure, someone might say patents inspire people to be innovative since it gives them an economic incentive, but we all know that's not true.

Then they have no valid reason for continuing to exist, and must be abolished. Period.

Re:I'm all for it

[mrchaotica]

But I can understand that the MPEG people want their cut for licensing out their technology

I think everybody understands that the MPG people want their cut; what some of us don't understand is why the law forces us to give it to them when it doesn't do society any good overall.

Re:I'm all for it

[LordKronos]

Wow, you are so smart. However, you might want to stop by random.org or do a google search for "truly random number generator". More than a decade ago, Intel even built a chip to do this and included it in many systems.

Re:I'm all for it

[alexo]

The random number is an unlock key. It cannot be read out of the processor by any means. It can, however, be compared to a user-supplied unlock key written to some special purpose register by software. Once again, the method for writing to this register need not be hidden. The unlock mechanism is simple: if the user-supplied key matches the factory-supplied key, the processor unlocks the upgrade functionality.

Since the factory-supplied key can never be read out, all that is necessary is to make sure they're secret and unguessable. Secret is a matter of keeping the data secure inside Intel as you said; the only way they let unlock keys out is if users buy them, and by using unique pairs of serial numbers and unlock keys Intel never gives out data that someone can adapt to unlock all processors. Unguessable is easy: just restrict the mechanism so it only accepts one user supplied key per second, or something like that, and make the random key string long enough that it would be impractical to brute force it.

Even easier than a per-second limit:

Have two registers:
R0 - Internal use, comes up as 0, set to 1 when writing to R1.
R1 - Write only, for entering the secret number.
The extra processor capabilities are unlocked when the user write to R1 and R0=0.

Now you have to power-cycle the processor between tries.

Re:I'm all for it

[TheFakeMcCoy]

It's like buying a game, and then paying to unlock the extra conten already included on the disk. That went over really well.

Re:I'm all for it

[BlackBloq]

If it costs me as a chip maker more to manufacture two different chips then why not make one chip and have the other "built in"? The problem is they have incurred the cost of EVERY chip, so that leaves one to ask if we are getting ripped off!". Following this logic, if a game company sold you a disk with 5 games on it and you would pay to unlock one at a time are they ripping you off?? Seems cool that all the product is there at your fingertips. Software doens't really incur costs pressing disks so what about trucks? If your truck came with 4 wheel drive built in and you could "upgrade" and be able to turn your 4 wheel drive button on; that would seem strange. The bottom line is that THE COMPANY has to be responsible for the cost of the base unit. If the cost of the base unit is still low and even competitive then there isn't a problem. Like if my two wheel drive trucks (with disabled 4 wheel) cost less then last years two wheel drive model without unlock, then they are adding value by dropping costs. I don't see how they could but if the product is a good price, then they are adding value.

Re:I'm all for it

[mysidia]

Plus, what's to stop you from brute-forcing the unlock procedure? Will the CPU permanently disable itself after a few tries?

No. The dedicated unlock circuit would have a 20 second countdown timer. After powering on the CPU, there would be some brief waiting period you have to endure before the CPU is ready to accept an unlock command. After a failure, you would have to wait 30 seconds before another attempt. Also, after 10 attempts, the security circuit would disable itself until the CPU is powered off for at least 30 seconds.

So the result is you can try approximately 2 unlock attempts per minute on average. If there are 128 bits, then it will take you approximately 10^33 years to unlock by brute force.

But what would be the chances of that happening? I think that storing millions of unlock codes for years when only a fraction of them will ever be used sounds like it's unneccessarily expensive.

If each unlock code requires storing 128 bytes, then 12 gigabytes is enough to store 100 million unlock codes, and 240 gigabytes is enough to store 2 billion unlock codes.

Enterprise storage is about $0.30 a gigabyte per year, and getting cheaper every year, so you are talking gasp $72 a year to store 2 billion unlock codes. There are costs involved, but storage is not a significant one.

Re:I'm all for it

[DarkEmpath]

I can understand having to pay extra for MPEG-2 support in a piece of software or hardware, as there are licensing issues with a third party, but you're Intel. This is your chip.

Not quite. Since AMD's 64 bit extensions to x86 became the de facto standard (over Intel's IA64 Itanium), Intel has had to licence x86-64 from AMD and is required to pay royalties. Yes, AMD has to pay royalties to Intel for x86 (and more) and they probably do some royalty-swap deals, but it's not quite as clear cut as "this is your chip".

But yeah, I agree this is a money-hungry cash grab and a dick move by Intel.

Re:I'm all for it

[WNight]

Software is licensed, not sold

Strange, I just asked at the store and they said it was sold.

There are many people who want you to think a license is required, but I don't see any proof of it.

Yeah, That'll Last

[MoonBuggy]

Crack coming out in 3...2...1...

Re:Yeah, That'll Last

[BlkRb0t]

Finally, we can pirate hardware.

Re:Yeah, That'll Last

[Anonymous Coward]

I love how people who don't know much about encryption pick up the phrase "one-time pad" and assume it refers to some magical cryptographic panacea.

Hint 1: "one-time pad" means a little bit of paper with a key printed on it. You probably mean "one-time key", which is a rather more general concept that is easier to implement in silicon.

Hint 2: there's a reason why vast resources have been poured into inventing encryption systems that can securely use a single key repeatedly, and it's not because one-time key systems are economically viable.

Re:Yeah, That'll Last

[Sir_Lewk]

Probably not if each slashdot poser knows what a one time pad is.

Re:Yeah, That'll Last

[WoLpH]

It won't be the first time that we are able to pirate hardware. Just like the expensive processors that get an unlocked multiplier... just another way to get more money without actually doing anything. But they too have been pirated in the past. Remember that people were fooling around with silver to unlock there processors?

Re:Yeah, That'll Last

[_Sprocket_]

Hint 1: "one-time pad" means a little bit of paper with a key printed on it. You probably mean "one-time key", which is a rather more general concept that is easier to implement in silicon.

Isn't a one-time pad a stack of said little pieces of paper - usually bound in some way?

Re:Yeah, That'll Last

[v1]

it would be completely impractical to try to hard code a different key onto each chip during manufacturing.

if they wanted to do this they'd have to have a fuse array in it like the satellite dish receivers use to decrypt their signals, set at the factory after production.

Possible... yes. Expensive... somewhat. Added complexity... absolutely. Likely.... not really.

It'll just come down to knowing a supasecret opcode sequence to execute that causes the chip to shift into higher gear until it gets reset. (powered off) So they'll send you a little bit of software that installs a startup item that loads and boosts the proc at some point in the boot process. It'll be just as (in)effective as any other software antipiracy method. I'm sure they'll try all sorts of tricks like keys customized to your windows key but in the end it all comes down to someone prying open the program and giving it some brain damage so it behaves more generously. Tho if it really is a simple opcode sequence, the pirates will probably chuck the program after they've decompiled it, and write a one-liner that unlocks it for you anytime you want.

Re:Yeah, That'll Last

[kestasjk]

Chips aren't individual, the only way they can be made cost effective is by being mass, mass, mass produced.. They can have a unique number flashed to them though. (Small nitpick but it's important not to get the wrong idea)

Re:Yeah, That'll Last

[Cylix]

"it all comes down to someone prying open the program and giving it some brain damage"

I immediately thought of Tron competing at the Special Olympics.

Re:Yeah, That'll Last

[whoever57]

it would be completely impractical to try to hard code a different key onto each chip during manufacturing.

Didn't Intel processors have a unique ID at one time?

Re:Yeah, That'll Last

[Ethanol-fueled]

No, a one-time pad is a butterfly-shaped absorbant sheet of cotton and/or synthetic fiber. It goes where no Slashdotter has gone before.

Re:Yeah, That'll Last

[_Sprocket_]

Is there nothing that can't have cryptography applied to it?

Re:Yeah, That'll Last

[troll8901]

You'd need to perform a dictionary and encyclopaediac attack on the description.

Butterfly = see Chaos Theory
Cotton = see Vegetable Lamb of Tartary
Fiber = see ITU-T G.652 "Characteristics of a single-mode optical fibre cable"

Re:Yeah, That'll Last

[KiloByte]

And they have it again. The first time, it caused a huge outcry, but now, as a part of the TPM chip, somehow no one seems to be upset, even though it's a much worse invasion of privacy and customer rights than just a mere serial number.

Re:Yeah, That'll Last

[Joce640k]

They're already building in a unique serial number. Why would adding random encryption key suddenly make things more difficult/expensive?

Re:Yeah, That'll Last

[Joce640k]

Why do they need a "one time pad" which is bigger than the plaintext? (ie. serial number)

Re:Yeah, That'll Last

[Joce640k]

Intel had chips with unique serial numbers back in the Pentium III days. I'm sure they still remember how to do it.

Re:Yeah, That'll Last

[ooshna]

I don't remember silver but I remember using the graphite in pencil lead to connect some leads or whatever that they would cut with a laser to lock the cpu

Re:Yeah, That'll Last

[drej]

"You wouldn't download a CPU"

Re:Yeah, That'll Last

[v1]

If they're already doing the serial number, then a unique random code wouldn't be much of a bother.

There'd have to be more to it than just that though. Usually in those cases (such as with the sat receivers) they have to put in special circuitry in the chip.

lets say they didn't do that and just put another random number in there, lets not even say they hash the sn, we'll give them benefit of the doubt that they use a real random number and just keep a table at the plant for sn and random unlock code.

Then when you "purchase your upgrade", they mail you a link to a downloadable program, and an unlock key. You download the program, punch in the key, and the program confirms the key matches your proc and its random number, and the program then tells the processor to bump.

See the problem? The program is the gatekeeper, not the chip. Ten minutes with a disassembler and you figure out how the program is talking to the chip to unlock it, and it's cracked.

First, a clarification. It's totally impractical to physically differentiate chips during production. They use lithography, and you can't just have a machine at the end of the line that goes in with tweezers making jumper changes. The chips instead have a grid of blowable fuses. At the end of the line, the machine with the SN list drops down the pins onto the chip and runs the initial tests on the chip, and then blows the fuses corresponding to the serial number of the chip. This is irreversible. This is how they all do it. And this is how they would set an unlock code too.

Getting back to the problem. If all you are doing is setting another number on the chip, you aren't protecting the chip, you're only using it as a way to store another number. The only reason you need to know the number to unlock the chip is because the unlock program insists on your license key matching the code it reads off the chip. The program is a very weak protection, easily cracked or decompiled.

The "correct" way to do it is different:

The unlock code is still blown fuses like the serial number, but with a difference. It's WRITE ONLY. There's no instructions you can send to the processor to ask it what its unlock code is. Instead, there's a new procedure added to the chip that allows the license program to SEND the unlock code to the chip. The chip then, internally, compares the provided number with the burned one, and if they match, it unlocks. If not, it doesn't. And done properly, it won't allow another attempt for some time, possibly until it's been power cycled. This prevents brute-forcing it.

(some of the more viscous methods used in the cable industry are to only allow a fixed number of attempts, and after so many fails, the chip bricks itself or becomes permanently locked etc, via burning another of its internal fuses)

This takes protection out of the hands of the program, and puts it in the processor, safely out of reach of most people. But it does require some additional changes to the chip. In retrospect, considering all that's IN the chip to begin with, I suppose it's not that big of a deal to add, but I just wanted you to understand it's not just a matter of writing another number to the fuses.

If they were stupid, the unlock code would be a hash of the proc sn. In that case, it's quite possible that the hash algorythm could be discovered, in which case anyone could write an unlocker or a keygen for the downloadable app.

But if they were using a hash, then it wouldn't be necessary to burn the unlock code into the chip, since the proc could run the hash on its sn itself and compare with the provided hash. But as I said, that would be the stupid way to do it, and I doubt it would save them much money in the long run even if it never was broken.

Re:Yeah, That'll Last

[zill]

You wouldn't steal a car. You wouldn't steal an Intel CPU. You wouldn't steal the unlock code.

Re:Yeah, That'll Last

[kestasjk]

They still have them today, but they do it by flashing them not because every chip is unique.

Re:Yeah, That'll Last

[toddestan]

You've been able to do that a long time. The only thing that changes is now you don't need a boat.

Re:Yeah, That'll Last

[KiloByte]

1. MAC addresses don't traverse the network beyond your first hop, and are trivially overridable.
2. Most current machines have TPM built in. It's not just a "niche value-subtracted(correction mine) component", it's actually there in, in all likelihood, your machine as well.
3. The problem here is that TPM provides that unique serial number, and more -- so it deserves at least as much attention as the Pentium3 one did.

Step right up, place your bets

[BBTaeKwonDo]

on the date of the first crack to unlock it without paying the fee.

Re:Step right up, place your bets

[blai]

Bollocks! Intel has Intel Genuine Advantage installed on the CPU!!

Re:Step right up, place your bets

[neo0983]

I am feeling generous today and will give Intel the benefit of the doubt that they are not completely incompetent. 1.5 days before the chips hit the general public.

Re:Step right up, place your bets

[teh moges]

Who cares if it can be pirated? It'll cost Intel very little to do this and the aim isn't to get lots of geeks paying $50, its to get average Joe at home to do it. They will still make money off this idea (so long as it isn't a huge PR disaster - selling crippled hardware...) with or without a patch.

Re:Step right up, place your bets

[BrokenHalo]

In any case, I don't really see what Intel has to gain from this. Why not simply sell their CPUs capable of their maximum speed and charge $50 extra in the first place? This would save them the trouble and expense of setting up and maintaining a lot of server architecture, and maintain customer goodwill into the bargain. A customer who has been blackmailed once is going to think twice before coming back.

Re:Step right up, place your bets

[c6gunner]

In any case, I don't really see what Intel has to gain from this. Why not simply sell their CPUs capable of their maximum speed and charge $50 extra in the first place?

Because they'd lose customers to AMD?

Re:Step right up, place your bets

[TheRaven64]

Why not simply sell their CPUs capable of their maximum speed and charge $50 extra in the first place?

You can already buy the CPUs for the faster speed costing $50 more. You can already buy them at the slower speed for $50 less. This means that you can buy the cheaper one now, then upgrade it to the faster one for $50 later, when you need a faster CPU. Most people don't upgrade their CPUs, they buy a completely new computer, which costs a lot more than $50, so it happens less often. By making the upgrade so cheap, they get people who would otherwise have bought the cheap chip to buy the more expensive one and defer part of the payment for a year or so.

I hope this doesn't fly ...

[ScrewMaster]

but if it does, it's a big opportunity for AMD. Of course, odds are it'll get cracked at some point and we'll be able to grab an "Intel Upgrade Service Crack" torrent.

Presumably Intel will be using the CPU serial number to keep track of legitimate users and so forth. But here it comes: have we bought a central processing unit which has now become our property because we paid for it, or are we simply buying a "license" to use Intel's "intellectual property"? If I go out and buy a penknife, I don't expect to have to pay more money if I want to be able to use the built-in compass. Will the BSA (or some similar organization) come down on companies that unlock their processors without paying Intel's upgrade fee? This has the potential to get ugly.

Re:I hope this doesn't fly ...

[Anonymous Coward]

That sounds an awful like circumventing a digital rights management technology. Remember, you didn't buy a CPU, you bought a /license/ to use the CPU.

Re:I hope this doesn't fly ...

[Greyfox]

IBM's been doing that sort of thing for years. They ship you a mainframe with more processors than you ordered or a disk array with more disk than you ordered, and you can pay them to turn it on. Some companies only turn on their extra processors for a short time each year (Like end-of-year transaction processing) and if you decide you need some more space in your disk array, it's much more convenient than having to have more disks installed or buy a new disk array.

Re:I hope this doesn't fly ...

[Anonymous Coward]

Except, I'm not sure how that really applies to hardware. Can you license hardware? Remember, modding consoles is illegal because you start fiddling with licensed software as well, not just the hardware you own.

Re:I hope this doesn't fly ...

[Shyfer]

Lately tech-related news is making me afraid of our future, seriously. Count how many good and how many bad news were on slashdot this year and you will see that things are not good.

Re:I hope this doesn't fly ...

[Kitkoan]

Since its a software unlock, it can come down to a DMCA violation, or it's a possible software style "codec" to designed to "optimize" your processor (which would make it software piracy to do it without paying for the unlocked). As a hardware only concept it might be a harder issue (but could work along the lines of modchipping laws possibly), but since it's software then the BSA could decide to have a go at you.

Re:I hope this doesn't fly ...

[Interoperable]

I would imagine that it would be similar legally to applying mods or cracks to game consoles or iPhones. Breaking a digital lock is, in itself, illegal in some countries and that would certainly qualify. Beyond that, I've never really understood the argument that you can't do what you want with your phone/console/chip.

My question is how is the patch applied? Is it a firmware update on the processor itself? (Do processors have firmware to update?) Bios update? If it's in the operating system, what does the processor look like to a Linux kernel and how would one unlock it in that case? No doubt someone here has a better idea of how it will work than I do.

Re:I hope this doesn't fly ...

[postbigbang]

Right. Intel. DRM. BluRay.....

Snort.

Re:I hope this doesn't fly ...

[Nyeerrmm]

I'm not a lawyer, or a computer/electrical engineer for that matter. However, I'm pretty sure, unless it was a direct copy of the of the intel unlock software, then the most you could do legally with regard to the software is some patent nonsense. Do a cleanroom operation to understand the process and host it from a location without software patents and its good.

As far as making modifications like that to your own hardware, the recent ruling on cell phone unlocking probably makes it pretty safe legally, although it's sure to void your warranty.

But then again, I don't really know the legal basis behind mod chip laws, and I'm no lawyer anyway. But I think a BSA issue is unlikely to be the avenue intel might try to stop it,

Re:I hope this doesn't fly ...

[rumith]

Oh. I see. They're going to lock down fully capable CPUs. Now that's a shame.

Re:I hope this doesn't fly ...

[kestasjk]

IBM's been doing that sort of thing for years. They ship you a mainframe with more processors than you ordered or a disk array with more disk than you ordered, and you can pay them to turn it on. Some companies only turn on their extra processors for a short time each year (Like end-of-year transaction processing) and if you decide you need some more space in your disk array, it's much more convenient than having to have more disks installed or buy a new disk array.

That almost sounds like .. a convenient, useful service.. But surely it can't both be good value for money, a good service, and use DRM can it?

Re:I hope this doesn't fly ...

[gnarlin]

And how many years did it take to crack that? Won't they upgrade the drm every year or so?

Re:I hope this doesn't fly ...

[Haeleth]

If I go out and buy a penknife, I don't expect to have to pay more money if I want to be able to use the built-in compass.

Not a great analogy. Can you try again only with more cars?

This isn't a case of you buying a Core i7 and Intel saying "by the way, we only gave you a Core i5, but you can have the full i7 you paid for if you give us another $50".

This is a case of you buying a Core i5 and Intel saying "here is exactly what you paid for, and by the way, if you ever decide you should have bought a Core i7 instead, we can magically teleport one into your computer for just $50".

If you want the pocket knife with a built-in compass, pay for the one that has a compass in it. If you deliberately buy a knife that says "KNIFE WITHOUT COMPASS (compass is available at extra cost)", you have no reason to complain when it turns out you have to pay extra to get a compass!

There's no bait-and-switch here. People are getting exactly what is advertised. Where's the problem?

Re:I hope this doesn't fly ...

[davester666]

Yes, you might be "buying" the hardware, but you are also licensing software within the CPU [as I believe both AMD and Intel cpu's use microcode within it to process x86 instructions].

Re:I hope this doesn't fly ...

[mysidia]

CPUs have built-in software, they are just commonly called 'Microcode' instead of software.

Hyperthreading is a software feature that involves using hardware to implement it.

So is "VT" / Virtualization Technlogy

There are even ATOM CPUs where the hardware is 64-bit capable, but Intel ships without the 64-bit capability enabled in the software.

So, yes, it's a technlogical protection that defends intel's exclusive right to license and distribute the Hyperthreading software.

Until you have been provided the code, Intel has not licensed the Hyperthreading software to you. It's just like an expired trial version of any sort of shareware you might have on your computer.

I wonder if Intel will offer a 30 day trial of Hyperthreading and Cache expansion <GT>

Re:I hope this doesn't fly ...

[Bob9113]

IBM's been doing that sort of thing for years. They ship you a mainframe with more processors than you ordered or a disk array with more disk than you ordered, and you can pay them to turn it on.

Is this leased or purchased hardware?

Leased hardware comes with a contract. Purchased hardware comes with a first-sale doctrine.

The purpose of the first sale doctrine is to maintain an information balance between purchaser and seller. "Purchase" is a standardized term so that we don't have to take our lawyers with us when we go shopping. If we had to take our lawyers with us, it would create friction in the economy, reducing the velocity of money.

Now, the idea of keeping the first-sale doctrine is, of course, predicated on the notion that an informed consumer is a good thing. That hypothesis is only true for business owners if they believe in the free market and maximizing GDP growth. While that is almost certainly their official line, actions speak louder than words. Offered the opportunity to engage in a market practice which reduces the probability of GDP growth but increases quarterly earnings, take a wild guess which one wins.

And, to bring the comment full circle, this brings us back to regulations like the first-sale doctrine. As Adam Smith himself pointed out, "People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public." The tool to prevent such behavior is simple and limited regulation like the first-sale doctrine.

If a company offers to lease me something, requiring me physically sign a contract that I can read in advance, no problem. If another company sells me something, then tries to tell me how I am allowed to use it afterwards, perhaps corrupting our government to achieve that end, that company is an enemy of the free market, and a progenitor of reduced GDP growth.

Re:I hope this doesn't fly ...

[ScrewMaster]

Except, I'm not sure how that really applies to hardware. Can you license hardware? Remember, modding consoles is illegal because you start fiddling with licensed software as well, not just the hardware you own.

Actually, I think it's more basic than that. You are using the bits of information stored in a particular media (CD, DVD, Blu-Ray, CPU microcode store, whatever) in a manner not authorized by the vendor.

Re:I hope this doesn't fly ...

[bieber]

...except that you did buy an i7, it's just that they didn't tell you about it. Just because a feature wasn't advertised doesn't mean I didn't pay for it when I bought the hardware, or that the price I paid didn't include the cost of manufacturing that extra feature. You shouldn't be going around critiquing other peoples' analogies if you're going to liken activating hardware that you've already paid for to magically teleporting new hardware into your computer...

Re:I hope this doesn't fly ...

[hairyfeet]

Yep sounds like they'd need DMCA, and if I was the CEO of AMD I'd be fricking dancing in the streets at this news. hell the commercial writes itself /shows sleazy used car salesman type selling PCs/ "Intel sells you hardware you can't use until you pay ANOTHER fee on TOP of what you paid for your computer. We here at AMD think you should get what you paid for, so we don't deal with such shady tactics" /sleazy salesman points at Intel box that side falls off and tries to cover it up/

hell between this and Intel getting investigated for shady dealings after having to shell out 1.25b to AMD to get them to shut up, it's like the gift that keeps on giving!

Re:I hope this doesn't fly ...

[clarkkent09]

I agree. However, the interesting question is what happens when some customers buy a pocket knife "without a compass" and then discover that the pocket knife does actually have a hidden compass and find a way to enable it. Will they get sued? In case of a pocket knife, no. In case of a CPU, I don't know but it wouldn't surprise me.

Re:I hope this doesn't fly ...

[GrumblyStuff]

I imagine the problem is akin to so called downloadable content for console games which is already on the disc but requires you to pay for and download some magic patch to enable it.

With the knife analogy, it already has the compass! The company already paid for it and only gets a lower profit margin if no one opts to "enable" it. With that in mind, they could charge a trifle more, have that extra bullet pointed feature on the list on the packaging, and come out ahead easily. Sell the knife sans compass for the same (compass enabling knife price point) price and there's even more profit!

The problem is that it doesn't make sense from the common sense point of view.

Re:I hope this doesn't fly ...

[Kitkoan]

The problem is you need to pay more for the equivalent of "flipping a switch" to get the most from your product. And they've made sure to keep it flippable on a software only level so you can't flip it yourself and it can literally/possible legally only be flipped by a "Intel certified" process.

Re:I hope this doesn't fly ...

[ScrewMaster]

IBM's been doing that sort of thing for years. They ship you a mainframe with more processors than you ordered or a disk array with more disk than you ordered, and you can pay them to turn it on. Some companies only turn on their extra processors for a short time each year (Like end-of-year transaction processing) and if you decide you need some more space in your disk array, it's much more convenient than having to have more disks installed or buy a new disk array.

True. My father used to work on a Hewlett-Packard mainframe back in the seventies, and he ordered some extra hard disk space. The HP tech came out, opened the casing of each drive (big freestanding units), reached in the back and flipped a DIP switch. Voila!, extra space. He even showed Dad how to turn on the entire drive if he wanted ... apparently HP didn't care (it wasn't a contract violation or anything) but wouldn't provide any support if you did.

Re:I hope this doesn't fly ...

[ScrewMaster]

The problem is that it doesn't make sense from the common sense point of view.

This may come down to economies of scale. If you only have to manufacture one processor, and simply deactivate parts of it, rather than tooling up for multiple chips, there might be a significant savings there. In fact, I'm rather inclined to think that that's all there is to it.

Re:I hope this doesn't fly ...

[M. Baranczak]

Bad analogies are like ex-wives.

Re:I hope this doesn't fly ...

[tgibbs]

It sounds like you are a bit unclear on the concept of pricing. Here's a clue:
The price of something has nothing at all to do with what it costs to produce or deliver--it depends only upon what it is worth to the customer.

So no, it doesn't matter whether the hardware you bought is capable of functioning as an i7, because you didn't pay for an i7, and therefore you didn't buy it.

Re:I hope this doesn't fly ...

[jpmorgan]

No, you bought a disabled i7.

Re:I hope this doesn't fly ...

[iCEBaLM]

Since its a software unlock, it can come down to a DMCA violation

I am not a lawyer and this is not legal advice. That said, good luck getting the DMCA to cover this. The DMCA makes "circumventing a technological measure that effectively controls access to a work" illegal. A "work" is defined as a copyrighted work. In this case what you are circumventing is access to features of s CPU, not a copyrighted work, like a DVD movie.

This would be beyond the scope of copyright law.

Re:I hope this doesn't fly ...

[Darinbob]

People are generally leasing those machines though.

Re:I hope this doesn't fly ...

[turkeyfish]

Yes, the next step is for the fine print to read, "this license gives you the opportunity to rent our CPU from us" for a period of 6 months after which time you will be required to pay an additional $50 per month to continue to rent it".

If consumers accept this, they can expect CPU rentals not to be far behind.

Re:I hope this doesn't fly ...

[skreeech]

In this case you have the physical hardware of an i7 but it is being restricted to an i5. The physical parts and any software is there and being restricted until you pay.

Re:I hope this doesn't fly ...

[Omnifarious]

Your analogy also fails. To go back to the knife and compass... the knives are sold without a compass, but secretly has one hidden behind a sealed metal shield in the handle. You are expressly prohibited from messing with the metal shield yourself, you have to have an 'authorized service technician' remove the metal shield for some extra cash.

And that's pretty darned stupid. I bought the knife, metal shield, compass and all. If I want to remove the metal shield and have the tools I'm darned well going to regardless of any silly rules the seller tries to impose on what I can and can't do with my knife.

Nobody should be allowed to impose post-sales restrictions on your property without a contract they negotiate explicitly with you. The key word being 'negotiate' implying that the terms can be modified in detail by you or the other party before either signs it.

Re:I hope this doesn't fly ...

[Kitkoan]

From what I know about the DMCA, it's illegal to reverse engineer a copyrighted software (and this software unlock will be copyrighted). So you won't be able to reverse engineer the program to "do-it-yourself", and the CPU could also have a hardware checks (to make sure the software is the correct Intel-only program) so you won't be able to reverse engineer the CPUs firmware under the same DMCA issues. That leaves pirating the software unlock which is still illegal under the DMCA.

Re:I hope this doesn't fly ...

[TheVelvetFlamebait]

But here it comes: have we bought a central processing unit which has now become our property because we paid for it, or are we simply buying a "license" to use Intel's "intellectual property"?

I'm not sure exactly how they would enforce it legally. This is not copy protection for some creative work, so the DMCA's anti-circumvention clauses surely don't apply. Is there some other anti-uncrippling section of the DMCA, or in some other law?

Re:I hope this doesn't fly ...

[ChipMonk]

If you want the pocket knife with a built-in compass, pay for the one that has a compass in it.

Uh, he did. The problem is that the compass is covered with opaque plastic, with a little note on it: "If you wish to have this cover removed, please contact your retailer for price quote."

If you want a car analogy: This is like a car that advertises for $22,000, but won't go over 35 MPH unless you pay an additional $3,000 to have the dealer install a jumper across two contacts on the car's on-board computer.

Re:I hope this doesn't fly ...

[ImaLamer]

I think modding consoles is only illegal as your intention is to circumvent the copyright protections. You can mod all you want to run homebrew software. That's the tricky part - are you turning on software on the chip (microcode, instructions) that are protected by this upgrade feature? Hacking the software to perform the unlock is certainly illegal, you must develop a standalone version.

Re:I hope this doesn't fly ...

[Anonymous Coward]

Mainframes are not consumer gear. In my opinion, this is a total scam and creating an artificial scarcity and demand where one does not naturally exist. If they can make the chips for $50 less, then charge $50 less or charge $50 more and profit. Sell BETTER chips that cost $50 more instead. CPU's are not bottlenecks anymore, they are milking what they can. It is close to the the hardware version of MS software assurance which may work for some businesses but would NEVER work for consumers. The days of selling a 2x cdrom one month and a 4x two months later and raising it Xx every few months after worked great until the max theoretical value was reached. Is Intel telling us that the CPU is at its theoretical max market value right now?

Re:I hope this doesn't fly ...

[vlueboy]

and if I was the CEO of AMD I'd be fricking dancing in the streets at this news.

You mention that and a future with AMD not copying this choice of CPU software paywalls. The shareholders bear more profit by copying today's move by Intel under the approach dubbed "everyone's doing this, we're just joining the cash bandwagon."

Once a similar stench hit the ISP industry a few years ago, ISP's didn't stand idly mocking competitors for the single aggro-generating move in the room. Instead, their quick and widespread annihilation of the free alt.binary usenet ensued --it was standard practice for decades to give it away free.

The economy is only the catalyst that we will indeed see the worst take place with AMD and all others. Even Microsoft has been shipping their CDs with the files in place to upgrade your Windows Home editions. We stopped owning our PC the day serial keys hit the drawing board.

Re:I hope this doesn't fly ...

[Mistlefoot]

And before I had to throw away or try to sell my old chip when upgrading. Now I just upgrade.

I really don't see why this is a bad thing...

Re:I hope this doesn't fly ...

[Chowderbags]

Let's say that you buy a house, but then the builder says "oh, yeah, there's this room in the back that's bricked up but it'll give you a bunch of extra space", can that builder restrict your ability to open up the back room using license agreements, etc (especially after all the other paperwork is done, filled out, and cash has been exchanged)? Do you have to pay that builder some more money to get the room which you weren't expecting but now know about? If you don't pay him and open it up yourself, do you lose the right to complain about a furnace breaking the day after you move in or toxic mold in the attic that wasn't mentioned?

The real question is if we should really encourage artificial barriers to computer performance just to improve a company's bottom line. To me it just feels wrong.

Re:I hope this doesn't fly ...

[Jane Q. Public]

No, you can't license hardware unless there is an agreement prior to purchase that you are doing so. U.S. courts have consistently ruled that a tangible product you buy retail is bought outright, not licensed, with complete disregard to any "licensing" language there may be on the package or elsewhere.

The only exception so far is software, and that's only because almost unbelievably, to date there has been no major court ruling on the matter. To be honest, I do not know why software should be any exception to a rule that applies to literally every other kind of product in existence.

Re:I hope this doesn't fly ...

[stonewallred]

And pirating movies is against some law or another. We see how well being against the law works in stopping that. If intel does this, it will be cracked and it will be torrented or shared through other avenues. Therefore, it is stupid to even try it.

Re:I hope this doesn't fly ...

[Jane Q. Public]

See my reply above. Unless you have made a prior agreement with Intel, if you get it from a retailer you are not licensing the hardware, you are buying it. Courts have been very consistent on this matter; there is 150 years or more of precedent.

If you buy it directly from Intel, and you have a contract ahead of time, then they could enforce licensing. But if they even do that kind of arrangement, it would apply to maybe 0.01% of end users.

Re:I hope this doesn't fly ...

[Jane Q. Public]

"And how many years did it take to crack that? Won't they upgrade the drm every year or so?"

(A) 4 years for a complete, "across the board" crack. (B) No, they didn't.

Re:I hope this doesn't fly ...

[Greyfox]

I'm pretty sure it's all purchased hardware, but mainframes are always a bit odd anyway. Buying a machine without a service contract, for example, is possible but quite stupid. I'm sure they're up-front with the customer, too. The disk arrays are actually less expensive to ship with extra disks, but the customer is aware that he's getting an array with 2tb of capacity. If he wants to upgrade to 11tb of capacity later, it's just a matter of turning on the extra disks. Same thing with the processors. You can pay more up-front if you want to, and have them all turned on, but a lot of times you really don't WANT that much processing power at the time.

I really don't have a problem with this as long as they're up-front with the customer about what they're getting and why. Somehow I don't think a Dell salesperson is going to explain that particularly well, though.

Re:I hope this doesn't fly ...

[Professr3]

That's one issue I have with those EyeFi cards - they all have the same hardware, but they just software-unlock the features for an extra bunch of money.

Re:I hope this doesn't fly ...

[stretch0611]

Even if it doesn't get cracked, we will see an "Intel Upgrade Service Crack" torrent. Malware authors love gullible people.

Of course, I am just saying that because even though it may take some time, eventually someone will crack it.

Re:I hope this doesn't fly ...

[iCEBaLM]

From what I know about the DMCA, it's illegal to reverse engineer a copyrighted software

Again, I am not a lawyer, and this is not legal advice.
No, you are reading the linked article wrong. It states that because copyrighted works may be protected by a technological protection measure, and defeating them is illegal under DMCA, and defeating them is required to reverse engineer, then by extension reverse engineering would not be possible. It's not that reverse engineering is illegal, just not possible because of the anti-circumvention clause of the DMCA. Infact the DMCA has a special clause which allows breaking DRM for interoperability reverse engineering:

Section 1201(f) states in part that "a person who has lawfully obtained the right to use a copy of a computer program may circumvent a technological measure that effectively controls access to a particular portion of that program for the sole purpose of identifying and analyzing those elements of the program that are necessary to achieve interoperability of an independently created computer program with other programs"

This however, does NOT apply is this case. What is the technological protection measure protecting the unlocking application? You could monitor everything the application does, every single command it sends to the CPU, without really doing anything to the application itself.

Keep in mind, the only circumvention prevented under the DMCA is that of protection measures that protect copyrighted works. A method cannot be copyrighted and by extension the method in which the intel software unlocks features of a CPU cannot be copyrighted. Methods fall under the realms of patents and trade secrets.

Re:I hope this doesn't fly ...

[mea_culpa]

Also similar are office copiers. Even simple things like postscript, touch screen, fax, scanning, email, etc are often disabled and enabled by a entering a key after purchase. Granted these are usually leases and could be part of the contract.

Re:I hope this doesn't fly ...

[Bing Tsher E]

In this case what you are circumventing is access to features of s CPU, not a copyrighted work, like a DVD movie.

I am certain that I have CPU chips made by third-party vendors that carry an Intel copyright notice, printed right on the body of the chip. I believe some of them are even AMD chips. AMD was a licensed second-source of Intel processors back in the 8088 and 80286 era. Intel couldn't sell their chips to certain customers back then without there being second-source vendors of the parts.

So what is copyrighted in those chips that magically makes it different than now?

Re:I hope this doesn't fly ...

[Score Whore]

A) When has bluray been cracked? I know that the HDCP master key is no longer a secret and I know that a few people have then said "bluray hsa been cracked", but those people are morons. HDCP is the encryption of the digital signal sent between a video generation circuit and a monitor/television/display. In theory you could build a small device that you hook to your bluray player and then capture a digital signal and reencode the uncompressed signal, but that's hardly a crack. It's just a reinvention of the analog hole. The fact that the HDCP master key is now known has zero relevance to bluray AACS/BD+ encryption..

B) Yes they do upgraded bluray encryption periodically. And the various tools that are used to copy said movies then need an upgrade. Happens pretty regular.

Re:I hope this doesn't fly ...

[ifiwereasculptor]

Yes, ex-wives are every bit as much of a pain the ass as bad analorgies.

Re:I hope this doesn't fly ...

[MaskedSlacker]

From what I know about the DMCA, it's illegal to reverse engineer a copyrighted software

What you know is wrong. DMCA does not outlaw the reverse engineering of software (all software is copyrighted, so talking about 'copyrighted software' is redundant). It outlaws circumventing digital locks--which is necessary in some cases of reverse engineering (say, DeCSS), but by no means all--I am free to reverse engineer any software so long as I don't bypass a digital lock in the process.

Re:I hope this doesn't fly ...

[MaskedSlacker]

Isn't it nice when you RTFA before posting nonsense that makes no sense in the context of the rest of the thread?

Re:I hope this doesn't fly ...

[MaskedSlacker]

Fuck em?

Re:I hope this doesn't fly ...

[AK Marc]

The price of something has nothing at all to do with what it costs to produce or deliver--it depends only upon what it is worth to the customer.

That's anti-free market. In a free market, someone using monopoly pricing (which is what you are describing) would have a competitor show up and instead price it on the lowest cost that generates acceptable profits. Now if only we had free-market capitalism, then your comments would be 100% wrong. So are you saying we don't have capitalism, or are you arguing against capitalism?

Re:I hope this doesn't fly ...

[Dodgy G33za]

Of course if you only have to MARKET one processor, rather than multiple, you costs come down even further, meaning the consumer ends up with a better price performance and everyone wins except marketing drones.

Re:I hope this doesn't fly ...

[lgftsa]

Assuming your frame of reference is the USA, then I believe that the recent Autocad ruling is applicable.
http://en.wikipedia.org/wiki/Vernor_v._Autodesk,_Inc.

Re:I hope this doesn't fly ...

[AK Marc]

Does the DMCA apply to all contracts? This isn't a copyright issue, it's a matter of contract law.

Re:I hope this doesn't fly ...

[AK Marc]

When has bluray been cracked?

I've seen a number of tools advertised to that effect. Are you claiming that there exists no AACS encrypted disk that can be cracked?

I know that the HDCP master key is no longer a secret and I know that a few people have then said "bluray hsa been cracked", but those people are morons.

It doesn't matter how strong you make the steering wheel lock in your car. If they can tow it, they can steal it. Even if AACS was unbreakable, "Blu-ray" has been cracked because you can get the content off in a manner designed for copying against the wishes of the content owners. Yeah, it's just a reimplementation of the analog hole, but that's still breaking the intended secure chain. You can crack "Blu-ray" as people perceive it without cracking AACS. "Blu-ray" as people see it is getting a movie off a disk, not the disk itself. They can see breaking it without actually breaking AACS. Evidently, you can't.

Re:I hope this doesn't fly ...

[rollingcalf]

I don't have a problem with them selling something advertised as X when it secretly has X+2 capability.

My problem is if they decide to take legal action against you when you find a way to enable the +2 part without the seller's help.

Re:I hope this doesn't fly ...

[gnasher719]

...except that you did buy an i7, it's just that they didn't tell you about it. Just because a feature wasn't advertised doesn't mean I didn't pay for it when I bought the hardware, or that the price I paid didn't include the cost of manufacturing that extra feature. You shouldn't be going around critiquing other peoples' analogies if you're going to liken activating hardware that you've already paid for to magically teleporting new hardware into your computer...

So here are three scenarios:

1. You have a choice of buying an i5 for $200, or an i7 for $300.
2. You have a choice of buying an i7 that pretends to be an i5 for $200, or an i7 for $300.
3. You have a choice of buying an i7 that pretends to be an i5 for $200, or an i7 for $300. If you pay $200, you can later for a payment of $100 turn it into an i7.

For me, choices (1) and (2) are identical, but choice (3) is without any doubt better. There is no situation where I am worse off than with choice 1 or 2, and in some situations I'm better off.

Re:I hope this doesn't fly ...

[icebraining]

You're assuming he's just talking about the bad:good ratio this year alone, while he may be talking about comparing this year's bad:good ratio to previous years' ratio and determining there has been a growth, which could in fact indicate a large number of bad news in the whole.

Of course, it could also indicate a change in the editorial decisions.

Re:I hope this doesn't fly ...

[AK Marc]

When you have informed consumers and zero barriers, then you have a free market. When that happens, you can't charge monopoly pricing (what the market will bear) because someone else will enter at a lower price. And no, the markup will not approach zero. If you think that, then you have no clue about finance (or are a liar trying to make others look bad by lying). Either way, that makes your opinion not worth listening to. And the fact you are an AC only supports the theory that you are a lying idiot.

Re:I hope this doesn't fly ...

[Solandri]

There's no bait-and-switch here. People are getting exactly what is advertised. Where's the problem?

The problem is that if these extras are so cheap that Intel figures they can afford to put them in every CPU even if only a few people buy them, then there's clearly a large disparity between the cost to produce the feature and the current market price for it. Long-term, this typically happens when there's a distinct lack of competition and a natural monopoly is arising. Normally, competition will drive the market price for features down to a small percentage above their cost to produce.

I'm pretty pro-free market and have eaten my share of down-ratings here for it. But that Intel is considering something like this is a pretty big warning sign that the free market isn't working as it should in this market.

Re:I hope this doesn't fly ...

[icebraining]

And do it all for $50? How?

Re:I hope this doesn't fly ...

[Joce640k]

Presumably Intel will be using the CPU serial number to keep track of legitimate users and so forth.

If you'd bother to read the article (yeah, I know it's a lot to ask...) you'd see it's done via a big blue scratch card which you can buy pay cash for in a shop.

Re:I hope this doesn't fly ...

[Joce640k]

Swiss army knives can be upgraded. eg. They sell the little screwdriver which fits in the corkscrew separately.

It costs you money so I guess that makes them a complete ripoff. I'd better start looking for another brand.

Not.

Re:I hope this doesn't fly ...

[Joce640k]

Almost every i5 in the entire world is really a perfectly capable i7.

Intel disabled bits of it purely because the distributor wanted some i5s instead of i7s that day. No other reason.

The PC you're using to read this has a deliberately-crippled CPU.

Have a nice day.

Re:I hope this doesn't fly ...

[Joce640k]

You're aware that the CPU you're using to read this text is deliberately crippled, right? Those little resistors on the back of the chip are only there to disable bits of it. They took a fully-functional top-of-the-range CPU and disabled bits of it because it's what a distributor ordered that day.

There's sooooo much more processing power in there but it's kept out of your reach because of some marketing guy in a suit.

You're welcome.

Re:I hope this doesn't fly ...

[hedwards]

Probably the same way it applies to software. They assert that you've bought a license, except when your copy gets damaged at which point you've bought a physical product and will need to buy another one.

Re:I hope this doesn't fly ...

[DaveGod]

There's no bait-and-switch here. People are getting exactly what is advertised. Where's the problem?

This has nothing to do with bait-and switch. The problem is that the strategy is only possible in a non-competitive position. Intel can only turn an i5 into an i7 if it is restricting supply of i7s through monopoly power.

Going back to the penknife analogy, the point is all penknives physically have the compass. All penknives have incurred the cost of having the compass. Clearly either (a) the cheaper penknife price actually includes the cost of the defunct compass or (b) the more expensive version is subsidising the cost of the compass in the non-compass penknives.

This would not be possible in competition because (a) there would be another company selling a knife with no compass and therefore cheaper, and (b) there would be a company selling a compass-knife that is not subsidising non-compass knives and is therefore cheaper.

This is not at all the same as binning, which is more or less where they try to make a compass-knife but the compass is defective so it is sold as a non-compass knife. As the compass does not work this always was a non-compass knife, and the cost of the faulty component actually is a cost of the compass knives (it's effectively wastage that has been recycled).

Re:I hope this doesn't fly ...

[hedwards]

The problem is that they're charging you for something you've already paid for. First sale doctrine applies and since you've bought the chip they don't have any legal right to extort more money out of you after they've sold the chip.

You're analogy is hardly analogous. It would be like buying a knife that says "Knife without compass" but somewhere in the manual it says for $10 we'll tell you where it's located. you've paid for the compass, but they want to charge another $10 to let you use it, after you've paid to have it made.

Re:I hope this doesn't fly ...

[Jeremy Erwin]

If you want a car analogy: This is like a car that advertises for $85,000, but is electronically limited to 180 kph unless you drive it at an approved race course.

Fixed it for you.

Re:I hope this doesn't fly ...

[lazyBob]

Can we receive $50 from Intel and downgrade to Core i3?

Re:I hope this doesn't fly ...

[Bronze Skarab]

The problem is that if these extras are so cheap that Intel figures they can afford to put them in every CPU even if only a few people buy them, then there's clearly a large disparity between the cost to produce the feature and the current market price for it. Long-term, this typically happens when there's a distinct lack of competition and a natural monopoly is arising. Normally, competition will drive the market price for features down to a small percentage above their cost to produce. I'm pretty pro-free market and have eaten my share of down-ratings here for it. But that Intel is considering something like this is a pretty big warning sign that the free market isn't working as it should in this market.

You don't understand economics. Just because a product has a low marginal cost does not mean their is a monopoly at play. The high upfront and fixed costs need to be recovered in order to break even. A zero-profit margin business could still have low marginal costs of sales and high revenue. That's not to say this industry doesn't have high barriers to entry through IP and the huge industrial infrastructure necessary to support the business. However. the latter at least is a natural monopoly due to the economics of the business and not something that can or should be legislated away.

Re:I hope this doesn't fly ...

[maestroX]

Not a great analogy. Can you try again only with more cars?

A V8 powered car with 4 cylinders disabled advertised as a V4 powered car.

People are getting exactly what is advertised.

Thank you for sharing the investment bankers perspective.

Where's the problem?

The deal has a bad smell to it; you buy a V8 and get the disabled V8. Whether you go for the bait or not, you either payed too much for your (disabled) V8 (at least for services you don't need such as this artificial limiting) or (and?) aided intel in chip dumping.
I think it would pretty naive nowadays to deny that complex misty constructions for selling a product are not in the interest of the consumer now and in the future.

Re:I hope this doesn't fly ...

[Threni]

In a similar situation, the same motherboards/code used to be used for all of a range of CD players (for example). A cheaper one had no remote, and could pretty much just play/pause/stop/rewind/review, but if you bought a `replacement` remote control unit for a more expensive unit, it would let you use all the other feature (ie not just remote control, but programming tracks, dividing up tracks into units of 45/60 mins etc etc). I imagine the same is true of most other consumer electronics (tv, microwave, etc etc) but just that it's not really possible to `unlock` those extra features.

Re:I hope this doesn't fly ...

[greylion3]

But here it comes: have we bought a central processing unit which has now become our property because we paid for it, or are we simply buying a "license" to use Intel's "intellectual property"?

Very good point. I'd say Intel might be looking at a class action lawsuit down the road, if enough customers buy into this, then figure they're being ripped off and decide to sue.
I suspect this came from some bean counter, who got the idea from shareware or demo versions with limited functionality, where you can purchase a serial number or key code to unlock the full functionality, and who figured this could work for CPUs too. Next up; enabling features in GPUs, and/or enabling the GPU in next-gen CPUs?

This has the potential to get ugly.

It could get even uglier - what if Intel decides to put a time-out on the upgrade, by using a tiny piece of flash memory embedded in the CPU for counting how long the CPU has been on with those features enabled?
Let's say you buy a one year upgrade, which means the CPU will run for a year (or approx. 526,000 minutes) with those features enabled, then switch those features off again.
That might come out to three years, if the PC is only on 8 hours every day.
Then those features will have become a subscription you pay for. Very ugly indeed.

Intel can be damn sure, that I won't buy into any of this.

Re:I hope this doesn't fly ...

[postbigbang]

Do you work for these guys? The HDCP master key allows the hole to be created without a "black box". The blackbox is a myth.

And therefore, it's entirely relevant to yet another black spot in Intel's long history of doing things badly, and often for the wrong reasons. That they can make CPUs is unquestionably right, and they do a reasonable job of that when they're not co-opting the competition (think HP) or attempting to destroy them (AMD). They can do that much.

Their other business units have come and gone. And good riddance.

Re:I hope this doesn't fly ...

[Nimey]

One hopes that Intel will have the price track the actual MSRP differences between models, at least. It'd be asinine to pay $50 for an upgrade when your machine's a couple years old and the parts are either EOL'd or the delta has fallen to $20.

Re:I hope this doesn't fly ...

[beakerMeep]

The situation when you are worse off is when they set the i5 price at 250 and i7 at 300 but upgrade costs $129.99

Re:I hope this doesn't fly ...

[Score Whore]

When I think, and I expect most people think, when I hear bluray has been cracked is that I can pull a bit-for-bit identical stream of the bluray disk that is in the drive, not just being able to record a copy.

Re:I hope this doesn't fly ...

[Score Whore]

The HDCP master key allows the hole to be created without a "black box". The blackbox is a myth.

That's interesting, please expand.

Re:I hope this doesn't fly ...

[sjames]

Of course, such systems were on really good support contracts and were monitored by IBM at all times. Your first notice that there was a failure might well have been the techs from IBM arriving to fix it. It wasn't a big step from having the extra hardware (that you weren't leasing at the time) sitting in a cabinet of spares or sitting in the machine ready to go but not enabled.

Re:I hope this doesn't fly ...

[Zcar]

The problem is that if these extras are so cheap that Intel figures they can afford to put them in every CPU even if only a few people buy them, then there's clearly a large disparity between the cost to produce the feature and the current market price for it.

Not necessarily. You're absolutely correct if you're talking only of the incremental production costs between the two feature sets. But, those aren't the only costs involved; Intel has to recover the R&D costs of the additional features.

Once you take these into account, you can make an argument that this is actually a good deal. This model can decrease Intel's production costs since there's only one type of chip to make. It can also decrease the cost for the limited version since they don't have to recover the R&D costs of the extra features. It probably also increases the cost per unit for the extra features because only the subset of customers that want/need those features are paying the R&D costs.

An alternative model would be to fold the R&D costs into the price of each CPU and only sell the non-limited version, raising the base cost.

Re:I hope this doesn't fly ...

[sjames]

Especially since this is coming at a time when AMD has removed the "enterprise tax" on CPUs that can live in a board with more than 2 sockets.

Re:I hope this doesn't fly ...

[Just Some Guy]

Ha! If this takes off, it'll be more like:

3. You have a choice of buying an i7 that pretends to be an i5 for $200, or an i7 for $300. If you pay $200 you can later for a payment of $200 turn it into an i7.

I mean, it's so convenient, right? That convenience has value that you should be willing to pay for!

Re:I hope this doesn't fly ...

[sjames]

There's no bait and switch, but people instinctively understand that the sale price is supposed to closely track the marginal cost of production. Speculatively including extras but leaving them disabled reveals in bold print that the market isn't sufficiently healthy to drive the price down to it's natural free market level. When it's physically separable extra hardware you can at least argue that it's just distributed warehousing.

People won't really be able to help thinking that Intel could afford to give them a better deal since flipping the switch costs nothing.

Re:I hope this doesn't fly ...

[postbigbang]

Everyone says you need something with the speed of a PLC or better to do realtime decode. That might be true if you want an inline-box, but an RTOS VM in a desktop hypervisor on a multicore machine ought to have sufficient strength to decode a realtime stream-- buffered correctly. There's enough left-over muscle on most video cards (hint: cell processors) to do it there, on the card, after an OS has buffered it all up. Objectionable aperiodic latencies could be stanched that way, too.

Re:I hope this doesn't fly ...

[greylion3]

But here it comes: have we bought a central processing unit which has now become our property because we paid for it, or are we simply buying a "license" to use Intel's "intellectual property"?

Very good point. Intel might be looking at a class action lawsuit down the road, if enough customers buy into it, then later figure that they're getting ripped off and decide to sue.
I suspect some bean counter came up with this, and got the idea from shareware/demo-versions, which can be unlocked to full versions if you pay for a serial number or key.
This is yet another example of hardware, which is defective by design.
http://www.defectivebydesign.org/
Next up; pay to enable some GPU features, or enable the GPU (or any number of them) in next-gen CPUs.

This has the potential to get ugly.

It could get even more ugly, if Intel decides to put a time-out on the upgrade.
There just needs to be a tiny piece of flash memory in the CPU, which keeps count of how many minutes it has been running with the features switched on.
Let's say you buy a six month upgrade, which will let it run for approx. 263,000 minutes. That might come out to 18 months, if your PC runs for 8 hours every day.
Then the features get switched off, and you have to pay to switch them back on again.
Effectively, the features are now a subscription which you pay for.

Intel can be damn sure, that I won't buy into any of this. And neither should you. These types of business models need to be discouraged.
Just buying the PC, and only letting the suckers buy the upgrade won't get the message through.
We need to actively warn people against buying PCs with processors like this, otherwise it's going to be that way with all processors in a few years.

Re:I hope this doesn't fly ...

[boxwood]

When you pay for a CPU how much of the price is for the R&D and how much is the cost of manufacturing?

What is the cost of maintaining a separate line running to produce and older chip which there is a demand for?

Do you really think that a faster processor requires more labour and materials to produce than an slower processor?

Quite likely intel compared the costs of keeping older production lines running vs. producing the same chip for everyone. Then checked the costs of each individual CPU. Since a newer production line is going to introduce more efficiencies to the process, they found the production costs of newer CPU weren't much higher than the production costs of the older CPUs. Given the fixed costs of running a whole other production line to make slower CPUs, it could actually cost the company more per CPU to produce the slower CPU than it is to produce faster CPUs.

Now what do they do? Stop producing slower CPUs? I'm sure AMD would love that. Continue producing slower CPUs even though it costs the company more to do so? Or produce slower CPUs on the new production line somehow.... and how do they do that? Thats right they disable features.

Re:I hope this doesn't fly ...

[boxwood]

No he's still right even in a free market.

If you competitor is offering a similar product for a lower price, then that lowers the price that the customer is willing to pay for your product.

Re:I hope this doesn't fly ...

[Garwulf]

"There's no bait-and-switch here. People are getting exactly what is advertised. Where's the problem?"

It's all about perception. Nobody is going to see it that way. What they're going to see is Intel selling them a crippled chip, and then demanding money to uncripple it. Then there's Intel's competitor, who will be perceived as selling them a chip that is exactly as advertised.

When I got into business, I didn't have a business degree, or formal training, so I had to figure a lot out for myself. I realized, after a great deal of thought, that a good, growing business ultimately came down to three rules:

1. Produce a good product that customers will want.
2. Make it as easy as possible for customers to buy your product.
3. Never give customers a reason NOT to buy your product.

Intel's move directly violates rule #3 in a pretty big way. I'd almost call it corporate Darwinism in motion, considering the PR hit Intel is about to receive.

Re:I hope this doesn't fly ...

[Alsee]

when I hear bluray has been cracked is that I can pull a bit-for-bit identical stream of the bluray disk that is in the drive, not just being able to record a copy

There can be some nitpicking back and forth on precise definitions, but with the HDCP crack software can be written to pretty much do that. The movie goes through a lossy compression, which is then recorded on the disk. That data gets decompressed and can be captured via the now broken HDCP. I don't know if anyone has bothered, but I am certain it is possible to losslessly recompress the expanded data back down to the format that was stored on the disk. This will produce either a bit-for-bit identical copy of what was on the disk, or a precise data synonym of the bits on the disk. I have not looked closely at video formats and how I'd write this sort of lossless recompression software, but to be fair I'll admit I can see reasons why it may (or may not?) run significantly slower than the usual step of running a new lossy compression.

-

Re:I hope this doesn't fly ...

[paramour]

So here are three scenarios:

1. You have a choice of buying an i5 for $200, or an i7 for $300.

2. You have a choice of buying an i7 that pretends to be an i5 for $200, or an i7 for $300.

3. You have a choice of buying an i7 that pretends to be an i5 for $200, or an i7 for $300. If you pay $200, you can later for a payment of $100 turn it into an i7.

For me, choices (1) and (2) are identical, but choice (3) is without any doubt better. There is no situation where I am worse off than with choice 1 or 2, and in some situations I'm better off.

You left out one significant scenario:

4. You have a choice of buying an i7, that acts like an i7, for $200.

Choice (4) is clearly best for me as a customer.

For Intel, (4) isn't any worse than (2), as clearly they think they can make a profit selling i7s at $200. For (3) it's only worse to the extent of their conversion rate, probably well under 10%, minus the cost to lock the chip, or less than $10. For (1), they are apparently just overcharging by $90 - $100.

Re:I hope this doesn't fly ...

[neovoxx]

Except that your i5 is already actually an i7.  When a CPU or any other hardware component can't be certified as fully functioning, it is re-binned as a lower model component.  When failure rates go down, they keep doing this, but rather than having a partially bad component, you have a perfectly good component with some functionality removed to it can be sold to you as a cheaper part.  This is done so the company doesn't have to manufacture so many different parts.  If you can make a CPU and just disable certain parts of it and sell it for less while keeping your manufacturing costs down because you're actually only making one part and reconfiguring it, do it.

Same thing goes for software.  Are you really going to complain that you bought a $10k piece of software but only paid $100 for it because your license key says parts of the software are disabled?  No.

Re:I hope this doesn't fly ...

[BitZtream]

No one thinks its about bait and switch, its that it is clearly nothing more than a way to rip people off.

People see Intel selling a i7 at an i5 price with parts disabled which costs the same as a normal i7 to make, just so people will pay more for the real i7.

Its just wrong. It may be 'acceptable' business, but it takes all of a millisecond to realize its wrong and shows just how amazingly greedy they are. When people see that level of greed they tend to back off and be weary of what other ways they are getting screwed.

Its obvious greed, and that bothers the hell out of most people. That sort of greed comes from people you KNOW you don't want to deal with cause its just a matter of time until they'll fuck you. Without lube.

Re:I hope this doesn't fly ...

[BitZtream]

Thats just because you're too stupid to realize that you are paying WAY too much for the i7 in the first place.

It bothers the rest of us because we realize that if they are selling it at an i5 price, then its clearly profitable enough at that price range, the extra cost to 'upgrade' is just greed.

That level of greed scares most people that recognize it. When someone is that greedy, most people realize that makes their screwing you in every way they can.

Re:I hope this doesn't fly ...

[Omnifarious]

And that's interesting to know, and I believe it. But it doesn't change anything I said. While it's lots more expensive than a software hack would be, I could go and get the crippling reversed.

Re:I hope this doesn't fly ...

[Jane Q. Public]

Yes, I had forgotten about that recent case. Bad precedent it is, too. But I still want to know: why should software be treated so differently that every other product under the sun?

Re:I hope this doesn't fly ...

[martyros]

But that Intel is considering something like this is a pretty big warning sign that the free market isn't working as it should in this market.

What do you mean "as it should"? If the overall result is that resources are saved by Intel only making one set of chips, instead of having to go through all the design and manufacturing process for two, then I think that's a good result.

Re:I hope this doesn't fly ...

[Jane Q. Public]

As far as the studios are concerned, I am sure, Blu-Ray is "cracked", since their interest (as they have always said) is to prevent digital copies of their material.

It's true that Blu-Ray in itself may not be cracked, but in some ways an HDCP crack is even better. Not just Blu-Ray but any other DRMed product that uses it it is now susceptible to lossless digital copying.

Reluctant as they have been to see reality (perhaps because the bigwigs are technologically ignorant), the "content industry" should take this as a sign: "DRM" is doomed to fail. Always. It always has, it always will. It is based on a failure to recognize a fundamental principle: in order to use the material at all, it has to be decoded at some point. Making that point difficult to get at may work for a few years, but then it will be broken and that massive investment will be shot to hell.

Re:I hope this doesn't fly ...

[dcposch]

I can see where you're coming from, but what if the majority of the cost of a CPU is simply there to amortize R&D? From what I've read, this seems to be the case. Intel spends $6 billion a year in R&D.

Even if the unit production cost on their current, mature production lines is a small fraction of the sale price, that doesn't necessarily mean that the sale price is too high, or that Intel has a monopoly.

By the way, I'm not suggesting that they don't have monopoly. Intel has a $100b market cap at this point, which is more than 20 times that of their main competitor, AMD. That certainly leaves room for suspicion. All I'm saying is that the fact that the marginal cost of adding cores to chips is small doesn't, by itself, indicate that Intel is behaving like a monopoly.

Re:I hope this doesn't fly ...

[Tacvek]

A perfectly competitive market would not have zero markup, that is true, since some markup is necessary to cover the other costs. That said a perfectly competeative market would have (asymptotically) zero profit, according to basic economic theory theory.

That said such theory is still incorrect, because it fails to account for the market of production factors. There is a fixed number of people willing to work minimum wage making your Widget. Let us say there is only one company in the market, they pay only minimum wage, and sell with a moderate profit (where profit=revenue - all costs (including wages)).

The theory says I could undercut that company, by starting up my own widget factory (these widgets are standardized commodities so my product is identical). There is no cost to set up, since there is no barrier to market. The theory says that if I sell for less profit, I will be selling for a lower price, and make all the profit. That said, I might not be able to hire a full crew at minimum wage, since there are only a limited number of workers willing to work making widgets at that price, and the other company has most of them. Therefore my costs are not equal to the other company, and I might not be able to undercut them.

Re:I hope this doesn't fly ...

[rundgong]

I am not going to argue that this is or is not a working free market, but I'd like to point out that the selling price not only covers the manufacturing cost. The selling price must also make up for R&D costs.

Re:I hope this doesn't fly ...

[ScrewMaster]

Yes, I had forgotten about that recent case. Bad precedent it is, too. But I still want to know: why should software be treated so differently that every other product under the sun?

I think this comes under the "why do dogs lick their balls" heading.

Re:I hope this doesn't fly ...

[ScrewMaster]

Making that point difficult to get at may work for a few years, but then it will be broken and that massive investment will be shot to hell.

Well, here's the thing. It's not about absolute security ... it's about keeping the vast majority of consumers under lock and key. And, even with DeCSS and this HDCP crack, they still are, and always will be.

The people that run the media companies are not ignorant: they understand very clearly the intent of DRM. This is about raising the bar high enough to prevent casual copying, no more and no less, and for most people, good old CSS does just that.

Re:I hope this doesn't fly ...

[Yaa 101]

Wow you are nice.

The economy works like this: you ask the price that you get away with!

Re:I hope this doesn't fly ...

[AK Marc]

That said a perfectly competeative market would have (asymptotically) zero profit, according to basic economic theory theory.

That's incorrect. Zero profit is a loss. When I can earn 1% in a bank account, I should sell the business and put my money in the bank. If profit is zero, then all businesses should close. That's basic economic theory.

The theory says I could undercut that company, by starting up my own widget factory (these widgets are standardized commodities so my product is identical).

No, basic economic theory is remarkable non-competitive. You don't try to "beat" others, you try to make a profit. That may require entering a market where there is room to charge less than the existing players and still make a profit. But "basic" theory is more about supply and demand which is not the kill or be killed cutthroat economics people assume.

Re:I hope this doesn't fly ...

[Jane Q. Public]

Well, the same is true about the lock on your front door. If most people understood how easy it is for a knowledgeable person to defeat the vast majority of them, they would be horrified.

Even so, I think you underestimate the "average user" these days.

Re:I hope this doesn't fly ...

[toddestan]

Exactly. It's going to be just like Microsoft's "Anytime" upgrades for Vista and 7, where the cost of a lower edition + upgrade is significantly higher than the cost of just the higher edition to begin with.

Re:I hope this doesn't fly ...

[alexo]

The problem is that if these extras are so cheap that Intel figures they can afford to put them in every CPU even if only a few people buy them, then there's clearly a large disparity between the cost to produce the feature and the current market price for it. Long-term, this typically happens when there's a distinct lack of competition and a natural monopoly is arising. Normally, competition will drive the market price for features down to a small percentage above their cost to produce.

By the same logic, it sure doesn't cost $2,599.00 in bandwidth for Adobe to let you download Creative Suite 5 Master Collection (warning: Flash).

Re:I hope this doesn't fly ...

[adn]

You are assuming that the price doesn't change when the option is 1, 2 ou 3. Also, the consumer pays the price expecting that the manufacturer had some less trouble and difficulty, less cost, to get to you the cheaper option. It's a perception issue. In your proposed comparsion, if the manufacturer has so many i7 so he may tune some down to i5, the consumer would find more fair if the manufacturer settled down the price on i7, just as the free market rules work. If that isn't happening, maybe it's a monopoly rising.

Re:I hope this doesn't fly ...

[sciencewhiz]

(4) is worse then (2) for Intel. With (4), they can't sell anymore i5s.

Re:I hope this doesn't fly ...

[Gaffod]

This is a case of you buying a Core i5 and Intel saying "here is exactly what you paid for, and by the way, if you ever decide you should have bought a Core i7 instead, we can magically teleport one into your computer for just $50".

Well if they can afford selling me an i7 for the price of an i5, even though I might never unlock, why don't they just go ahead and do so? I don't understand, is there a reason why processor must cost a certain amount of money, even if all the expenses going into making them are less?

If indeed the company runs out of low end processors, and starts selling crippled high end ones, why bother crippling them? Just sell the high end unit for the low end units price, hell call it a sale or limited time offer or whatever. Won't more people buy an i7 than an i5 if they were the same price? It's not like they're stuck with it later either, I don't think Intel is seriously unable to say, "sorry guys, our stocks are properly balanced now, back to the usual prices".

Re:I hope this doesn't fly ...

[Tacvek]

A quote from an article covering perfect competition: "In contrast to a monopoly or oligopoly, it is impossible for a firm in perfect competition to earn economic profit in the long run, which is to say that a firm cannot make any more money than is necessary to cover its economic costs."

Claasical economic theory ignored the possibility of erning interest by investing the money in alternative measures. Only neoclassical economic theory accounts for that. In neoclassical theory, profit is defined including subtracting out the interest that would have been earned by low risk investment instead.

Can you hear that?

[Anonymous Coward]

Can you hear that?

That's the sound of so many informed geeks switching to AMD.

Re:Can you hear that?

[hawkingradiation]

Sometimes it is harder to get an OEM computer to use AMD (like apple) but according to AMD's website: Powering ultrathin notebooks to blade servers, all AMD processors shipped are designed to use AMD-V features. Where as Intel has been a little less free and more cumbersome. For instance most Atom processors by Intel do not support virtualization but all shipping AMD (and it has been a while) do. Also computer models such as the sony viao (undercapitalized for a reason) use the "feature" provided by Intel to disallow virtualization through the BIOS, meaning that you have to turn in on before booting. Along with other technology that AMD has developed makes you wonder why Intel is so dominant in the space. So for an informed geek, switching to AMD was already a good move, if only the manufacturers would follow.

Re:Can you hear that?

[0123456]

Along with other technology that AMD has developed makes you wonder why Intel is so dominant in the space.

Intel are bigger and better able to tune products to different markets. Leaving out virtualisation on Atoms means anyone crazy enough to consider doing so on an Atom will have to buy a more expensive CPU instead, and AMD don't have much to compete in the Atom's market.

AMD end up having to sell working quad-cores as dual-cores at significantly lower prices because they can't afford to build two different chips. In this case Intel seem to be selling intentionally crippled i3s for almost the same price as a fully functional i3, and then charging $50 on top to enable the i3 functionality that's been there all along... you'd be better to buy the real i3 in the first place.

Re:Can you hear that?

[0123456]

Why? Seriously, exactly what is it about this that would piss off a rational human being? Are you _seriously_ claiming that _not_ having the ability to upgrade your processor for a fee is a good thing?

Are you claiming that buying a car with a wheel-clamp on one wheel and being allowed to pay to upgrade to four working wheels would be a good thing?

Being sold something that's deliberately crippled even though all the hardware is there really, seriously pisses off an awful lot of people.

Re:Can you hear that?

[Sponge Bath]

This is a test by Intel. If it fails, Intel will drop it. If it succeeds, AMD will adopt it.

Re:Can you hear that?

[thegarbz]

Just like all those informed geeks switched to Intel when it was revealed that AMDs could be unlocked with a pencil, and all that separated one processor to the other was a little trace printed on the outside of the chip? The companies have been doing this stuff for years, releasing chips that don't quite meet quality control as a slower chip, and when the distribution channels crash due to unforeseen circumstances, ship identical higher speed chips with the clock turned down even though there's nothing wrong with them. It was more a matter of time before someone would monetize this.

If the price is right, all you'll be hearing is the sound of business as usual, or perhaps even a boost in Intel sales as like all things when you give something out on good faith, someone will crack it.

Re:Can you hear that?

[Kitkoan]

Why? Seriously, exactly what is it about this that would piss off a rational human being? Are you _seriously_ claiming that _not_ having the ability to upgrade your processor for a fee is a good thing?

Are you seriously claiming that being sold a crippled product that you have to pay more then the listed price to become fully functional is a good thing?

Re:Can you hear that?

[precariousgray]

How is this really any different from shipping a CPU with a locked multiplier? Things have been this way for a while.

Re:Can you hear that?

[MadUndergrad]

The AMD thing should attract geeks because hey, all you need is a pencil to upgrade! No, this is different. You have to imagine all the bullshit Intel will have to shovel to make sure no one jailbreaks their chip. You might wind up seeing the BSA, the DMCA, and god know what else involved. No one wants that, lease of all geeks.

Re:Can you hear that?

[trickyD1ck]

Yes it is. You are still able to buy the full functioning product for the full price. Additionally, you are not able to save money by buying a reduced spec product.

Re:Can you hear that?

[makomk]

The companies have been doing this stuff for years, releasing chips that don't quite meet quality control as a slower chip, and when the distribution channels crash due to unforeseen circumstances, ship identical higher speed chips with the clock turned down even though there's nothing wrong with them. It was more a matter of time before someone would monetize this.

Except that Intel are guaranteeing that all their CPUs have the ability to unlock this functionality (so they can't ship CPUs where it's been disabled for not meeting quality control) - they just won't let you use it unless you pay. Whereas AMD's chips often genuinely didn't pass QA at the higher functionality, and when they do work AMD seem to be OK with people enabling entire CPU cores that weren't advertised as working even though they don't get a penny from this.

Re:Can you hear that?

[Joce640k]

Guess what, they even do it with chips that made quality control. They make the best quality chips they con possible make and set the clock speeds according to what the distributors have ordered, not according to what the chips are capable of.

Re:Can you hear that?

[Joce640k]

How on earth did that get modded 'insightful', it's stupidity incarnate.

The ONLY difference between this chip and most of the other chips you've ever owned (AMD included) is that the 'locks' aren't soldered onto the chip. If this is a ripoff then all those other chips were too.

Re:Can you hear that?

[thegarbz]

Isn't that what I said ;)

Re:Can you hear that?

[kenh]

I think not. Feigned moral outrage over a non-story like this won't shift market trends.

How many "informed geeks" will be impacted by this experimental move in one low-end laptop?

If anything, they may wind-up buying this laptop to try and crack the upgrade code - it could actually help sales.

Re:Can you hear that?

[Jeremy Erwin]

It's more like buying a pickup with an optional fifth wheel. Don't need to haul a trailer? Don't unlock the fifth wheel.

Re:Can you hear that?

[yakovlev]

It's actually a little more complicated. You see what speeds the chips come out at, and then look at how you can set multiple price points to make the most money by selling the slowest chips at a discount and selling the fastest chips at a premium while throwing away the very slowest chips.

Sometimes, that means you have to sell chips that would work at a higher sort point in a lower bin.

The idea here is to sell something like this:

2-core: $200
4-core down-binned to 2-core: $220 (Market as 2-core upgradeable to 4-core)
4-core: $300

Charge $100 to upgrade the upgradeable parts.

For the consumer, this means:
1.) If you need a 4-core now, buy one.
2.) If a 2-core is all you will ever need, buy that.
3.) If you might need a 4-core sometime in the future, but you might buy a new PC instead, buy the upgradeable chip.

Now, my guess is that the average consumer is going to buy either the fastest chip (which they often don't need) or the cheapest chip (which is usually good enough), so I don't think anyone would buy the upgradeable chip. I also think they'd be right, in that if you don't need the 4-core now, you are probably better off just buying a new PC when you do.

Re:Can you hear that?

[toddestan]

It's going to be more like the budget CPU is $100, the nicer budget CPU is $125, but if you buy the budget CPU then upgrade it you get to pay $150.

Also, I fully expect Intel to screw you over with incompatible sockets to, so you can't just buy a better CPU a few years down the road for $50 and drop it in.

Re:Can you hear that?

[toddestan]

So far, I haven't seen AMD adopting some of Intel's more obnoxious market segmentation techniques, such as intentionally removing features on low-end CPUs like virtualization and ECC memory support, not to mention the multiple incompatible sockets.

Re:Can you hear that?

[jonathanleong]

Also computer models such as the sony viao (undercapitalized for a reason) use the "feature" provided by Intel to disallow virtualization through the BIOS, meaning that you have to turn in on before booting.

The decision to disable VT in the Sony VAIOs was apparently done out of security concerns as determined by their engineers. No details were provided so I can't say if this was actually true or not. (And I'm too lazy to find the link). In Sony's defense,once they realized there was strong demand they did release a BIOS update and promised to not disable VT in future models. Problem solved. I can go back to testing operating systems with all the advantages of VT support.

models such as sony viao (undercapitalized for a reason)

btw, why the under-capitalization? Something against Sony?

Re:Can you hear that?

[kenh]

R U Serious? How many "informed geeks" are currently "loyal" to Intel, but will switch to a new CPU mfg. (AMD) for no other reason than Intel sells low-end processors that can be upgraded?

Intel has always sold parts that were capable of higher speeds, set at the factory to lower specs at a lower price, it's called binning (I'll bet AMD does the same think, disabling one core in a quad core CPU and calling it a three core CPU, but that's just a guess), difference is, AMD won't give you any way to activate that fourth core - in this instance, Intel is making the upgrade available for $50.

No one crys foul when a 3.2 GHz capable CPU is sold as a 2.4 GHz part, permanently hobbled at the factory to identify as a 2.4 GHz part and requires various BIOS-foo to realize it's full potential, but when Intel provides a convienient path to overclocking/upgrading a CPU that is not otherwise upgradeable, suddenly the outrage is palable?

I have to believe you are writing from an emotional point of view, based on a child-like view of the CPU industry.

Let's hope they used the same key as HDCP

[laing]

I believe that HP/Agilent was the first company to do this. They would manufacture something with gobs of RAM and then charge you extra money to enable the 'option' that was already present. It costs less for a manufacturer to produce a single version of their product than for multiple versions with different capabilities. Intel realizes this but their marketing people are full of shit (just like HP's were). They didn't lose any money when they sold you the processor. The software unlock is 100% pure profit. It's really annoying to know that you have paid for and posses capability that you cannot use.

Re:Let's hope they used the same key as HDCP

[SIR_Taco]

Intel used to do something similar with the Celeron series... they were just chips where some of the cache didn't work on. It was caught during manufacturing/QA and deemed a lower grade. Maybe the manufacturing process has reached a point where there are much fewer 'duds'.

I have no problem with a hardware company selling lower-grade hardware for a discounted price.

I do have a problem with a company selling a piece of hardware that is [i]intentionally[/i] down-graded. The cost of manufacture is equal to it's counterpart.... both should be equally expensive/cheap.

Re:Let's hope they used the same key as HDCP

[Anonymous Coward]

The product you refer to is called HP Instant Capacity (iCAP)

https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B9073BA

Re:Let's hope they used the same key as HDCP

[icebike]

When was that?

I specifically remember working on a Control Data Corp CDC3200 (circa 1970) which was nothing but a 3300 with delay lines installed to slow it down. Simply removing a specific card (secret) did the upgrade.

Re:Let's hope they used the same key as HDCP

[clifyt]

"Intel used to do something similar with the Celeron series..."

And it was happening years before that too. I owned a 486SX machine and was looking to upgrade (Doom was big at the time...the original) and played badly on the SX, but was decent on the DX. Turns out that much like the celeron, the math coprocessor was often damaged on the chip and they just downgraded it to the version without the math. But after a while, it was nearly impossible to find one that was actually damaged, but it was still sold this way.

I forget what it took to change the chip...I think there was a jumper that was physically cut and one could solder it, or maybe put it into a different socket that enabled it (it was like 92....I can't remember that far back!) but I remember it wasn't that expensive or difficult -- and much cheaper than buying a new chip. And worked perfectly (and even on the ones that it didn't, there were math libraries that were optimized to look for the errors and would revert to the standard if errors did occur).

But yeah, this has been going on for YEARS....

Re:Let's hope they used the same key as HDCP

[LordLucless]

That's a bit different. They disabled a chip, and sold you the board. You could turn it on if you wanted - with no guarantee that it'd actually work. They do the same with multi-core CPUs. They build em all with X cores, during QA, if one of the cores fails a test, they'll just disable the core, and sell it as X-1.You're perfectly free to try and re-enable your disabled core - but there's a chance it won't work. They can't sell them as X-core processors, because they can't guarantee all cores will work.

This, on the otherhand, is them selling you a piece of hardware with functionality *that they know is 100% functional* (or they wouldn't be able to offer the upgrade) and are trying to make you pay for what you already own.

Re:Let's hope they used the same key as HDCP

[alanshot]

Packard Bell used to do something similar.

my first 486 system I bought back in the early 90s came in two flavors: the SX version and the DX version.(For those that dont remember, the DX had a math coprocessor, the SX didnt.)

  It was about a $50 difference in price between the two models, and so I bought the cheaper one.

One day I was skimming the manual looking for a motherboard jumper and found a cryptic note for "J12 1-2 SX/2-3 DX). On a whim I swapped the jumper.

Whadda ya know! suddenly my bios reported DX processor.

Apparently it was cheaper to build one box, then apply a case sticker/jumper setting to differentiate the two.

makes perfect sense really.

Re:Let's hope they used the same key as HDCP

[thegarbz]

Actually due to economies of scale this kind of process is applied not only in a shitload of products (Tektronics produced oscilloscopes that where limited only by firmware and their input low pass filter long before Agilent was even spawned by HP), but actually to CPUs in general. Ever wonder why some CPUs seem to overclock so much more gracefully than an identical other product? One may not be able to meet the required quality control, the other may be the result of the manufacturer not meeting the required supply contracts. It's quite common practice to cripple a product and send it out for cheap since economies of scale are improved if you don't create 20 different product lines.

Or we could take it to extreme and look at Creative X-Fi soundcards. How did that go again? Creative didn't give a shit about it's customers so it let it's customers write their own 64bit drivers only to find out that all the cards in the range are identical and the drivers themselves are the only thing that separates them? Or what about the NVIDIA Quadro graphics card I had in my computer that I bought off the shelf for $250. The card may not have said Quadro on it but at least the software recognised it as such and enabled the required hardware acceleration in the CAD programs I was using.

Re:Let's hope they used the same key as HDCP

[Jane Q. Public]

This, on the otherhand, is them selling you a piece of hardware with functionality *that they know is 100% functional* (or they wouldn't be able to offer the upgrade) and are trying to make you pay for what you already own.

It isn't different. They did that with the 486 as well. But there was a HUGE customer backlash about it and they ceased the practice.

Intel later claimed that the SX processors it sold had 487 (math coprocessor) units that didn't pass the necessary tests. (E.g. -- numbers pulled from the air -- the 487 portion of the chip might only operate properly at 8 MHz while the main part functioned fine at 10 MHz, or other such problems.) However, the sheer number of SX chips that were selling would have meant that Intel's manufacturing yield was really abysmal, if all the 487s were in fact defective. They weren't believed and there was good reason.

They quickly backtracked and changed direction on that. But they should have learned their lesson. Looks like they haven't. Sadly, there will probably be less backlash this time because the average user is less technically savvy than back in '92. (I don't mean that people in general are... but the majority of PCs were still business machines back then, administered by some kind of IT tech.)

I hope I'm wrong. This is a bullshit way to market products. Imagine if you were sold a factory automobile, perfectly capable of 100 mpg with no loss of performance, but which had to be "tuned" with special software to do so, at a cost of an extra $1000. It would piss a lot of people off!

Re:Let's hope they used the same key as HDCP

[precariousgray]

Paid for and possess capability that you cannot use? No, you paid for X GHz. If you want Y GHz, you pay more.

This is not news to anyone who has been paying attention for the past decade, at the last.

Re:Let's hope they used the same key as HDCP

[precariousgray]

Speaking of paying attention to things that have occurred over the past decade, it's good to see I still have to manually enter line breaks via HTML and can't edit my posts to correct alcohol-infused typographical errors because this is Slashdot.

Re:Let's hope they used the same key as HDCP

[jeti]

It costs less for a manufacturer to produce a single version of their product than for multiple versions with different capabilities.

That's not why Intel is doing it. If you produce computer hardware, you only get a fraction of what it sells for in the store. The retailers take a big chunk and the taxes and import tarriffs stack up. Selling a software upgrade - preferably online - is pure profit.

Re:Let's hope they used the same key as HDCP

[giorgist]

Yes but if $100 is what intel woudl like to get for the processor on average,
would you not mind if intel sells you three processors at $80, $100 and $120 ?

Or what about if you can get the processor at $10, and intel knows 50% will pay up $100 later for an upgrade, but you can also buy it up front at $100.

If $10 gets you 40% the $100 CPU, are you not ahead ?

Re:Let's hope they used the same key as HDCP

[yoshac]

"Imagine if you were sold a factory automobile, perfectly capable of 100 mpg with no loss of performance, but which had to be "tuned" with special software to do so, at a cost of an extra $1000. It would piss a lot of people off!" No need to imagine... http://jalopnik.com/cars/news/so-long-guvnor-mercedes-will-unlock-top-speed-on-amg-models-in-the-us-for-a-price-154226.php

Re:Let's hope they used the same key as HDCP

[boxwood]

not 100% pure profit.... the price of the software unlock represents the cost of R&D for the new chip.

Re:Let's hope they used the same key as HDCP

[yakovlev]

That's a fascinating analysis, and I think it makes a lot of sense.

This could be a way of Intel screwing the PC manufacturers by selling directly to the consumer. To the consumer it may look like an upgrade costs the same as the more expensive processor, but Intel makes more profit while the PC maker makes less. If an upgraded part costs less overall than an initially full-spec part then this is almost certainly the case.

Re:Let's hope they used the same key as HDCP

[BitZtream]

Lets also not forget that Intel tends to spew lies in order to force changes to get their way, like reduce compatibility.

Remember 'slotted' processors ... which were supposedly because BGAs and PGA sockets were no longer capable of handling the high frequency communications needs ... (When any EE knew better instantly) and it turns out really that they switched to slots just so AMD wouldn't be able to follow them due to patents, hence forever ending the time when you could drop an AMD or an Intel processor in the same socket and actually use the machine. Funny how we're back to BGAs and PGAs again.

Ridiculous...

[Shyfer]

This is so ridiculous (and should be illegal)... And I tought we had control of our PC hardware, I guess in future PC hardware and architecture will be closed like game consoles are now. Also, I have read that they only unlock extra cache that was already there but showed some problem. So they are selling 'unlock cards' to unlock defective cache? Sounds good.

And Engadget seems to support it, at least their post make sound they do, corporate drones...

Re:Ridiculous...

[dimeglio]

Look at it this way: you buy a CPU at $200 with one core. A year later, you need more performance. Instead of trashing the entire computer (ram, cpu, and motherboard at least), you simply pay a mere $50, unlock 3 more cores, booth the clock by 100% and throw-in hyperthreading. You'll extend the life of the unit for at least another year saving a few hundred dollars. Make it 6 months and another 6 months but the idea is the same.

I might work great if the price and options are right.

Re:Ridiculous...

[Shyfer]

Yeah, unfortunately you are right. Good old times when we had complete control of our PC hardware.
I still cant understand what a company earns closing everything (I'm looking at you apple), I feel that if a company projected a brand new PC architecture, very optimized and stuff, completly open, linux support, etc they would be very successful. Not only geeks would fall in love with it, but also small (and big) companies that need a new architecture but dont have the money to develop it would. Sadly I dont see this happening =/

Re:Ridiculous...

[Kitkoan]

Look at it this way: you buy a CPU at $200 with one core. A year later, you need more performance. Instead of trashing the entire computer (ram, cpu, and motherboard at least), you simply pay a mere $50, unlock 3 more cores, booth the clock by 100% and throw-in hyperthreading. You'll extend the life of the unit for at least another year saving a few hundred dollars. Make it 6 months and another 6 months but the idea is the same.

I might work great if the price and options are right.

Or look at it this way: you buy a CPU at $200 to have access to 100% of it, and then a year later you don't need more performance since you already bought all of the chips performance at the beginning (like most physical objects). No needing to spend more money just to open up what you already bought and paid for.

Re:Ridiculous...

[Shyfer]

Look at is this way: you buy an $200 CPU and get what you pay for, instead of having to pay another $50 to unlock what you already paid for just because the company that sells it to you is too greedy.

Re:Ridiculous...

[Nemyst]

It also means that for a full year, you've been wasting 3 cores for no purpose at all when you could've used them right away otherwise. I'm sorry but whichever way I'm looking at, this is the very definition of getting shafted. Would you want your Ferrari's 5th gear to be an unlockable?

Re:Ridiculous...

[Pentium100]

No. This is pure money grab. You buy a CPU that cost the same to manufacture as the higher end CPU, but is crippled until you pay more money. Which means, that the higher end CPU also costs the same to manufacture, which means that the company is too greedy.

This is the same as buying a car with a lock placed over the radio, unless you pay some more money. The radio is still there, so the car with locked radio was not cheaper to make, yet they ask for additional money for it.

Yes, I know that the price does not really depend on the cost of manufacture, but the companies do not need to actually show that to everyone.

Re:Ridiculous...

[__aajfby9338]

Except that I've already paid for that hardware with the original $200, and Intel made a profit on it unless they were daft enough to sell it to me at a loss. It cost a fixed amount to build that chip, based on wafer cost, die size, test time and yield. It'd be one thing if they took a bunch of chips in which some of the nonessential features failed final test and then sold them at a lower cost instead of throwing them away, but these proposed feature-locked chips are necessarily fully-functional chips in which they've chosen to hold some of the features for ransom. This is simply price gouging.

This is just like paying $20,000 for an SUV, and then later paying another $5,000 for the key that opens the back doors and the cargo area once I've decided that two seats and a glovebox aren't enough for me.

Re:Ridiculous...

[Z34107]

Remember, this is Intel. He's spot-on.

Re:Ridiculous...

[RightSaidFred99]

Please present a rational reason for your dislike of this. Note than hand-waving and nebulous comments about "control" don't cut it as it's not rational.

Scenario one, you buy a 3GHz processor that has 512K cache, and it's not upgradeable. You pay $99.

Scenario two, you buy a 3GHz processor that has 512K cache for $99. You can pay $50 to upgrade it to 3.2GHz and 1M cache.

I defy you to logically explain how scenario one is better.

Sure, you'll make a facile point that they should just charge you $99 for the 3.2GHz 1M chip since they sold the physical item for that price. I'll point out that how Intel bins, markets, and sells its own product is its business. They don't charge you based on what they pay for the chip, they charge you what the market will bear based on competition and availability (supply and demand).

Re:Ridiculous...

[mysidia]

Wasn't there recently an article about Nevada governer candidate wanting to sell 'speeding passes' for the ability to drive your car 90 MPH? (with the only caveat being you will need to pay the fee to activate the electronic beacon, and get an annual safety inspection of your vehicle in order to do so)

Re:Ridiculous...

[RightSaidFred99]

Opportunity cost, genius. If I sell you a my car for $300 and I could have gotten $500 for it, I just lost $200.

The entire response to this thread is just depressing. I had no idea people were so clueless and irrational about business.

Re:Ridiculous...

[Shyfer]

If the 3GHz processor that has 512K cache costs $50 to manufacture, the upgradeable version will not cost $50 too. So if today you pay $99 for that 3GHz processor that has 512K cache tomorrow you will pay $150 for it +$50 if you want the upgrade.

Also, my main point is that it is ethically wrong. Yes, I do know that 'they charge you what the market will bear based on competition and availability', and I'm saying that this is wrong (capitalism is wrong), you should pay what it costs not more or less (actually only more ) depending on the market demand.

Re:Ridiculous...

[mysidia]

Actually, I think this is more like... your car came with a satellite GPS system, but you need an activation code to use it, or you need to buy software to use it. Or you need to pay an annual fee for the ability to use the radio receiver to pick up traffic data that is broadcast over certain frequencies.

GPS manufacturers sell annual updates to map data, by the way, and if you happen to have 2 or 3 units of the same model number, you still have to buy the update 2 or 3 times. You get the software in a form that can only be activated once, so they intentionally prevent you from updating the data on multiple units.

What's interesting here is.... this... Intel already does this. They already send you CPUs with features disabled when you buy the lower end model. They just simply provide no means whatsoever to turn back on what they have disabled.

If they hadn't disabled it in the first place you would pay more

So i'm not entirely convinced their change to allow you to activate this functionality you aren't paying for up front is entirely bad.

(The functionality definitely has a market value, you would definitely pay more if the functionality was not disabled)

Re:Ridiculous...

[-noefordeg-]

Wow! That was stupid. Actually a bit retarded!

Would you say the same about cars, books, reading lights, washing machines, and what-not?

When you need a larger car, you send in $10000 to the car company and receive some new keys which unlocks the passenger doors? You save $20000 over buying a brand new car. Now THAT is almost a steal! =O
When you are almost done reading the book, you fork off some extra money to be able to finish the book completely. Saves you money from buying another book...
When the book you upgraded comes with smaller font, you shell out some cash to increase the power of your reading light? Saves you money from buying a brand new light. Or when times come to get a bigger TV, you just pay $500 and get to unlock the last 10 inches on your television. This is SO CLEVER!

Maybe you should unlock the rest of your brain... Maybe you wouldn't try to make this sound like something smart, clever, and beneficial for the human society.

Re:Ridiculous...

[Shyfer]

"It amazes me that nerds can have such irrational, retarded points of view sometimes." I say the same thing.

Yes, of course they are fucking greedy, they are a company, and that is still wrong.

And yes, that would be terrible, and I already said my reason.

Re:Ridiculous...

[mysidia]

In which case, the CPU would cost you $100 instead of $50. Because that's what they determine to price those cores and cache at otherwise.

It's not and never was a question of "costs to them" in order to manufacture your individual CPU. Many costs are support-related and also...

The cost to them is more complicated than that, and includes things like the opportunity cost of providing you a more powerful CPU that will cause you to not make another purchase from them in a year.

Their possible lost sale of a CPU to you one year from now is a cost to them, that they incur, if they provide you extra performance for free. Charging for the unlock capability (or charging you the full market value for the additional cores) presumably offsets that cost.

Re:Ridiculous...

[thegarbz]

Shafted is a matter of economics. If the processor costs the same every other 4 core processor then you're getting shafted. If it's even half the price and you have no need to spend the extra money on 3 cores that would otherwise be sitting idle ANYWAY then it makes a heck of a lot of sense. I have a 4 core system and it spends most of it's time wasting 3 cores I paid for.

For your car analogy. If I had a Ferrari I probably wouldn't take it on a highway cruise. Our local speed way has a pathetically short straight and is otherwise a big mess of twists and turns. So yes a 5th gear that I never use would be something I would be happy to not have to pay for ... providing there's a cost saving involved.

Re:Ridiculous...

[RightSaidFred99]

God save me, why am I arguing with you people. You have no idea how business works, much less how semiconductor manufacturing works.

First, both versions may very well cost $50 to manufacture.

Second (sigh, I'm explaining this to an adult?) businesses do not sell things based on the price of manufacture. They sell based on the price they can get for the product.

You don't know anything about business or business ethics.

Re:Ridiculous...

[thegarbz]

No. This is pure money grab. You buy a CPU that cost the same to manufacture as the higher end CPU, but is crippled until you pay more money.

No actually this is just the way CPUs are sold and have been for the last 20 years. Not all low speed CPUs are high speed ones which fail quality control. This is exactly why some chips can be overclocked to an insane amount yet others with identical stepping and design fail miserably when overclocked even slightly. You don't seriously think Intel run some 30 different production lines all at the exact right amount to meet the retail market demands do you?

The only thing new here is that Intel are giving customers the option of upgrading without requiring physical new hardware. Providing the price is right this could be a great change in the economics of processor distribution.

Re:Ridiculous...

[Shyfer]

No, selling low-end products should not be illegal, selling the same product that has the same cost to manufacture as high-end or low-end and charging different prices for it should. Yes yes, it was advertised on the box and etc, but we shoudnt pay more just because a company wants more money...

"I thought we lived in a capitalist society, not some kind of totalitarian Stockholm state". I'm a communist, do you have any problems with that?

I'm not talking about control in this sense, I'm talking about control as you knowing exactly how it works, waht you can do with it, etc.

Re:Ridiculous...

[Kitkoan]

But you didn't pay for all of it. You got extra cores without paying for them. So you won't be able to use them until you pay another $50.

It's cheaper for Intel because they have to produce only one line of CPUs instead of 2 or 4.

I paid the full/listed price for the CPU and everything built onto it and take it home with their blessing. That means yes, I did pay for it. I shouldn't have to pay more on top of the listed price for the CPU.

Re:Ridiculous...

[Pentium100]

Intel already does this. They already send you CPUs with features disabled when you buy the lower end model.

I was under the impression that the lower end models were in some ways defective, that is - maybe part of the cache does not work, maybe the chip does not work right at high frequencies, in which case they could throw it out, or disable access to the non-working parts and sell the chip for cheaper.

I remember some years ago it was common to unlock additional pipelines in a video card. The manufacturer would test some chips and if they had some defective pipelines, the whole platter would be sold with the pipelines disabled. You can try to unlock the pipelines but if your chip has some defective ones, you will have problems. If the chip was fully working, you'll get a better card for cheaper. Same thing about overclocking. If some chips cannot do the higher frequency the platter is set at a lower frequency, but some chips can actually work at the higher frequency or even higher than the max stock frequency.

However, with this, Intel will have to sell the chips that have been tested to be fully working, otherwise the $50 software won't work on them.

Actually, I think this is more like... your car came with a satellite GPS system, but you need an activation code to use it, or you need to buy software to use it.

OK, so I paid for the hardware, now I need to pay for the software, or find a way to install my own software. Same thing as a PC without an OS.

Or you need to pay an annual fee for the ability to use the radio receiver to pick up traffic data that is broadcast over certain frequencies.

As this is an actual service provided, as, say, paid TV or internet connection, I have no problems with it.

GPS manufacturers sell annual updates to map data, by the way, and if you happen to have 2 or 3 units of the same model number, you still have to buy the update 2 or 3 times.

Just like new versions of Windows. This is a service that is provided - that is, it costs money for them to create the update, just like it costs for Microsoft to create a new version of Windows.

What I object to is buying a three head tape deck, with wires to the record head cut off and having to pay for the permission to solder them back on. The head is there, the wires are there, the electronics are there but I still need to pay for the permission to use what already was in the tape deck from the time I got it from the store. It's not like they would save some money if I chose not to pay for the permission.

All this activation is like saying "Well, I make the more expensive chip cheaper than I ask for the cheaper chip, since both of them are the same, just with artificial locks put on".

Re:Ridiculous...

[Shyfer]

If you dont want to argue just dont and stop saying this like you just said.

If you know so much about semiconductor manufacturing ok, I dont. For me the 1GHz processor is cheaper to manufacture than the 3Ghz processor.

And yes, I do know that businesses do not sell things based on the price of manufacture and I do understand enough of how business work. I said that I disagree with how economy works currently, have you read my comment?

Re:Ridiculous...

[dbIII]

By the time you want to upgrade the socket has changed and it's DDR2 or DDR3 instead of what the old one had.

Re:Ridiculous...

[Khyber]

"Last I checked, CPU upgrades don't necessitate RAM and motherboard changes."

You didn't check hard enough, then. Suddenly upgrading the CPU from one core to 3 or 4 removes the bottleneck from the CPU and puts it elsewhere - likely the RAM (we're not including harddrives for the sake of the argument,) or it's entirely possible the motherboard can only handle 2 of those 3 or 4 cores.

All kinds of things can go wrong.

Re:Ridiculous...

[turkeyfish]

Looks as if the Intel marketing folks and bean counters are crawling all over slashdot. If this flies its just a short jump to hardware licenses and rentals, with chips scheduled for deactivation if additional recurrent fees are not payed every 6 months.

Re:Ridiculous...

[Culture20]

the higher end CPU also costs the same to manufacture

The higher end CPU actually costs fractionally less to manufacture; they are all created unlocked (because they have to be fully tested unlocked), but then the lower-end CPUs have an extra assembly step of getting locked down. And the company has to spend the money on the whole unlock sales-force.

Re:Ridiculous...

[Khyber]

"You have no idea how business works, much less how semiconductor manufacturing works."

Hi, I'm actually a pro in both areas you just mentioned.

And you're full of shit.

"Scenario one, you buy a 3GHz processor that has 512K cache, and it's not upgradeable. You pay $99.

Scenario two, you buy a 3GHz processor that has 512K cache for $99. You can pay $50 to upgrade it to 3.2GHz and 1M cache.

I defy you to logically explain how scenario one is better."

Scenario one doesn't involve MAFIA TACTICS and breaking the law, nor does it involve a 50% markup for MOST CERTAINLY NOT 50% performance increase.

Which is why my company threw their support 100% behind AMD when we heard of this.

Now sit down and shut up, real professionals are having a discussion.

Re:Ridiculous...

[mrnobo1024]

Anyone who's taken basic microeconomics knows that economic efficiency is maximized when price = marginal cost. If two goods are manufactured the same way with the same machines, their marginal costs are equal, so if there's a price difference there must be at least one inefficiency there. It's not just that the consumer is getting screwed, there is less wealth in total to go around then there should be. The proverbial pie is being shrunk.

Re:Ridiculous...

[hydromike2]

This is especially a better option for not having to test new hardware in critical systems and simply 'pressing a button' to make it faster.

Re:Ridiculous...

[Shyfer]

Well, IMO it should, anyone trying to sell anything above its real value is wrong, is damaging the whole society and should be illegal, but that is a matter of opinion really.

Well, I do. Being a communist means that you're either stupid or evil. Either you haven't paid attention to the past 100 years of human history or you have and actually want to bring about the kind of things that communism encourages.

OMG, hope you are kidding... Guess that the cold war and all the USA propagana against communism made you think like this.

Re:Ridiculous...

[TheVelvetFlamebait]

I read the first two sentences of your post and saw instantly that you don't really get it. Everything a company does is technically a pure money grab. The term "money grab" is just a purely arbitrary and subjective label that people put on a company's actions when they don't approve of those actions. Whether or not an action by a company is deemed a money grab by a given consumer is (or should be) irrelevant when you go shopping. What should be relevant is the product you're getting versus the amount you're paying.

Also, even though manufacturing may cost the same between the locked and unlocked CPU, Intel may well be unlocking a previously untapped market, for people who want a CPU, but are unsure of the amount of power they need. This provides something close to what they want: a CPU that can be upgraded for next to nothing, with next to no hassle involved. If such a market were to exist in any significant size, then releasing the locked CPU could provide intel with a greater market share, and greater profits. Greater profits then allow intel to eat into their newer, fatter margins to lower their prices when competing with AMD. I'm not saying it will happen, but it could happen. Don't be so quick to dismiss more creative business models.

Re:Ridiculous...

[mlts]

There is a lot a company (and I'm using this in the generic sense) would gain by having a completely closed and locked architecture that is geek hostile. If a company can make a product that is extremely hack-resistant and forces consumers to always pay for incremental stuff (DLC, ability to play used games, monthly fees for single player games, monthly media access fees), they are going to be rich and have a guaranteed income source on the books each month.

MBA 101 -- squeeze customers. It will take a lot of squeezing before people show a company the middle finger and stop using a product they just bought. Especially if the company had a product that was the only game in town.

Re:Ridiculous...

[Ekdar]

You did pay for the extra cores, but you paid for "locked" cores. I'm not sure I'm too fond of the whole locked-cores notion, but I don't necessarily see anything intrinsically wrong with it.

I mean, I don't think they're going to mislead you into believing that you're paying for 4 usable cores only to reveal at boot time that 3 of the 4 are locked. They're going to advertise to you that you're getting 1 functioning core and you'd pay for the hardware with the understanding that you were getting a single functional core (and not more).

It sounds like, for you, there's just something about locked functionality you don't like, even if the price you paid was a fair one based upon the non-locked functionality. But, for most average computer users this is probably a non-issue. It might even be helpful for someone who doesn't have the money now to pay for a feature they may want down the road.

Re:Ridiculous...

[macshit]

"Last I checked, CPU upgrades don't necessitate RAM and motherboard changes."

You didn't check hard enough, then. Suddenly upgrading the CPU from one core to 3 or 4 removes the bottleneck from the CPU and puts it elsewhere - likely the RAM (we're not including harddrives for the sake of the argument,) or it's entirely possible the motherboard can only handle 2 of those 3 or 4 cores.

All kinds of things can go wrong.

On the other hand, it's certainly not necessary that "things go wrong", and there are cases where a new CPU works out -- it may very well be that the system pre-new-CPU actually had excess memory bandwidth, or your application happens to be very cache friendly, or the new CPU's bigger cache is enough to handle it.

Re:Ridiculous...

[number17]

It will cost more to purchase this chip though. Not because of production costs but because of the implementation of this program. They will not have to hire people to either answer phones to give you a code or hire programmers to create a web app. It adds little value for more cost.

Re:Ridiculous...

[emt377]

I was under the impression that the lower end models were in some ways defective, that is - maybe part of the cache does not work, maybe the chip does not work right at high frequencies, in which case they could throw it out, or disable access to the non-working parts and sell the chip for cheaper.

It's certainly possible that an i7 with a defect can be sold as an i5, but that doesn't mean all i5's are defective i7's, or that a Core 2 Duo is a crippled i7, or that a defective i7 can be sold as a mobile i7. Nor are Atoms defective i3's; Atoms are fabbed by TSMC so even if this were even electrically possible the logistic complexity would soon consume any benefit.

Re:Ridiculous...

[Lord Kano]

Well, IMO it should, anyone trying to sell anything above its real value is wrong, is damaging the whole society and should be illegal, but that is a matter of opinion really.

How do we determine value? An item's value is determined by what people are willing to pay to get it. I'm into retro-computing. Look on Ebay. On one hand, a 15 year old Sparc might be worth $10, but if there are enough people who want that specific configuration it would be worth $200.

OMG, hope you are kidding...

I'm not.

Guess that the cold war and all the USA propagana against communism made you think like this.

Well, that and the ~40 million people that were exterminated by Communist regimes during the 20th century.

LK

Re:Ridiculous...

[Shyfer]

I'm talking about companies selling stuff to customers. For example, if an item costs $100 to manufacture (include researching, etc) and they sell it for $1000 it is wrong, even if lots of people want it, if it costs $100 people should pay $100 and no more. Paying more than it costs is only giving money to the company owners, and IMO that is just wrong.

Well, that and the ~40 million people that were exterminated by Communist regimes during the 20th century.

I'm not talking about specific regimes, I'm talking about the communist theory. Also, how many people were exterminated by capitalist regimes during wars that only beneficiated the small elite?

Re:Ridiculous...

[Lord Kano]

I'm talking about companies selling stuff to customers.

Makes no difference. Economics is still economics.

For example, if an item costs $100 to manufacture (include researching, etc) and they sell it for $1000 it is wrong, even if lots of people want it, if it costs $100 people should pay $100 and no more.

If no one could sell anything for more than the cost to manufacture it, what would be the incentive to make it in the first place?

Paying more than it costs is only giving money to the company owners, and IMO that is just wrong.

Well yes. That's what companies exist for. Companies exist to provide profit to their shareholders. If there were no profit, there would be no companies. Now, before you start harping on what a good thing that would be, let me remind you of a few things. EVERY pharmaceutical breakthrough since my grandparents were born has been expensive, too expensive for any single person to afford to pay for it. Companies that are large enough to pay for medical research or build new fuel efficient cars or tasty beverages are why we have all of those things.

I'm not talking about specific regimes, I'm talking about the communist theory.

We don't live in a theoretical world. We live in a world where all of these theories are put into practice by imperfect human beings.

Also, how many people were exterminated by capitalist regimes during wars that only beneficiated the small elite?

One could argue that Capitalism leads to hunger and poverty in the third world and that poverty leads to untold deaths, and that is an arguable position. However, Capitalists have engaged in far less of the "lining people up and machine gunning them into a mass grave" kind extermination that the Communists.

LK

Lock in the BIOS or in the CPU?

[crow]

It would be relatively simple for the BIOS to turn off CPU features in such a way that they can't be turned back on without a reset. So the easy way to implement this would be for Intel to partner with a PC vendor and charge for the BIOS upgrade that doesn't disable the CPU features in question. With such a system, it would mean that you could pull the CPU and put it in a different motherboard, and get all the features, but that's not going to be a concern for the business model until they're talking about hundreds of dollars for the added features.

Putting this into the CPU would require that the CPUs be designed specifically to support this, which is not as likely to be the case, but would be much more difficult to defeat.

Re:Lock in the BIOS or in the CPU?

[BitZtream]

which is not as likely to be the case

Well, if it isn't the case than Intel will look like some serious schmucks when it gets cracked. Considering Intel also produces tamper-proof chips for cryptography stuffs for the goverment, I would presume that putting it into the CPU would be trivial for them ... its what they do.

I'm not saying they wouldn't do it in CPU, but lets face it, they know if they don't it'll be cracked, and they know how to do it in CPU already, and it would look bad on their other divisions if it does get cracked ... I just cant' see them being stupid enough to do it any other way.

Similar experience at bestbuy

[js3]

So I went to buy a laptop from bestbuy and this dude from the geeksquad told me they could make my cpu go 18% faster. I was baffled, I asked him, "so you're saying i'm buying a cpu that's 18% slower than what it says on the specs?". After giving him sarcastic replies for like 3mins, he finally told me they didn't have that system in stock, lol. The asshole probably wanted to sell that shit to some unsuspecting mom. Fuck the geeksquad.

Re:Similar experience at bestbuy

[Mitsoid]

Most of those services remove the default useless bloatware from your new PC for you -- which means, yes, it can go faster.

However, most of us that read sites like these already know to clean slate their new PC anyway...

And just a note... that bloatware is not installed by Retailers -- it is installed by the PC manufacturer, so in some cases people do appreciate these services as they do not know how to remove them themselves.... Not any of us though, since we know how to deal with the bloat...

I do not believe their claims though, but I do agree you can get *some* improvement by removing them

Re:Similar experience at bestbuy

[Kitkoan]

Fifteen enormous cocks raping every orifice in your body.

I couldn't describe hiring the geeksquad any better then that.

Re:Similar experience at bestbuy

[GrumblyStuff]

I could only think of places for thirteen and the uncomfortableness (quite the understatement) would be staggering. I am intrigued yet frightened.

Re:Similar experience at bestbuy

[mysidia]

I would ask them how they are going to prove they improved the performance by exactly 18% and not 15% or less? :)

Re:Similar experience at bestbuy

[Osty]

Most of those services remove the default useless bloatware from your new PC for you -- which means, yes, it can go faster.

Many, yes. But not Geek Squad. Their concept of "removal" is to delete the program icon from the desktop without actually uninstalling it. Unless the software is something that competes with their own offerings, so that they can upsell you a piece of crap antivirus solution after removing the norton or mcafee trial that came with the computer (never mind that you're better off installing something free like Microsoft Security Essentials). There's also the problem that as soon as they take your new computer into the back room where they do the work, you have no idea what you'll get back. Will it be missing RAM? Will it have the right power adapter (or any power adapter)? Will you get the right Windows license paperwork and other bits and pieces that came in the box? Honestly, Best Buy should give you a discount if you use their Geek Squad service, since at that point you're essentially buying an open box computer of unknown completeness.

And just a note... that bloatware is not installed by Retailers -- it is installed by the PC manufacturer

Well, yes and no. Yes, the bloatware is installed by the manufacturer, but if the retailers really didn't want it they could exert pressure on the manufacturers to sell clean PCs. Microsoft does this in their two retail stores (and online) -- you can buy the exact same HP or Gateway or whatever as you'd get from Best Buy, but without any of the bloatware crap. And without the Microsoft store folks having to do any of the cleanup. They get the machines that way from the manufacturers. Best Buy, etc, actually prefer PCs with bloatware because it means they get to upsell their customers. Thus while it's correct to blame the manufacturers for putting this crap on their PCs in the first place, the retailers should receive just as much blame.

Re:Similar experience at bestbuy

[bloosheep]

Fifteen enormous cocks raping every orifice in your body.

I couldn't describe hiring the geeksquad any better then that.

Please note that "enormous cocks" is what the Geeksquad is, not what they have.

Re:Similar experience at bestbuy

[Nimey]

The retailers don't want crap-free computers because:

1) The price of each machine would go up a few tens of dollars, because the crapware purveyors pay the manufacturers to include it. Unless all the other retailers did it too, it'd be a death sentence.

2) They couldn't offer very profitable services to remove the crapware.

Cue:

[fyngyrz]

(to the tune of the Intel commercial):

Bum-bum-bum-bum!

Re:Cue:

[Nerdfest]

More accurately, "dumb, dumb, dumb". What's next, using the IBM mainframe approach of charging for how many MIPs you use?

Re:Cue:

[fyngyrz]

Yes, that was the point. Whoosh much?

Good Idea

[ifiwereasculptor]

Probably cheaper to produce and cheaper to buy, at least for the end user who knows his way around the interwebs. So everybody wins.

not new

[Lord Ender]

IBM has been doing this in with mainframes for a while. As long as you sell these to businesses with lawyers who will flip out if they hear of IT breaking contracts, Intel should be fine, too.

Re:not new

[mlts]

Mainframes != PCs.

When a company buys a mainframe, they really don't care how many CPUs are sitting in the CEC, versus ones paid for. The reason enterprises buy mainframes is because big iron is made to be working with 100% uptime, and if a part failed, some redundant component takes over, IBM is servicing the down part, and the end users are not noticing a single bit of downtime. Companies buy mainframes for the service agreements and the reliability.

With a mainframe, a company expects to get the box installed, carve out LPARs, install applications and DB servers, and have a good assurance that 3-5 years later with a 24/7/365 business, the amount of unscheduled downtime they have had to deal with is numbered in seconds (barring network issues, obviously.)

PCs are different. x86 hardware is intended for commodity use where reliability is not really as much a concern as price. PC hardware is decently reliable for the most part, but any IT departments expecting 3-5 nine reliability from x86 stuff without multi-machine clustering would be just plain delusional. With PC hardware, what is under the hood is the important thing people are buying.

Remind anyone of IBM?

[StuartHankins]

This model works fine for large business servers where downtime is expensive -- unlock the resources when you grow -- but the audacity of doing this on the x86 platform is fail.

Intel Fanboys

[r6_jason]

I wonder if the Intel Fanboys have udders, because Intel sure is trying to milk them.

Re:Intel Fanboys

[MobileTatsu-NJG]

I wonder if the Intel Fanboys have udders, because Intel sure is trying to milk them.

Heh yeah because AMD is sooOOoo altruistic.

Re:Intel Fanboys

[tokul]

I wonder if the Intel Fanboys have udders, because Intel sure is trying to milk them.

Or Intel has hard time in selling high end CPUs to Intel users. So they sell those CPUs at lower price and hope that some customers will pay extra some day when hardware upgrade requires more resources than price for flipping some CPU microcode switch.

Re:Intel Fanboys

[hedwards]

An Intel fanboy in this age? How udderly preposterous. Do they sit around playing games with Intel IGPs?

Hey, I don't mind..

[xemc]

As long as I get a $50 break on a new CPU!

Seriously, if you get less you should pay for less. They'd still be competing with their other chips (and AMD's).. so it's not like you shouldn't get what you pay for.
That being said, this is like Intel creating a similar avenue as overclocking.. getting more performance from a cheaper chip. I'd really be tempted to get a crippled / cheaper CPU and just crack it to get the full-price speed.

Re:Hey, I don't mind..

[sdnoob]

As long as I get a $50 break on a new CPU!

You already are... by buying the Pentium instead of the more expensive i3 that already has the extra MB of L3 and HT enabled.

_____

Intel and AMD have both been shipping chips with certain features disabled to meet market demands for years. Nvidia and ATI do the same with GPUs. Sometimes the disabled parts are actually defective, but sometimes not. Then you have two chips that cost the exact same to manufacture sell at two different price points, with the manufacturer intentionally choosing to sell some at a lower price (with the plan of making up the difference through higher sales).

Owners of certain AMD processors have been able to unlock entire cores along with extra cache for some time now. Intel is just trying to profit from it. I just don't know how well that idea will go over with the uninformed masses. I think many will be just a bit pissed-off that they were sold an intentionally-crippled computer. Unfortunately, any backlash will be aimed at the company who's logo is on the box, not Intel.

Re:Hey, I don't mind..

[Kitkoan]

Doubt they would give you a $50 break on the new CPU, I don't think the markup is that high and they aren't going to sell the CPU's at a loss hoping everyone will pay for the upgrade (because I doubt most would bother since most people only use their computer to surf the web, listen to mp3's and play something like FarmVille, not the most CPU intensive). And they won't bump up the markup because that would give companies like AMD one hell of a competitive advantage by AMD just cutting their markup to below $50.

Re:Hey, I don't mind..

[RyuuzakiTetsuya]

I kind of agree but this is seriously a dick move.

At least in older times they took the effort of cutting the traces or removing the units all together ala 486SX/DX chips.

Re:Hey, I don't mind..

[Nikker]

That's a slippery slope my friend, are you sure you want to start this garbage of a game with Intel's marketing dept? "Sure we'll give you $50 off of our product, we were going to sell you a sigle core 1.4GHz for $250 but now we will drop the price to $150 clocked @ 1.2GHz, that's a better deal than you were going to get to begin with!". Now realize the numbers involved were exaggerated but I hope you get the jist. Once you start bartering over what you were supposed to be getting in the first place you end up with nothing.

Re:Hey, I don't mind..

[sayfawa]

But this is backwards. In the sane world, the reason why paying less gets you less is because the cheaper product was cheaper to make. Not only is their locked down product not cheaper to make, but it's actually making the whole line more expensive. Some of our dollars are actually going towards the developers for the purpose of making the product worse, by locking it down.

Re:Hey, I don't mind..

[mysidia]

Unfortunately, any backlash will be aimed at the company who's logo is on the box, not Intel.

You seriously think Intel's logo won't be on the box?

I see Intel Inside Logos almost everywhere. Usually they are happy to have their name everywhere, and PC manufacturers stamp the logo prominently, probably to take advantage of the capital in the Intel brand and customer familiarity with Intel.

Re:Hey, I don't mind..

[mysidia]

It's not the case of "you get a $50 break". It's a case of: they would sell this CPU to you for the same price anyways, even if they weren't experimenting with an upgrade feature.

If they weren't experimenting with this upgrade concept, it's not like your CPU would be more highly specced otherwise. It would just be still locked without any opportunity to upgrade it.

Just like most low-end Intel CPUs are. Artificially locked; loaded with code that disables certain features.

Re:Hey, I don't mind..

[sdnoob]

You seriously think Intel's logo won't be on the box?

Consumers don't buy their computers *from* Intel, they buy from Acer, Dell, Gateway, HP. And many of them couldn't tell you who made their processor (or what a processor is, for that matter) even with that little 1x1 case badge.

Re:Hey, I don't mind..

[Kitkoan]

Yes, but with many low-end Intel CPUs, if your willing to "get your hands dirty" you can unlock them (most people won't, but those who want/can might just do it). With this, that option/power has been removed from you and been given a price tag instead.

Re:Hey, I don't mind..

[sjames]

You already are... by buying the Pentium instead of the more expensive i3 that already has the extra MB of L3 and HT enabled.

What it actually means is that the market is sufficiently unhealthy that Intel can get away with charging you $50 more than the free market value of the i3. Or, at least Intel believes that it is. We can hope AMD eats their lunch and proves otherwise.

Re:Hey, I don't mind..

[kenh]

"Intel is just trying to profit from it. I just don't know how well that idea will go over with the uninformed masses."

I think you got that backwards - the uninformed will be happy the laptop was sold at a discount, and they will also be happy they can upgrade (if they want) for a relatively trivial amount of money at a later date (helping them avoid anticipated buyer's remorse).

Oddly, I think it is the informed masses (or should I say "those who mistakenly believe themselves to be informed") that are having a problem with it - I think they are upset that anyone with $50 can goose more performance out of a low-end CPU - prior to this being made available by Intel, overclocking was considered a black art by many. Now, and Tom, Dick, or Mary can go down to Best Buy, drop a $50 bill on the counter, and in a few hours have a faster laptop without picking up a screwdriver or trying to manipilate BIOS settings.

That's OS licensing...

[Qubit]

IBM's been doing that sort of thing for years. They ship you a mainframe with more processors than you ordered or a disk array with more disk than you ordered, and you can pay them to turn it on.

Yes, but I'm pretty sure that's all predicated on IBM service contracts and/or the license on the IBM OS/application software running on the system.

If you're running a completely-FOSS debian install on top of these new Intel processors, what leverage do they have on you?

Re:That's OS licensing...

[shutdown -p now]

I guess they'll just put EULA into BIOS, and will flash "Press whatever to read EULA" alongside everything else on boot.

Re:That's OS licensing...

[derGoldstein]

That's the most likely solution, but would that be enforceable? Will Intel really claim: "it was clearly stated in the EULA, in the BIOS, which flashes for 2 seconds if the vendor screen isn't turned on by default". You can't hide the EULA in an encrypted microscopic engraving somewhere, and you can't have a EULA that's activated by "if you wait 5 seconds without pressing DEL, then you are accepting the EULA" (at least I think you can't -- it's supposed to require some active, conscious effort on your part).

Re:That's OS licensing...

[shutdown -p now]

I think it would be fairly trivial to e.g. make it pop up the first time you boot the PC after purchase, and boot not proceeding unless you choose "I agree".

Anyway, don't you worry. That's what expensive lawyers are for. They'll figure something out.

Re:That's OS licensing...

[Eivind]

That would not work.

The agreement to purchase was made at the time of purchase - additional conditions disclosed AFTER the sale, can not have any influence on the sale itself.

If I walk into a store and say: "I would like to buy this computer", the clerk says: "Very well sir, that'll be $899 then." i pay, and receive the computer. Then what just happened was a very classical, fairly typical sale. (except clerks tend to be less polite)

They cannot then hours or days later, when I turn on the computer that I bought, pop up some text that says: "Infact you didn't buy this computer, instead you need to agree to this /agreement/ which forbids you from doing certain things with it."

Similarily, if you paid by check, and when the store tried to cash it in, they got a EULA from you, wherein they need to agree to certain conditions of yours, to get the cash, that's -also- not be legally binding.

Clauses which are not negotiated and included in an agreement BEFORE the agreement is entered into, can obviously not form part of that agreement.

Re:That's OS licensing...

[icebraining]

You do know that's how software EULAs work right now? With most pieces of software, you only get the EULA after buying, yet ProCD vs Zeidenberg found this kind of shrinkwrap licenses valid if the software shows you the license and you have to click "I agree" before using it.

Re:That's OS licensing...

[Lehk228]

FTFA At the time of purchase, Zeidenberg may not have been aware of any prohibited use, however the package itself stated that there was a license enclosed.

Re:That's OS licensing...

[BitZtream]

Heh, they don't need any more leverage. If you're trying to run Linux on these machines you're going to go out of business real quick.

We're not talking about PCs running a normal OS, we're talking about main frames ... which run slightly more advanced operating systems.

Re:That's OS licensing...

[icebraining]

Said there was a license, but he could only access the conditions after the sale. Parent said:

Clauses which are not negotiated and included in an agreement BEFORE the agreement is entered into, can obviously not form part of that agreement.

Which is clearly not true for software EULAs, since you can't even see the clauses, let alone negotiating them.

Re:That's OS licensing...

[Tim C]

If you're running a completely-FOSS debian install on top of these new Intel processors, what leverage do they have on you?

Leverage for what? If you're running an OS that they don't port their "make CPU go fast now" software to, you simply get what you paid for - a CPU with X cores running at Y GHz, and they don't get to charge you $Z for the upgrade.

Maybe, just maybe, if you crack it they could smack you with the DMCA (depending on the implementation of course), but how would they know you'd done it?

Don't let the marketing get to you...

[Anachragnome]

Don't let the marketing get to you, and do not encourage it.

If you are shopping for processors, simply disregard the "upgrades" and treat the product accordingly. Does it compare with fully unlocked competitors?

No? Then don't buy it. Yes? Then buy it but don't upgrade.

Re:Don't let the marketing get to you...

[RightSaidFred99]

Exactly. Most of the replies in this story are just nonsensical. I really don't understand how some people see the world.

Consider this an option, and one you can ignore. If you can buy a better processor for the same price and don't care about upgrading, buy that one. Otherwise, this is just an option that you can choose to use or not use later on.

Re:Don't let the marketing get to you...

[_Sprocket_]

I really don't understand how some people see the world.

Some folks get a bit punchy when they feel like they're being manipulated. And this is manipulation.

This is a method for Intel to simplify logistics for themselves and their partners without abandoning their long-standing business of carefully guiding their technology through price-point gates. If anything, it might even make it easier to set up more gates as a single unit could potentially cover many price points designated by licensing tiers. That's great for Intel. It's probably just as good for OEMs and retailers as this is a chance for them to "up-sell" every unit. They can get their foot in the door with a bargain price and then either push for the "better deal" right there or collect on the "modest upgrade fee" later on.

However, as consumers, it'd be better for us for them to run in to logistical walls where they simply have to move their price points up to the next better-performing product.

Re:Don't let the marketing get to you...

[RightSaidFred99]

Exactly, it's better for Intel, OEMs, and retailers. And those efficiencies bear out in the price consumers pay for the product. Now they can sell the consumer a less expensive product but also allow the same product to be upgraded to what would normally be a more expensive unit in software. Normal consumers do not upgrade CPUs. Now then can easily.

The problem people have with this appears to be a lack of understanding of how businesses price products, especially the semiconductor industry.

Like the PSN fiasco, which was also due to consumer ignorance, this one could blow up in Intel's face. In the case of the PSN, even technical people turned off their brains. It was as if there weren't any of a hundred ways for someone to track your computer already, they thought the PSN was some privacy issue. It was ridiculous, and any backlash about this is equally ridiculous.

Re:Don't let the marketing get to you...

[_Sprocket_]

Well, yes. If you want to ignore that this creates a price point scale that's favorable to Intel and not consumers, then sure. It's good for consumers too.

I like how you hand-wave off the PSN issue also. Very well done.

Capitalism sucks

[log0n]

I'm getting so fucking tired of companies creating ways to further nickel and dime.

There's no chance this coupon is going to bring down the price of a computer by $50 to correspond to the loss of features, this is just another way to make some coin after the fact.

We never end up saving money, it's all bullshit.

Re:Capitalism sucks

[overtly_demure]

But Comrade, it's really great bullshit!

Re:Capitalism sucks

[thegarbz]

Depends. If they don't price it right then you can just go to the competitor. On the other hand if the chip is now $50 cheaper then I'll be lining up to get one providing what they have locked is a feature I don't need. (Bonus points if hackers figure out a way to make it work anyway).

It may seem like nickel and dimeing but really processors have sold this way for years. All they've done now is monetize overclocking of chips which actually pass quality control but needed to be sold as a lesser product to meet market demands (quite a common practice).

Re:Capitalism sucks

[MobileTatsu-NJG]

There's no chance this coupon is going to bring down the price of a computer by $50 to correspond to the loss of features....

Why, did AMD stop making CPUs?

Re:Capitalism sucks

[TheVelvetFlamebait]

If you buy a crippled CPU, you are not losing features, you are gaining them. You're just not gaining the amount that you would if you had paid the extra $50.

Intel could easily have made the same crippled CPU without the ability to upgrade, at (roughly) the same price. Does that make it a superior CPU? Does it make it any less crippled? It has the same features at the same price. Well, it would have one less feature: the ability to be upgraded to a slightly faster model for peanuts, which is a pretty useful feature, if you think about it.

Re:Capitalism sucks

[soupforare]

We never end up saving money, it's all bullshit.

We've been saving money. You can get a non-value tier quad core for under a hundred now, that's ridiculous. I paid well over that for every other CPU I've had. Even prebuilt machines are cheap, it's a wonderful time. This intel thing is for computer illiterate people to get an easy performance boost postsale. The only people who should be bitching are sales guys who will be missing out on spiffs for selling slightly higher-end machines to grandma.

Re:Capitalism sucks

[HyperQuantum]

There's no chance this coupon is going to bring down the price of a computer by $50 to correspond to the loss of features....

Why, did AMD stop making CPUs?

What's to stop AMD from doing the same thing? (though I hope they don't)

There are only two vendors to choose from if you want an x86 or x86-64 CPU. Not much choice for the consumer IMO. They could do almost anything they want and still not lose that many customers.

MS Windows?

[overtly_demure]

Isn't this how MS Windows works since a couple of versions ago?

Re:MS Windows?

[0123456]

Isn't this how Windows works now? There's a window somewhere which says that for $80 I can enable the 'Pro' features of Windows 7 that are already installed on my PC.

Re:MS Windows?

[6ULDV8]

Yes. Microsoft Windows Anytime Upgrade (WAU)
http://en.wikipedia.org/wiki/Windows_Anytime_Upgrade

The hardware aspect is used in quite a few products. MGE's Galaxy PW UPS in the 100 - 225 kVA range are identical, except for the license key used during startup. "field upgradeable up to 225kVA without the expense of added equipment or installation costs" http://www.keyitec.com/keyitec-MGE-UPS.html#GalaxyPW

Could be a huge marketing tool

[mikeabbott420]

I think the possibility of a crack is irrelevant.

Maybe this is just an extension of the normal binning any semiconductor process produces and a demonstration that at this particular tick/tock point the yields are excellent.
More likely though is that CPU performance has become so complicated to predict, and so often irrelevant to many classes of users, that this is intels best guess at how to upsell CPU's to users who have no idea how different CPUs perform.

Some people pay for good reason, Some people buy intel instead of AMD because a crack means they can "get something for nothing" but mostly imagine all the people getting frustrated with some poorly written piece of software and thinking "50$ .. worth a try" .

It's hard out there for a pimp, when pimping cpu's to farmville players.

Re:Could be a huge marketing tool

[Corporate Troll]

It's hard out there for a pimp, when pimping cpu's to farmville players.

Never played FarmVille, eh? It's unbearable on an AMD Turion X2 (Machine is from 2007, runs Linux) and just okay on a Core i7 iMac. Perhaps it's better on the Windows platform, but FarmVille (or Flash, or both) is a pure CPU hog.

Re:benefit from embracing customers

[Kitkoan]

Wouldn't Intel benefit from embracing it's customers and just allowing this to be done?

Maybe, but embracing this would benefit Intel's stockholders even more.

Re:benefit from embracing customers

[kestasjk]

Think of this service as customers who need those full capabilities subsidizing the customers who don't need them.

Whether Intel produces chips with or without these extra features it costs them the same, though it costs them more to design the features initially. This means that giving the features to everyone has to raise the price for everyone, even those who don't need the extra features.

I know making money is evil, but let's try to take a balanced view of this.

What's next, Windows only CPUs?

[Da w00t]

How come the software to "unlock" this capability appears to be windows only?

Re:What's next, Windows only CPUs?

[ILuvRamen]

Probably because it's a software lock and the CPU is unlocked without it lol. Btw I'm sort of new to programming but can't someone just flip the boolean bit in memory with some sort of memory editor (I hear they exist) for the software that represents the variable responsible for validating the keys? I mean don't most programs just do their fancy encryption keys and then flip a boolean switch or at least use a temp one inside an if statement? So flip that and the thing thinks it's validated. No need for a key generator.

Re:What's next, Windows only CPUs?

[Kitkoan]

For the same reason BIOS and other firmware upgrades are typically Windows only.

Re:What's next, Windows only CPUs?

[Mashiki]

I don't know of any motherboard that doesn't support some form of Qflash and loading the bin directly from a flash drive, that's been made in the last 4 years. That includes CPU and HDD firmware updates.

Re:What's next, Windows only CPUs?

[tomz16]

EXACTLY! I'm new at this stuff too, and I think you are on to something. You just have to boolean the bit that controls the memory hash function pointer. Then you can probably just decrypt the parity bit endian stack!

Re:What's next, Windows only CPUs?

[ILuvRamen]

gee, thanks for the helpful and completely non-sarcastic reply. If you were a little bit smarter, you'd realize that at some point a program effectively asks itself what the result of the decryption was and it relies on memory to do that. Was the key entered valid, yes or no, is a single bit held somewhere in memory and it probably loads between more obvious variables in a detectable pattern. Aren't there hex editors or something like that out there to scan memory and find it and then flip it? This type of attack is exactly why all MMORPGs have the server generate all random numbers for decisions at the server and not the user's PC because memory editing to affect outcomes is very possible.

Re:What's next, Windows only CPUs?

[Laxori666]

There's a reason you're modded 1 and the "unhelpful" reply was modded 4. Even if the code is something like: "int keyIsValid = validateKey(); if (keyIsValid) turnItOn();", and "all you have to do" is flip "keyIsValid" to 1, you'd have to: 1) do it at the right second, 2) find where in memory "keyIsValid" is which... is like finding a needle in a haystack. And every time you run the program it's a different haystack with the needle somewhere else.

Re:What's next, Windows only CPUs?

[_DangerousDwarf]

Hi this is Hollywood calling,

Can you we please talk about a potential job for you a script write on the latest Hollywood blockbusters?

Thanks, your technologically challenged script writer.
DD

Re:What's next, Windows only CPUs?

[Anonymous Coward]

Don't you realise this is written by Intel and not some two-bit DRM-ware vendor? They could very well make it so these chips do the validtion on the CPU (therefore no bits in RAM that can be flipped) and make each CPU require a unique key. Doesn't mean it will be uncrackable, but it isn't the same as DRM, so I wouldn't count on it.

Re:What's next, Windows only CPUs?

[Alsee]

That won't work unless you rotate the bit harmonics first.

-

Re:What's next, Windows only CPUs?

[Jesus_666]

I think we may need to write a GUI in Visual Basic first so we can capture the IP address of the L5 cache.

Oh-my-effing-Gawd

[Astronomerguy]

Grrrr....I can feel the nerd rage building...must...punch....pillow...!!! Seriously, this is pure nickle-and-diming of the consumer. Fuck the moron in Marketing that grunted out this steaming pile of idiocy, I prefer a nice black-and-white binning model with related price points to this micro-purchase crap. This is so asinine that I'm sorely tempted to take a serious second look at AMD. Note to Intel: WT-effing-F???

Re:Oh-my-effing-Gawd

[Lehk228]

they already bin perfectly good high end CPU's down to lower end models for marketing purposes, if this system replaces that with an option to upgrade later it's just offering a stopgap later, rather than taking the time and money to replace your CPU drop the $50 and upgrade it and keep using for another year.

Re:Oh-my-effing-Gawd

[Astronomerguy]

they already bin perfectly good high end CPU's down to lower end models for marketing purposes, if this system replaces that with an option to upgrade later it's just offering a stopgap later, rather than taking the time and money to replace your CPU drop the $50 and upgrade it and keep using for another year.

What I like about the current model is that, depending on your motherboard, you can get a significant increase in processor speed for nothing. I bumped my Core 2 Duo 3Ghz to a stable 3.6Ghz on air cooling alone. That took all of 15 minutes to do. 4Ghz was almost stable but the voltages need to be tweaked a bit. Intel's binning processors to a lower speed benefits those consumers who don't mind digging under the hood to get better performance. Locking/crippling functionality and charging for an unlock code to regain functionality is just plain wrong. If this actually hits the marketplace, I'll nuke a bag of popcorn and watch the hackers go to town on creating and releasing a free unlock. Hardware is not licensed like software is (and don't get me started on that...). I bought it, I own it, and I can do whatever the heck that I want to it.

$200 should have bought full functionality then

[mrnobo1024]

They wouldn't have sold the crippled CPU to you if $200 wasn't a fair price for at least the full quad-core CPU, since that's what they had to manufacture. Whether you keep it as single-core, or pay extra for the upgrade, you are with absolute certainty being ripped off.

Re:$200 should have bought full functionality then

[flux]

You know, there are other costs involved than just manufacturing the chip. For example recouping the R&D costs. I would be very surprised if the price difference between dual-core and quad-core CPU's reflected the difference between the manufacturing costs of those two.

Re:$200 should have bought full functionality then

[Twinbee]

You're not just paying for the hardware, but the time and effort that went into researching, designing and testing it.

Re:$200 should have bought full functionality then

[Thruen]

Which has always been, and will still be, included in your initial purchase price. Enough trying to justify Intel's actions by comparing it to what HP or IBM did in the past. I don't know the details surrounding mainframes sold with additional hardware to be activated later, but it could just as easily be as wrong as this is. You're not understanding why this is a problem. The problem here is that in order for Intel to appropriately price the chip so they'll make money, they have to assume not everyone will buy the upgrade. In fact they should assume most people won't buy the upgrade. So if the actual value of the upgrade is $50, and they assume 20% of people will buy the upgrade, they have to recover $40 per chip sold, meaning that $40 will just get tacked onto the initial purchase price. Right there, is where we all get ripped off. If they can afford to let people not upgrade the chip, it means they've included the actual cost in the chip price already, and that's the rip off. If you don't upgrade, then $40 of your chip price goes to cover hardware you're not using, and if you do upgrade, then $40 of the $50 you paid is actually to recover money lost due to other people not upgrading. If you still don't understand the problem, I can't help you.

Re:$200 should have bought full functionality then

[Twinbee]

I get the whole "getting hardware that you're never going to use" thing. But you're still overlooking factors.

Firstly how do you even know that it's much more expensive for the extra hardware? For all we know it's better off overall for them to unify two chip designs into one. It reduces complication and therefore cost.

I'm also not convinced about your idea that they "must assume that no one will want the upgrade". For all we know, they may break even, or gain/lose a little if no one upgrades, and all the profit lies in the upgrade. These things are hardly black and white.

Please use paragraphs in future btw - just makes stuff easier to read without backtracking a line to see where I've read up to.

Re:$200 should have bought full functionality then

[SpectreBlofeld]

Nah. This is really no different than Microsoft's six editions of Windows - Starter, Home Basic, Home Premium, Professional, Enterprise, and Ultimate. It doesn't cost MS any more to stamp out an Ultimate disc than a Starter one, so why not just have the Ultimate edition only?

The reasons are economic. If you only had one version of Windows, what do you charge for it?

I'm not going to bother to look up the actual prices for the editions, so let's make 'em up. Let's say that Home Premium is $50 and Ultimate is $199.

If we take away the tiered pricing and MS sells Ultimate only, they'd probably settle on an average price, let's say $129.

The people that really needed or were willing to pay $199 are now getting a bargain at $129, but Joe User who was previously willing to pay $50 is either shut out of the market or forced to pay more, for features he didn't need.

In chip manufacturing, like any other manufacturing, it's cost-prohibitive to have too many production lines going on at once. You can have tools, materials, and manpower divided into two production lines pumping out low-end chips and high-end chips - that probably cost the same to manufacture once R&D is finished, actually - or you can streamline, and produce one chip. Which is what Intel's doing. They're offering a Intel Home Basic edition that's affordable and an Intel Ultimate Edition that's pricier. Both prices reflect what the market will bear.

This allows Intel to market to people of differing needs and socioeconomic strata. I guarantee you, if they didn't do this, and only sold the 'unlocked' chips, then the chip's selling price would be higher than what the locked chip is right now.

Nobody's getting ripped off, this is just how economics.

Re:$200 should have bought full functionality then

[dcposch]

Not quite. Hypothetically, what it only costs them $50 to manufacture, but it costs them $6 billion a year in R&D to develop?

Then, selling i5s for $200 and i7s for $300 might be a perfectly fair price, and doing that by selling $200 chips with a $100 optional software upgrade might be reasonable as well.

Re:$200 should have bought full functionality then

[hob42]

But less so than you already are today. At least with this approach, it's a lot cheaper to later upgrade such a chip, because you don't have to buy an entirely new chip to get the features that were disabled on your old one. Given the two options, I'd choose this one.

Windows 7 only unlock!

[scrib]

"Currently, CPU upgrades are available on selected Windows 7 systems."

It installs the application. Does it run every time your computer boots? Does that mean the unlock isn't permanent? If I pay to unlock the chip, and then reboot into Linux, is the CPU still unlocked? If I have to reinstall Windows, do I have to reinstall (or re-purchase) the upgrade?

No thanks...

Re:Windows 7 only unlock!

[mlts]

This reminds me of an old 486 upgrade chip for the 386DX that was pin compatible. It would run the same speed as a 386DX, but one had to install a .SYS driver in MS-DOS to turn on the internal clock-doubling and such. No driver, no performance gain.

I wonder if it is the same stuff, where the CPU is fed some sequence to have it allow access to the full cache and such. Of course, I will be almost 100% sure that this driver will be not something open-sourced, so expect the performance boost by "unlocking" the chip to be only in Windows, and no other OS.

I just hope Intel doesn't spread this crap beyond the bottom of the barrel chips where profit margins are razor thin. For the low-end market where price is everything, maybe. However, for mainstream i5 and i5 chips, much less Xeons -- hell no.

Re:Windows 7 only unlock!

[jpmorgan]

Yes. Modern CPUs can be given what's called a 'microcode update' during system boot by driver. Microcode updates are volatile, so they need to be reapplied on each boot. Generally this is done to fix minor bugs that slipped through testing. In this case, Intel is allowing the microcode to unlock additional capabilities.

Only available for Windows 7 right now, but generally microcode update drivers are available for all common platforms (e.g., Linux). If Intel is serious about this business model, it's likely that they'll roll out updated microcode drivers for other supported platforms soon.

Re:Windows 7 only unlock!

[AK Marc]

There's a perception that releasing such tools for Linux will result in cracking them. Such "delicate" tools are often not released to Linux, so I'm doubtful your prediction will come true.

Re:Windows 7 only unlock!

[hedwards]

They'd better be very careful, because if that's how it works, they could very easily find themselves in a world of anti trust hurt. Plus if that's how they do it, somebody is going to pirate it, and there's not a damned thing they can do to enforce it.

Re:Windows 7 only unlock!

[TheGatesofBill]

More likely, the systems that its available on ship with Windows 7. I think you read too much into it.

intel web site has place holders on it

[Joe The Dragon]

https://retailupgrades.intel.com/Page.aspx?Name=WhereToBuy

piracy isn't even required

[mentil]

Unlock codes aren't copyrightable, so wait until the required code is reverse-engineered

web site has place holders on it.

[Joe The Dragon]

https://retailupgrades.intel.com/Page.aspx?Name=WhereToBuy

Is this some kind of flash update or os based?

[Joe The Dragon]

Is this some kind of flash update or os based?

So will it be $50 per os reload?

Will you be able to buy it one time and make a image and mass deploy it?

Will Linux just auto unlock the cpu?

Will some MB auto unlock the cpu?

Re:Is this some kind of flash update or os based?

[jpmorgan]

It's probably implemented via microcode update, so it'll require a driver that runs on every boot.

Re:Is this some kind of flash update or os based?

[Khyber]

Or just replace the BIOS with the proper microcode already in the ROM, or just put the processor in another motherboard instead of the original one.

Re:Is this some kind of flash update or os based?

[Nimey]

I'd hope there's a tiny amount of non-volatile memory inside the CPU for this, otherwise the unlock is only going to be good for one system. This would, of course, suck if your motherboard fails but the rest of the machine is OK.

Re:Is this some kind of flash update or os based?

[BitZtream]

Is this some kind of flash update or os based?

No, its likely a fuse in the CPU that only gets blown by sending the CPU a special series of commands and the proper password for that particular CPU. After it validated the passcode, it blows the fuse and on next power on, its a different chip.

And to all your other questions:

No, that would be stupid.

Any bets it only works in Windows?

[knorthern knight]

If it's a software upgrade, what OS does it require?

Really paranoid vision, or is it merely realistic... what if the the upgrade requires Windows not only to activate, but handshakes with the OS on bootup? Would the upgrade only work under Windows?

Yeah well,

[wonkavader]

I tought I taw a putty tat.

this could be interesting

[mentil]

What if you could buy a code that unlocks your CPU's multiplier? Instead of buying a special K-model CPU with the multiplier unlocked at a $100 premium, you can buy the normal processor and then unlock it later if you want. Or you can get a cheaper code that unlocks the multiplier within a certain range, allowing for some but not extreme overclockability.

I can foresee a problem

[sea4ever]

The summary says something about special software to unlock the extra cores etc.
However, everybody knows that special software doesn't run on all operating systems. You see this kind of thing with BIOS-flashing tools.

So if I hypothetically get a machine with a crippled CPU like this and I wanted the full deal, would I have to install windows, download the special software, run it, and then wipe windows to return to my main OS?
It sounds cumbersome.

Ive seen this somewhere....

[metalmaster]

Salesman: Thank you sir. Thats all the paperwork. Here are the keys to your brand new Toyota Camry. Oh wait, there's just one more caveat. We'll need you to pay the "accelerator calibration" fee

Joe: Nah, i'll take my chances.

The rest is history...

They've been doing this for forever.

[Jartan]

I'm a bit surprised everyone is going nuts over this. Intel has been doing this almost since the start. The only difference is instead of paying for the difference upfront here you get the chance to "upgrade" if it proves useful later.

Re:They've been doing this for forever.

[ChipMonk]

[citation needed]

What Intel did was find the maximum spec at which a chip WOULD work, then sell it as the next spec down (to add some extra margin of error). The practice on Intel chips first came to public awareness with 80386SX vs DX (16- or 32-bit) and i486 SX vs DX (FPU disabled or enabled).

Re:They've been doing this for forever.

[Joce640k]

PS: So has AMD. And every other chip maker.

Re:They've been doing this for forever.

[Joce640k]

Not quite. What Intel did was make the best chips it could possibly make, then it set the clock speeds according to what the distributors ordered from them.

Intel has NEVER set the clock speed/abilities of every chip according to how well the chip did in QC testing.

Price Discrimination

[tukang]

This is an example of price discrimination. Apple basically does the same thing with all their hardware when they offer iphones/ipods/ipads with different hard drive sizes. For example, "NAND flash chips on 16GB, 32GB, and 64GB iPads cost $29.50, $59, and $118, respectively". This means they're making an additional $70 and $82 profit on the higher priced items. I guess it's not as blatant as Intel but it's essentially the same strategy.

Re:Price Discrimination

[tgibbs]

No, price discrimination is when identical goods are offered at different prices to different markets. But the chip with the extra two processors enabled is not identical to the chip without those processors enabled, as you can easily prove by doing a benchmark. And a 16 GB iPad is not identical to a 32 GB iPad, because the latter has more memory. Having different profit margins on different products does not constitute price discrimination.

Re:Price Discrimination

[tukang]

Please see the "Premium pricing" section in the link I provided.

The 16GB product brings Apple a profit of $240 while the 32GB brings them a profit of $310. The products are essentially identical and only differ in the hard drive size. The reason it's price discrimination is because the larger hard drive is being charged at a far higher profit margin.

Car companies do this when they offer different "trims" for the same model and you can also see MS and Apple do this when they offer different versions of their OS. The products are essentially identical yet the "premium" product is charged at a far higher profit margin.

Re:Price Discrimination

[tgibbs]

But a 32 GB product is not "essentially identical" to a 16 GB model, because larger memory gives you substantially more options in what you can do with the product, and may even increase its usable life (since applications tend to require more and more memory over time). Consumers correctly perceive the 32 GB model as being substantially different and substantially better, and are willing to pay accordingly.

By the way, Apple has never offered differently priced versions of their OS.

Re:Price Discrimination

[tukang]

What the buyer perceives (correctly or incorrectly) is irrelevant in determining whether price discrimination is happening. If you read the coffee example in the link I provided, it could be equally argued that the premium coffee is made from a totally different bean, it tastes better, is healthier but in the end it's still coffee and what matters is that the cost to the seller was only slightly higher than the "basic" coffee yet the price was far higher.

The only thing that matters in determining price discrimination are the cost to the seller and the price i.e. the profit margin. If the seller offers a "premium" version of the same product - and that's what the 32GB version of the ipad is - and the seller prices it at a far higher profit/cost ratio then that's a way for the seller to get buyers to show their willingness to pay.

I'll give you that it's more subtle than what Intel is doing but it's textbook price discrimination (Apple is literally used in some textbooks as an example).

... if I still didn't convince you then I probably never will :)

Re:Price Discrimination

[tgibbs]

What the buyer perceives (correctly or incorrectly) is irrelevant in determining whether price discrimination is happening. If you read the coffee example in the link I provided, it could be equally argued that the premium coffee is made from a totally different bean, it tastes better, is healthier but in the end it's still coffee and what matters is that the cost to the seller was only slightly higher than the "basic" coffee yet the price was far higher.

It is. And since pricing in a free market is more dependent upon what a product is worth to the buyer than what it costs to produce, it costs more. So if you are going to call charging different prices for different products "price discrimination," then "price discrimination" ends up being just another word for a free market.

the law also says you can jailbreak stuff what BSA

[Joe The Dragon]

the law also says you can jailbreak stuff what does the BSA and others think about that?

What if Intel had a cpu that only booted windows but for $50 you can unlock any os?

That's a nice little CPU you've got there...

[thewils]

...shame if anything should "happen" to it...

Re:the law also says you can jailbreak stuff what

[ScrewMaster]

What if Intel had a cpu that only booted windows but for $50 you can unlock any os?

I'm sure they'll try something like that as soon as they can figure out how to get away with it. The problems would be more legal than technical, but as anyone knows a liberal application of funds to the proper politician can resolve such issues very cost effectively.

OEMs

[RightSaidFred99]

I think it's fairly obvious the primary target of this would be OEMs, who can now sell one unit either as an e.g. 2.8GHz desktop or as a 3.0GHz desktop with the same hardware. Less development cost, fewer very slightly different models they have to deal with, etc...

Like most such stories, people react based on emotion without any real thought. It's both hilarious and sad at the same time.

Re:OEMs

[Joce640k]

This whole page is hilarious...

The punch line for next time...

[copponex]

Crack coming out in 3...

Oh, there it is.

Re:the law also says you can jailbreak stuff what

[Kitkoan]

the law also says you can jailbreak stuff what does the BSA and others think about that?

Sorry, but thats not right. The law is very exact in how its phrased, being "bypassing a manufacturer's protection mechanisms to allow "handsets to execute software applications" is permissible". This is what makes sure things like modchips and modding consoles is still illegal. Only effect handsets aka cellphones/smartphones.

Re:Easier than disabling?

[mysidia]

I'm fine with that as long as they make the activation simple, painless, one-time, not involve permanent bloat of the BIOS or system, and not require Windows or Linux software. Specifically, the activation should be either via a boot ISO, or via a PCI card with an option ROM, and involve a permanent change to the hardware, such as breaking an internal fuse.

But no permanent drivers, system daemons, or code built into the BIOS or that the system has to run continuously to maintain the activation.

Re:Try another vendor

[hairyfeet]

If you feel this way like I do please put your money where your mouth is and support AMD. There isn't a single thing other than having the biggest ePeen that would force you to buy from this company, who has been paying off OEMs to try to kill the free market. You want an Atom? Try the AMD Neo, I've sold several and liked them so much I bought one for my dad. since they pair the Neo single and dual with Radeon GPUs you actually get decent multimedia performance from your netbook. Low end dual? Try the Athlon II or for a few dollars more the Athlon X3 or X4. for most day to day tasks they run like a champ and one can even get low wattage CPUs to help save on electricity. Finally if you're like me and like to pound your system with video transcoding get the Phenom II, I pound mine for hours with virtualdub batch transcoding and it purrs like a kitten and remains decently cool even with a stock HSF.

The only way a free market works is we vote with our dollars and don't take mistreatment from vendors. Intel has already been trying to rig the game by paying off OEMs not to stock AMD, and now they are trying to gouge for more money after the sale. As many have pointed out this is nothing but a greedy pure profit money grab, so to show your distaste do as I've done and only build and sell AMD. And when you can get a fully loaded X3 Phenom kit for just $280 it is certainly easy on your (or your customers) pocket. We need real competition, which means a healthy and strong AMD. Otherwise it'll be like the bad old days where Intel charged a king's ransom for even the junk chips and your choice was take it or do without.

Sounds as if

[turkeyfish]

they are determined to hand AMD a PR club they can use to beat Intel about the head.

I can just see the ad bylines now, "Why pay extra to have the IQ of your microprocessor raised to average, when you can get one with a higher IQ with no additional hidden costs?"

This is purely a marketing ploy to see if they can sucker consumers into accepting, so that can generate an additional profit line.

Re:Sounds as if

[BlueKitties]

Actually this move makes a lot of sense. Microprocessors aren't pricey because of materials, they're pricey because of the research and development required to make them. Now they can have less manufacturing lines, and charge for the tech accordingly. The other option is to manufacture an outdated chip (which will require similar priced materials) and sell for less because R&D has been covered by previous sales. This just cuts that step out, and manufactures a single chip and then charges for R&D accordingly. Notice how Microsoft Office 2003 is cheaper than 2010? Same concept.

Re:Sounds as if

[interkin3tic]

This is purely a marketing ploy to see if they can sucker consumers into accepting, so that can generate an additional profit line.

That was my thought too. $50 is not a huge amount of money, if they simply raised the price of this chip $10 and didn't bother with the unlock technology that is going to be cracked anyway, does anyone doubt they'd make the same amount of money as they will from this?

Re:Sounds as if

[bhcompy]

As mentioned, this already occurs. For example, AMD's 3 core Athlon II's are 4 cores with the 3rd core disabled, either to meet a quota or because it didn't pass QA.

Re:Sounds as if

[Bert64]

Yes, but AMD aren't trying to sell you an upgrade...
You may have a 3rd core which is defective, or you may get lucky and its just disabled and you can re-enable it with software for free. Either way you bought a cheaper chip.
Intel on the other hand are selling a chip which is definitely fully working, and then trying to charge you extra to make use of the hardware you've already bought. AMD aren't trying to screw money out of you, you *may* end up getting a bargain out of them.

Intent is all important.

Re:Sounds as if

[TheRaven64]

I don't have access to Intel's market research data, so I can't meaningfully predict if they would make more or less money with that approach. It's worth noting, however, that Intel's entire business model revolves around largely artificial market segmentation. The difference in production cost between their high end and low end chips is significantly smaller than the difference in sale price. They used to have much higher yields of the cheaper chips (which were just the good chips where not everything passed the tests, e.g. the FPU on the 486sx), but they quite often the yields don't fit with where their sales team wants to put the market segments and so they just cripple some chips before selling them. AMD does the same thing.

This is just a way of more dynamically adjusting the segments, as well as making upgrades cheaper. Now, rather than buying a crippled chip and replacing it with a non-crippled chip later, you buy a crippled chip and undo the crippling in software. This, effectively, lets Intel sell you the same CPU twice.

Re:Sounds as if

[JamesP]

Yes, but AMD aren't trying to sell you an upgrade...
You may have a 3rd core which is defective, or you may get lucky and its just disabled and you can re-enable it with software for free. Either way you bought a cheaper chip.
Intel on the other hand are selling a chip which is definitely fully working, and then trying to charge you extra to make use of the hardware you've already bought. AMD aren't trying to screw money out of you, you *may* end up getting a bargain out of them.

Intent is all important.

No

Most AMD Durons 1.2GHz were overclocked (multiplier unlocked), and able to be converted into Athlon 2000+ (1.6Ghz) by means of recreating bridges on the packaging of the processor.

Re:Sounds as if

[MightyYar]

Microprocessors aren't pricey because of materials, they're pricey because of the research and development required to make them.

I'm in the industry. I don't mean to downplay R&D, but the front-end (jargon for the actual chip manufacturing process) yield has a huge influence on cost. This is the main reason that die-shrinks are so important - less surface area means less chance of errors. Big chips are fantastically expensive. Another huge driver of cost is the capital involved in setting up a process. The masks that are used to "print" the microchip on the wafer are horrendously expensive.

And then you have labor. Despite a stunning array of robotic equipment, the industry is still reliant enough on manual labor that it actually pays to set up in China despite the learning curve and initial quality hit.

Re:Sounds as if

[uglyduckling]

What will be interesting is Intel's approach to third-party unlockers. They can copyright the unlocking software, but pirated versions will appear pretty quickly, but I'm sure someone will reverse engineer it and create proper third-party software. I won't use a pirated version, but from my point of view if I buy a CPU it's for me to use as I please, and I would have no qualms about using a third party unlocker.

Re:Sounds as if

[squiggleslash]

Not necessarily. Look at it this way: the AMD MegaCore 3000 and the Intel Leelu Multicore are equivalent CPUs. The MegaCore costs OEMs $99, and the Multicore $105. You're a laptop maker, which chip do you order a million of?

Now, change the equation a little: the Leelu Multicore Lite costs $85, and while less powerful than the MegaCore, can be upgraded (by the OEM if necessary) for $50. The OEM can order a million, and knows that (a) many consumers will buy the slightly less powerful machine anyway knowing they can upgrade it later for $50, and (b) they can serve that minority that both (a) needs the extra power and (b) isn't prepared to wait by making a slightly more expensive laptop for $50 more (and cater to that market in real time based upon actual, rather than predicted, demand) by upgrading the CPUs they've already ordered a million of.

End users may want to get the most powerful CPU they can get, but OEMs don't, and OEMs make the majority of chip buying decisions. This might very well help Intel in the long run.

Re:Sounds as if

[dbc]

Yes and no. Early on in a process life cycle and a processor life cycle, the yield to the top bins is usually quite constrained. Over time the fab guys tune the process and the circuit and logic guys tune the mask and drive good yield to the high bins. Then somebody says, "Suppose we had a bin even higher? Could we squeak out enough yield to launch a new speed grade?" Lather, rinse. Repeat.

And then you move the processor to a new process. Or start over with a new processor.

What you say is true some of the time. But it is a gross over generalization. Yields and bin outs are a moving target with a lot of smart people continuously adjusting their aim, with management moving the goal posts on a regular basis.

Re:Sounds as if

[hazydave]

Yup. I think the problem is only that this one may be flawed -- it may not actually speed up the system.

Intel has to price their CPUs according to an increasingly artificial tier system -- lower CPUs cost less, otherwise OEMs would buy the better CPU for the same price. They have such a demand, the actual cost of production is rarely if ever a real factor in CPU pricing these days.

Intel did well, and really loved, the days of old, when people routinely bought FPUs and CPU upgrades in retail boxes, at retail prices. So much more per-unit profit for them. Given that, for most people doing most things, any old CPU is fast enough, this trend has largely shrunk to hobbyists making their own PCs.

I thought this was wrong and stupid and evil and all kinds of dirty, nasty things when I first read about it. But think from the Intel perspective, and not just on this chip, but in general. Intel's not getting the upgrade market they used to, and yet, actual cost vs. retail or wholesales is more decoupled than ever. That's why people overclock -- Intel may sell 2.0GHz, 2.5GHz, and 3.0GHz versions of the same CPU, but if the process pretty much guarantees 100% yield at 3GHz, the lower levels are there just to keep the price of the 3GHz chip high -- they're making all they need to on the 2GHz chip.

So, given that, how about Intel makes a new chip, with a bunch of features turned off. It does the same thing the $50 OEM priced chip did last month, the OEM sells it exactly the same way, and everyone's just as happy as they were last month. Only now, the consumer drops $50 retail on that chip, and it goes faster -- Intel's back in the retail upgrade business, for essentially zero cost to them. And it's a no-brainer impulse buy -- no need to crack open that 6-month-old PC and do frightening things.

I'm not suggesting this is a Good Thing, or not evil, or implying any goodness... only that I can see how Intel might have thought of this as a good thing, and might be confused about its nearly inevitable market rejection. The real reason for that is simple: they have so many performance tiers already, they can't really offer much of an upgrade unless EVERY chip in the line has the same performance boost option. And the more of these that exist, the greater the near inevitability that it gets cracked, and everyone runs it. And that's probably not even illegal -- virtually none of the media protection laws written for content, software, etc. apply to hardware. For example, it's perfectly legal to reverse engineer your PC's motherboard and publish the resulting schematics (Sams Photofacts actually had that as a business model -- was a time you could walk down to your local radio/electronics shop and order up a schematic for just about any consumer device). I can't see DCMA or other things, as written, applying to a few magic codes dropped into an undocumented CPU register.

First it was software...

[macraig]

...and now we'll be having hardware subscriptions, too? Lovely.

We've done a pretty good job so far of putting the kibosh on software subscriptions (short of so-called Web apps, subscriptions in disguise); are we going to be as lucky with this one, with Intel's monopolistic weight behind it?

I'm not

[symbolset]

My first impression was "whoa - Gateway is still in business."

After that though, yeah. Dumb idea all the way around. They're going to get such a roasting over this. Viral Youtube videos, blog crusaders polluting every tech forum and newsgroup with this one issue, the full Tonight Show treatment. The hate that this spawns will be worth far less than all the money it could possibly bring in.

And then of course comes the question: if ideas this bad come to market, who's running the ship up there? And then the stock takes a hit.

Re:i don't understand...

[mysidia]

they already need to check if everything on it is working, because they can't offer an upgrade if it might not work hardware like, so its not as if they are asking extra for the broken chips... its just weird.

They probably don't make additional checks. If the silicone was slightly defective, and turned off for that reason, the CPU presumably won't go to one of the batches they are selling the upgrade option for.

At this point, few are likely defective.. in most cases they take perfectly fine CPUs and disable capabilities to sell them as low-end chips without hurting their market for high-end chips. It's all part of a business practice called Market Segmentation. Segment your market and sell optimal products to different segments with features those segments need at prices those segments are willing to pay for optimal profitability.

Intel's been doing this for over 10 years, nothing new about this. Taking a chip sold to one segment and then having an upgrade option, instead of a throw-away and buy a new one option.. that's new.

And more environmentally friendly, I might add.

I've got a better idea..

[lionchild]

Hey, why don't we just add $50 to the price of the processor, and skip the whole upgrade idea all together?

I mean honestly, how many folks would really rather by a 'Celeron' type processor when for $50 more they can get the real deal, and not the disabled one?

Think of all the money Intel would save on marketing alone, by just having one new line instead of two! Hey, can I have 10% of the marketing budget I just saved Intel? I suspect I could retire really early.

Re:I've got a better idea..

[headLITE]

This may come as a surprise to you, but most people who buy computers want something that can run Word and is cheap. $50 less for a difference they won't notice is a very good deal.

Re:I've got a better idea..

[lionchild]

There's something of an art, or perhaps science, to not offering bad choices to clients. They like you better when you help them spend $50 on the front end that saves them $200 on the back end when you come back and charge them for your time and the built-in upgrade. They tend to be aggrivated at the hardware you recommended, and at you for charging them more on the back end when they realize the mistake they made on the front end for buying a cheap system to run Word. :-(

It seems to me to be a good idea to make good recommendations on the front end and omit [cheap] options that will cost more later.

Genuine Intel Disadvantage

[turkeyfish]

Get your new Genuine Intel Disadvantage built in! Sounds like a great marketing strategy to let competitors brand their processors as genuinely, intentionally "brain dead".

Re:Genuine Intel Disadvantage

[Jeremy Erwin]

Marketing is bullshit. Branding is bullshit. Pay attention to the benchmarks, and pick the CPU that fits your budget and your performance requirements. If it's normal CPU, so be it. If it's a crippled CPU, so be it. If it requires a upgrade to meet your performance expectations, while still fitting your budget, so be it.

You like this car?

[Culture20]

"It's a really nice car. Great mileage, never driven except test drives. A real beaut."
"You can't afford the sticker you say?"
The dealer takes out a baseball bat. *Crash Smash*
"So with no passenger headlight and mirror, you'll save $200. I could save you $4000 total and still have the car street legal. Or we could repair the mirror and headlight for $400 added. A real value! So we have a deal?"

I hate the cheating concept of downgraded hardware; if it's not a loss-leader, then the re-enabled hardware is overpriced.

Re:You like this car?

[Vegeta99]

I had a SAAB back in the day with a computer-controlled turbocharger. I don't remember the specifics, but it was a 900 model. I had a black turbo controller. If you had the red box, you got a few extra pounds of boost. That was the only difference between two trim levels as far as the engine was concerned, same internals.

Re:You like this car?

[beaviz]

I had a SAAB back in the day with a computer-controlled turbocharger. I don't remember the specifics, but it was a 900 model. I had a black turbo controller. If you had the red box, you got a few extra pounds of boost. That was the only difference between two trim levels as far as the engine was concerned, same internals.

I still have a 900 turbo - wonderful car :)

The box you're thinking of is the APC (Automatic Performance Control) unit that adjusts boost level according to engine knock and other parameters. It's really easy to modify for better performance. Around here it's hard to find an APC unit that isn't modded.

Here's a page describing the mod:
http://900aero.com/main/tech_main_apcmod.htm

I wonder if we see this for hard drives next

[mysidia]

Or in filesystems...

You install Windows on a 1000gb hard drive you bought, and Windows only provides you 500gb of usable space. A few years later, when you are running out of space....

Dialog box: "Warning: C: only has 5 gigabytes of space left out of 500gb. You can visit http: upgrade . microsoft .com / morespace to expand your system storage capacity.... Your storage software is currently: Bronze Edition (limit: 500gb); you can upgrade to Silver (limit: 750gb) for $99.99 or Gold (limit: 1000gb) for $199.99. Or platinum for $299.99 to allow you to add a second hard drive to your computer"

And then we could have hard drive manufacturers sell 1TB hard drives that can be upgraded to 1.5TB or 2TB hard drives by running a program and inputting an activation code from a web site......

Re:I wonder if we see this for hard drives next

[tomz16]

Why not?
I really fail to see what the issue is here, as long as
#1) it's a free/competitive market
#2) you know exactly what you are paying for

Let's say there are two (competitive) brands of hdd manufacturers A, and B, and two sizes, 1TB, and 2TB.

The hypothetical prices are as follows

Company A
$49 1TB
$99 2TB

Company B:
$50 1TB non-upgradeable
$60 1TB upgradeable to 2TB
$50 software upgrade 1TB->2TB
$100 2TB non upgradeable

Where's the problem? *sure* you can say that the hardware in the 1TB upgradeable model is IDENTICAL to the 2TB model, and that the *evil* HDD manufacturer is pocketing the extra $40 they make on 2TB drives, and gimping *your* hardware to lock down that last 1TB... but you CHOSE to pay a competitive rate to get that drive!

Now call Company A (AMD), and Company B (Intel), and replace TB with GHz... etc. etc. Who cares whether silicon for cache is physically missing, traces are laser-cut to set the multiplier, or there is some voodoo microcode keeping the full hyperthreading features of the hardware from you. The end-result is the same --> The chip was still sold on the free market at a particular price, with a particular set of features, and you chose to buy it. You knew exactly what you were getting when you handed the cash over!

Note #1) personally, i think this new scheme is a great idea compared to the current alternative. As you get further into a lithography process, the yields go up. This means that you no longer bin based on chip capability, but based on market demands. Therefore in the current system both intel and AMD are gimping perfectly functional chips by locking multipliers, cutting traces for extra cores/cache, etc. If the gimping were done in software you could buy that same cheap CPU today, and have the option to painlessly unlock it later with a few mouseclicks. Less frustration, and less computers in landfills.

Note #2) I'd personally like to see a multiplier unlock program. Pay extra for a hardware-unlock to your multiplier which also openly extends your factory warranty for the overclock range that was unlocked.

Note #3) Those predicting that this will be easily cracked/pirated may be premature. If it's implemented in BIOS via a microcode update, then it'll be trivially easy (esp. if one of the mainboard manufacturers chooses not to play ball). However, e-fuse technology now exists, and cpu's have unique serial numbers. Therefore it's entirely possible to bury a shared-secret key in the silicon, and use a cryptographic hash function to generate an unlock-key based on that secret AND the serial number of the individual CPU. The correct sequence is checked at the microcode level and blows the e-fuse, marking the CPU as upgraded. The secret will be difficult (if not impossible) to extract from the silicon without months->years with an electron microscope and thousands of sample chips.

Re:I wonder if we see this for hard drives next

[Mysteray]

Look at the photos of the actual card. There's a barcode with a bunch of numbers. There are some digits on the left, a bunch of zeroes, and then a number that's just a bit over 2^31.Presumably, there's an unknown code under the silver. We don't yet know what data gets sent over the web during the un-downgrading process, but it's quite possible that upgrades may be performed even while that system is offline, perhaps by reading codes over the telephone.

My guess is that the left-justified digits identify the Intel project within the upgrade card network. The ones on the right are the card's unique code with 32 bits of entropy. If the uncrippling process can be unlocked over the phone, there's a probably a brute force attack against the CPU. Humans just can't read long streams of digits that accurately.

So, if the key or ID space really is something like 2^32, how does a gigahertz CPU resist brute force attack? Just a theory, but it may be that after too many failed attempts, the CPU burns itself out. (That's just the kind of heavy-handed solution these customer-hating DRM types seem to love to implement.)

This raises the real possibility that the un-downgrade application contains the seeds of either a crack, or permanent hardware destruction of the affected Intel products.

Re:I wonder if we see this for hard drives next

[Anonymous Coward]

This raises the real possibility that the un-downgrade application contains the seeds of either a crack, or permanent hardware destruction of the affected Intel products.

Sounds like an awesome feature for the virus-writers of the future.

Re:I wonder if we see this for hard drives next

[Mysteray]

That's certainly possible, they may have done a great job with the implementation. Going by history, there's a good chance that they did not. They might not have wanted to pull their top engineers off of whatever they were working on for a marketing trial.

Making a hack run early in the boot process is probably a bit difficult. Maybe not many people can do it, but it only takes one. That might make it all the more interesting for those who can.

Looks like I can get a quad core AMD and mobo under $200. I heard ATI finally had some decent X drivers. I think a shopping trip is in order today.

486sx

[Spazmania]

Anybody remember the Intel 487sx math coprocessor that was really a complete 486 processor which simply disabled the original 486 processor on the motherboard? Sold like hotcakes it did. Intel at their most ingenious.

Re:486sx

[Nimey]

I looked at the 487 back in '94 or so, but went with the DX2-50 Overdrive because it was actually cheaper.

Re:I think many forget that this is a good thing

[Culture20]

In this particular case, selling the unlock is the solution to the new problem -- there is less manufacturing defects. So what would you have Intel do? Actually create a line that produces low-end products? Or intentionally break high-end products? Or stop selling low-end processors because it simply isn't profitable to do so?

"Stop selling low-end processors" sounds like a good idea. If they can make a profit selling X, no matter how it's configured, then they can make a profit selling X with the default configuration (unlocked). They should stop intentionally breaking high-end products.

Re:Ubuntu should auto-crack these.

[Kitkoan]

Linux Mint isn't illegal, it's hosted/maintained in Ireland so they aren't affected by US law. Thats how they can add the extra codecs and stuff because they are fully in compliance of Irish law, unlike Ubuntu which is hosted/maintained in the US and is thus bound by US law.

Re:the law also says you can jailbreak stuff what

[ImaLamer]

Wasn't this the idea (roughly) around trusted computing? Are kernel devs going to pay for the signing? Only MS et. al. would and thus kill free software.

HDCP Code Anybody?

[Nom du Keyboard]

Maybe Intel is using HDCP codes to unlock the processors.

Re:The 70's called, they want their meme back

[metageek]

A few years ago Larry Elison and the folks at Sun (what a strange combination ;) ) tried to do this by promoting the "network computer", a java thin client running on a diskless terminal that connects to a mainframe - Time sharing, as it was known in the 70's). It turns out the main opponent of this thing was Microsoft. Perhaps Intel finally saw a way of pulling Microsoft into this idea.

Of course, this is what cloud computing is going to lead to...

Corporate Greed

[kainalu]

Between I3, I5, I7, this speed, that speed, this feature, extreme edition, don't they already have enough ways to make money? This just sounds like corporate greed to the finest extent. People buy what computer they want, and they get what they pay for when they buy it. Not even joe User is going to pay $50 after the fact, when that same $50 at the time of purchase could have likely made the difference between .5GHz faster, or even between I3 & I5. This will blow up in their faces.

Re:Corporate Greed

[goodforusers]

What has this to do with corporate greed? It seems Intel is trying a new business model and in the end it is up to the Customers if they are willing to spend additional 50$ or not. Also, how is this different from a software upgrade? When anyone wants to upgrade to Win 7 premium from Win 7 home...what do you think actually happens? Microsoft just enables some part of the code which was previously disabled. Intel is just giving us some more options.

The real scam...

[Chris Snook]

...is that most users of low-end CPUs won't notice the difference going from 2 threads to 4, or turning on extra cache. They'll just notice their Windows 7 system getting slow, as Windows systems are wont to do, and then pay $50 only to find out that it's still just as slow, because it did nothing for their memory-starved, I/O-bound, single-threaded workload.

Re:The real scam...

[Mysteray]

Yep.

Looks like the retailer's got a piece of the action too: http://www.intel.com/cd/channel/reseller/asmo-na/eng/404392.htm "You will be eligible for a revenue share from Intel if/when your reseller customer installs an upgrade."

Looking at the card: "Effortless movement between multiple applications". Really? Wow. That's a pretty wild product claim to make for 1MB of L3 cache.

Heck, I had decided to turn off Hyperthreading at my next reboot. For some things it's a net slowdown.

We might guess that the folks buying one of these stripped-down, crippled notebooks are likely to be students or otherwise budget-constrained. It's going to suck for them to get their expectations up and part with 50 hard-earned bucks just to find out it's not that all that big of a difference in performance. I suspect they might feel a teeny bit ripped-off even.

It'll be measurable on benchmarks, but like you said, it's not going to exactly breathe new life into a low-end laptop that's sucking wind because of malware, anti-malware scanning, general Windows bloat, and/or the 10 different applications that load themselves in the system tray and memory on startup.

This was not a good move for customer loyalty, Intel. Anyone want to bet they'll end up giving all the affected customers free un-downgrades and refunds?

The take-away here is...

[Mysteray]

The take-away here is that when I buy an Intel processor, I'm not getting the best performance, I'm not getting the best price, and I'm not getting the the best value. At best, I'll get crippleware. Crippleware sucked and I'm glad it died out of the marketplace back in the late 90s.

Some Intel products open security holes on your system with their defective DRM: http://extendedsubset.com/?p=30 . I just figured they couldn't get competent C programmers after what they did to Randal Schwartz http://www.lightlink.com/spacenka/fors/ . The HDCP leak was yet another example of fail. But now they want to bring this level of quality engineering directly into the CPU? Haha, no thanks guys.

Imagine the APT malware that would be possible if the CPU microcode update protections get busted wide-open like HDCP just did.

Now was it really such a good idea to hand the Elbonian Business Network a way to sell cracks for who-knows-how-many millions of CPUs for $50 each? Congratulations Intel, the black market value of a crack on your microcode just went from $100k to $M++. Did you stop to consider the fact that some of the top supercomputers on the planet are botnets? That's right: the adversary has the computational resources of a state actor and he doesn't even pay his own power bill.

I'm sitting right now within arm's reach of 14 Intel cores I've bought within the last year or two (from Atoms to i7's), never mind the stuff I have a voice in professionally. My next general purpose CPU is coming from AMD.

Good for users - more options

[goodforusers]

This is not as bad as it looks. In fact, this could be real good for the Customers. This model will allow me to pay less for a machine when I don't need that much performance. If the performace is not meeting my needs, then I may use the upgrade card and increase the life of my machine. I don't see why folks here are making big deal out of it. How is software upgrade different from hardware upgrade? Even Microsoft and Apple do the same where they charge different price for different features and technically charge less for features by disabling some. So I think software and hardware upgrades are analgous. I actually like that Intel is thinking out of the box and trying to do something different. This only means better and more options for the Customer.

Re:Good for users - more options

[Mysteray]

So, did you create that account today or was your mom really Mrs. Goodforusers?

Re:Good for users - more options

[linhares]

He should tell his boss to go AC next time. But nice try, Dr. PR GoodForUsers.

Re:Good for users - more options

[Chowderbags]

You're not paying less to for a machine with less performance. You're paying more for the machines where you do get performance. If the "low end" chips weren't making a tidy profit, they wouldn't be sold. What they're doing is akin to if Coke came out with a "Coke with added fecal matter" brand coke priced where the regular coke is now, but doubled the price of regular coke (insert joke about New Coke here). Neither one has higher production costs, yet clearly Shit Coke was implemented only to gauge more profit out of consumers. Now imagine that coke sells 85% of all drinks (assume you're trapped in Atlanta), and has performed a large number of underhanded tricks to undermine all the other competition. Are you going to be singing the praises of Shit Coke as being the coke you drink when you don't care if something tastes like ass, but you'll gladly pay 50 cents more for a pill to go in your Shit Coke to remove the offending taste when you want something a little nicer? Is Shit Coke providing "more and better options" for the customer who may not have taste buds?

as long as AMD doesn't do it

[thetoadwarrior]

I've been using AMD for years so I don't really care. In fact I hope Intel does it and it has a considerable affect on their market share.

Intel's antics

[jkeelsnc]

Now seriously Intel? Do you really take users for being this kind of a moron? Its easy! If I see something like this I'll simply not buy a computer with an intel chip in it that I have to pay extra money to activate a feature? This is about the same as the cable companies, etc that were trying to experiment with "metered billing" and then now the CEO gets mad because customers threw a pie in his face for treating the customer as if they are really that dumb to consider it! Come on! Seriously! Absolute BS. AMD is STILL making alternate CPU's after all. Hmmm, Intel! *knock knock* are you in there!?

How is this similar?

[Ecuador]

When you try to unlock an extra core or to overclock a processor there is no guarantee it will work. The manufacturer tells you what the specs of the unit are , which is what you paid for, and from then on you are on your own.
Here we are talking about a case where the cpu has features disabled on purpose but guaranteed to work as long as you provide a ransom fee. While I can find some logic in it, they are in fact telling the consumer that they make a good profit already with the price they charge for the "crippled" unit, since they are willing to sell it at that price. Then the extra $$ is the "idiot tax" they will get from some users.
I really hope AMD returns to its early Athlon days so that Intel can be in check. Judging from the previews of their netbook APU (http://www.anandtech.com/show/3933/amds-zacate-apu-performance-update) they might have something to show next year...

Brillian

[EvanTaylor]

This is brilliant. Intel's current processors (and those in the foreseeable future) beat the snot out of AMD offerings in performance.

They can further get the low end by offering upgradeable CPUs for bargain prices (to Dell, HP, etc) further putting pressure on AMD's margins, then recoup even higher profits by offering users a upgrade path that directly pays intel.

Intel's yields are ridiculous right now, few of their procs are disabled due to defects, but to offer different market segments options.

I will pay them $250...

[Hymer]

...if they provide me with a code or utility to disable hyperthreading.
Hyperthreading is killing my VM hosts and new servers do not have the "Disable Hyperthreading" in BIOS.

Re:I will pay them $250...

[forkazoo]

You can bind specific processes to specific cores. The exact methodology will vary depending on the OS you are using, but just bind all of your VM's to only even numbered cores and you should be able to bypass any problems with HyperThreading. Though, honestly, HT generally doesn't hurt performance. I'd be curious to know how you have measured the performance drop, and what exactly led to the conclusion that HyperThreading was specifically the culprit. There may potentially be some issue with you analysis method.

Re:I will pay them $250...

[volcan0]

You can disable it when you re-compile your kernel. You'll need to manually edit your kernel config though.

Why bother with hyperthreading? it's crappy.

[Toasterboy]

For most applications where CPU power matters, hyperthreading actually hurts performance, because the scheduler assumes that the hyper cores are real cores, when in fact they are not. A hyperthreaded core is not a full core, and will bottleneck waiting for the bits it doesn't have on the real core. I prefer not even having hyperthreaded cores unless the os and applications are aware that some cores are gimpy hyperthread cores and takes that into account; i.e does not schedule on them, or only schedules tasks that are not actually crippled by dependency on the real cores on the hyper cores.

Try running a database on hyperthreaded cores....it blows chunks. It's better if you just disable them in BIOS.

Well I just checked the System monitor....

[JDmetro]

to see what kind of cpu I had. Built this system myself 4 gb ram, 4 seagate hard drives (actually gets confusing when in fdisk or your favourite equivalent.), yada yada. I went to the computer store I said I needed a motherbord/cpu that didn't have any issues with my flavour of OS. The dude behind the counter said this one will do fine blah blah. My point is I was concerned about the Hard Drives, ram, case, and how many slots were on the motherboard. Nowadays I buy based on the best deal cpu manufacture/ speed doesn't matter to the desktop user / possible occasional gamer. But $50 means I could buy extra ram or Hard Drive space.

It's not bad idea in principle

[JamesSharman]

The physical difference between your uber cpu and a z80 is half a teaspoon of sand and some subtlety in the arrangement. You don't think you actually paying that much for the physical material in your processor are you? If a cpu manufacturer just sold their top cpu design at it's best configuration with the development costs spread evenly then they would find themselves priced out of the entry level market (sell far less chips and the top ones would end up being far more expensive). All the variations in cpu's are a way to spread those design costs around while not forcing people to pay for what they don't need. What's being proposed here is brilliant in principle, put the extra stuff on the chip (Which doesn't cost them much) and give people the upgrade opportunity, which should be far cheaper for all concerned than stamping out another piece of nearly identical silicon when the customer discovers the new generation of games aren't quite fast enough. My primary concern is that if this is a boot time driver update then Intel's "upgrade" only applies to whatever operating systems they deem fit to support.

Re:It's not bad idea in principle

[chrysrobyn]

The physical difference between your uber cpu and a z80 is half a teaspoon of sand and some subtlety in the arrangement. You don't think you actually paying that much for the physical material in your processor are you?

No, I don't think I'm paying for the physical material in a processor. I think I'm paying for the time on the lithography machine to expose the 40-60 mask layers -- the most expensive part of wafer processing. Assuming Intel C4 bound instead of area bound, they were going to spend the same time here, so their costs are fixed. The second biggest component is wafer and module test -- the "locked" cores must have been fully qualified in order to advertise on the box that they are unlockable -- so Intel has already spent that money too.

Every cost that goes along with locked cores is already paid. This is a pure profit move on a piece of silicon I hypothetically bought and fully own. That doesn't set right with me.

Re:It's not bad idea in principle

[JamesSharman]

It's an interesting argument. Exactly the same argument could be made on both sides for commercial server software that is locked to given number of users. The comparison is very similar and I'm sure there are many people who would make the same counter argument in that case.

An early poster pointed out that it's common for cpu manufactures to hard lock features out (either because of defect or purely to create bigger range of product), do you object to this as well?

You argument that every cost that goes along with locked cores is already paid however just doesn't fly for me. The R&D costs of chip development are astronomical and it's exactly this portion of it that Intel are offering a compromise over.

About time.

[John Pfeiffer]

I've always wondered how long it would be until you could pirate hardware upgrades.

Privacy issue.

[leuk_he]

in PIII times intel introduced a "Processor Serial Number", where every cpu is unique. this was later not implemented to to privacy issues (you could be tracked by this number). Micorcode is already updateable to fix some bugs after production, so a unique code added to every cpu is not an issue. But in this sitation to make a hack proof solution they need

-internet acces to store unlock codes on a server (to prevent one code used multiple times)
-Unique identification of the CPU (to prevent that a unlock code from the server is replayed on other cpu's)

Oh no. Not again!

It's mostly for mass-bought office PCs

[Briareos]

Considering that you need a machine/chipset that does AMT to do this upgrade (not to mention an OEM processor you're not likely to get sold on the market) it's probably not something the regular home user is going to have to worry about. Yet.

Primitive notion of the free market

[tgibbs]

That's anti-free market. In a free market, someone using monopoly pricing (which is what you are describing) would have a competitor show up and instead price it on the lowest cost that generates acceptable profits. Now if only we had free-market capitalism, then your comments would be 100% wrong. So are you saying we don't have capitalism, or are you arguing against capitalism?

That is true only in a static, equilibrium world in which no new products or services are ever created. But in the real, dynamic world, competitors require time to develop and bring to market a product that offers equal value. So in a truly free market, a creative individual company will be able to maintain a lead over its competitors, and sell products at a price commensurate to their value to the customer.

Why sell low-end upgrades?

[kenh]

What is wrong with this? Gateway is getting cheaper silicon in their lower-end hardware, and users can upgrade performance for a few buck vs. buying a new laptop.

Intel sells the chip to Gateway at a discount, Gateway sells the laptop at a lower price, and the end user that wants better performance can choose to upgrade at a later date - those who don't want the upgrade got what they wanted, cheaper hardware.

If The performance jump were greater, say from dual to quad core, this would be great - Upgrading CPUs in a laptop are what we like to call "forklift upgrades" - albeit a very tiny forklift, this is an alternative.

A little premature?

[Peterct]

Perhaps this was announced a little early? The list of supported systems at seems to only have placeholders.
I've mirrored the placeholders at for when Intel fix the placeholder problem.

I don't get why people are upset

[GrumpySteen]

OMG! My cable company transmits a bunch of channels that my cable box could decode, but won't because I haven't paid for them to be unlocked. This makes me so mad! They should either make cable boxes that can't receive channels or let me watch all those channels for free!

Sounds fucking retarded, doesn't it?

That's what I see in 90% of the comments here.

Are these chip parts more likely to be defective?

[dave420]

Are they disabling possibly-slightly-defective areas of CPUs, selling them, and then selling a way to enable the possibly-slightly-defective areas, and $50 seems to be just enough to charge to cover the actual likelihood of any part actually having developing a fault in the newly-enabled areas (and then Intel having to issue a replacement)? That seems a pretty good deal, I guess.

car manufactures doing this for decades

[doug141]

Keyless remote receivers are built into every car whether you pay for the option or not. There's an aftermarket keyless remote industry to exploit it.

Re:car manufactures doing this for decades

[Larryish]

Got a url?

Re:car manufactures doing this for decades

[doug141]

search google shopping for "keyless remote" and your model of car.

Re:I think many forget that this is a good thing

[holophrastic]

Great. So the only CPU left is the extreme edition. I can afford it. Can you?

price gouging?

[reiisi]

Let me help you out here a bit:

...
2. You have a choice of paying $300 for a processor that should only cost $200.
3. You have a choice of paying $300 for a processor that should only cost $200.

If INTEL were some poor back-water underdog CPU manufacturer unable to turn a profit, this might be an acceptable temporary tactic to meet market prices and move towards profitability.

If INTEL were selling you some sort of service in addition to turning the function on (support contracts, or such, but what do end-users need supported for this?) this might be acceptable as a way to pay for said hypothetical service.

(Hmm. They send out an engineer to test the RAM or something?)

If INTEL were not determined to turn their near-monopoly of the (desktop/workstation) CPU into a monopoly, and were not determined to push that monopoly into the server market, with chips that burn at least an order of magnitude more energy than they should for the amount of work they do (and far more energy than necessary for the applications areas they are marketing to), there would be less negative reaction to this.

This is just evidence that the i7 is overpriced.

(I think it's overrated, too, but that's a topic for a different thread.)

Re:price gouging?

[Your.Master]

Where are you getting this notion for what things "should" cost? What economic model are you using? In a free market, there's no such thing as what something "should" cost you, the consumer. Everything is worth what its buyer will pay for it, and all that. You appear to be advocating some sort of command economy where everything must cost the least amount you can possibly charge while still turning a profit, which is only the case in an idealized free market after an infinite amount of time with omniscient consumers. But even that doesn't make your statements follow from the premises.

Let's have a thought experiment:

Widgets are a strict superset of gadgets. Widgets can be converted to gadgets for free, but gadgets cannot be converted to widgets.

You can produce either widgets OR gadgets for the following costs (it's no cheaper to produce a gadget):

Fixed costs = $1000
Variable costs = $5 / widget

You can produce 200 widgets, so obviously it takes $10 a widget to recoup your costs. But only 150 people need or want widgets and they only want one widget apiece; while 50 people are willing to buy gadgets but not at $10. So it's unprofitable to ever make widgets or gadgets, and none ever get sold. You make no profit or loss because you don't produce widgets, and nobody has widgets or gadgets even though they want them.

But 100 of those 150 would even be willing to pay $20 for a widget, and the remaining 50 would settle for a gadget that was less than $10. So you sell 100 widgets for $20 apiece, and you sell the 100 widgets as gadgets for $6. Now you're making a net profit of $100 on your $2000 investment, instead of going out of business, and 100 people have widget, and 100 people have gadgets. Everybody involved is much happier with this situation.

But then you come around and say that this proves that widgets "should" cost $6. Now you're losing even more money than before, and if any widget is defective and has to be a gadget, you now likely have a market shortage because the 100 people who are after gadgets will buy widgets at the same price, and then 100 that really wanted widgets (so much so that they would pay $20 for them) can't because only gadgets are left.

What do we learn from this thought experiment?

We learn that it is not necessarily the case that market segmentation makes things worse for *anyone*, let along making it worse for *everyone*. That doesn't mean it's always good, but there's no in-principle reason why it's bad in a market economy. It really doesn't make a difference whether Intel is a poor or profitable company. It doesn't even make a difference whether the i7 is overpriced, although it really isn't overpriced in the common sense of the word unless there's a market glut of them.

douchebag inside

[Jeek Elemental]

intel seems to be at the forefront every time some new douchebaggery scheme emerges (hdcp, chip id, in bed with ms etc.), do they really think this is a good long term strategy?

Its FAR worse than you think, be VERY upset

[voss]

https://retailupgrades.intel.com/Page.aspx?Name=Upgrade

No Linux version of the upgrade software. Its an MSI file so they cpu upgrade is only available to Windows users.
Windows users get a cpu upgrade for $50, Linux users get a CPU they cant upgrade.

Mark me redundant, mark you redundant

[reiisi]

As so many have replied to the identical assertion further up the thread,

AMD (and many others) are disabling functions that don't meet test.

You enable the function and get lucky that it works well enough for your purposes. This is different from INTEL selling you a chip that they know meets spec.

What INTEL is doing is artificially propping the price of the i7 up. And that when they don't need the profits.

I'll add a bit of non-redundant information for you.

AMD et. al. do not test every chip. (Nor does INTEL.) That costs way too much.

They use statistical testing. If they find one chip in a batch that is bad, the whole batch is sent to the recycling bin. If they find one chip in the batch that doesn't meet spec, they do some more sampling to make sure the batch is not likely to be bad, then mark the whole batch with the lower spec. And/or they disable functions. (Disabling functions to make the chip warranteeable is cheaper than testing, yes.)

So, what is happening in the AMD et. al. case is that AMD is actually pushing the price down by not testing.

If AMD tried to do this selling an upgrade trick with their downgraded parts without testing, well, they'd have to be ready to lose a lot of the upgrade fees to handling bad part returns and data damage claims.

Does that help you see the difference?

Re:Mark me redundant, mark you redundant

[m.dillon]

They do both. They disable functions that don't meet test but they also disable functions to match production to market demand. Unlocking those functions is a hit-or-miss proposition.

-Matt

Re:Mark me redundant, mark you redundant

[reiisi]

I'm actually okay with the practice, as long as it's not the market's 800 pound gorilla doing it.

Still, I'm wondering whether you claim actual, first-hand knowledge of this, or whether you are repeating what others have said.

Or, from the phrasing you use, whether you understood my post.

I know a lot of people seem to be working from the assumption that any re-enabled functionality that actually works must have been disabled strictly for marketing purposes.

And that assumption makes it hard to explain to them that there are other, valid reasons, including managing the costs of testing and scrapping.

Market leaders have certain responsibilities that non-leaders do not. If they fail to assume those responsibilities, they are showing that they are more interested in today's profit margins than the health of tomorrow's market.

One time fee?

[nurb432]

Or is that a monthly subscription?

Intel's "intellectual property"

[nurb432]

have we bought a central processing unit which has now become our property because we paid for it, or are we simply buying a "license" to use Intel's "intellectual property"?

It has always been that way. You bought the silicon, but you purchased a perpetual license to use the code in it. But this new proposal is just plain scam. The bad thing is the masses will fall for it and they will make billions.

What is next, if you don't pony up more after a particular time all the cores turn off? "Demo/Trial" CPUs?

A difference...

[Junta]

It makes no sense for a manufactured part to work like this from a business perspective. The per-unit cost to produce these parts must be enough to cover full capability (or else they are selling at a loss, which is very sketchy to do). They can deliver the full function at the same price point, and in a 'theoretical' competitive market, this means they would be forced to do so by competition. The fact they can arbitrarily bump up the price on the same exact part by 50 bucks and get away with it is an indication of either bad business judgement or a very broken microprocessor market.

In your example, your provider is using broadcast to mitigate costs. Your household not decoding the bandwidth alloted for premium content does not affect the per-unit price of the content. Similarly, software incurs no per-unit manufacturing cost for unusable content, so it can work for them.

This would be like your cable company unicasting to each house all the content they pay for. It simply wouldn't make sense to unicast unusable content to your house as the incremental cost of doing that per non-paying customer would be significant.

Re:A difference...

[gnasher719]

It makes no sense for a manufactured part to work like this from a business perspective. The per-unit cost to produce these parts must be enough to cover full capability (or else they are selling at a loss, which is very sketchy to do). They can deliver the full function at the same price point, and in a 'theoretical' competitive market, this means they would be forced to do so by competition. The fact they can arbitrarily bump up the price on the same exact part by 50 bucks and get away with it is an indication of either bad business judgement or a very broken microprocessor market.

Assume there is a million people willing to pay $100 for a certain product, and half a million willing to pay $200 for a better version. Sounds quite reasonable. Now where the software market and apparently the semiconductor market are different from other markets is that a lot of cost is in developing, but actually producing the cheaper item would cost only very little less than the more expensive item. Actually, the savings from producing only one item is so much that producing 1.5 million expensive items might be cheaper than producing the mix. _And_ it is cheap and easy to disable functionality in the expensive item, and making that reversible.

And now lets say development + production cost are $130 per item if you produce 1.5 million. So what should that manufacturer produce?

They could sell the better item for $100 and sell 1.5 million - making a loss. They could sell the better item for $200 and sell half a million - making a loss because with half a million says they don't get back the development cost. Or they could sell 1 million cheaper items for $100 and half a million better items for $200 and make a profit.

Re:Guess what?

[ScrewMaster]

How did you get an all cap post past the filter?

It was so loud it overloaded the preamp.

Re:The 70's called, they want their meme back

[Anne Honime]

The 70's called, they want their meme's back.

What's next? Pay-per-hour internet?

Pay-per-use computing?

Dumb Terminals?

Pay-per-byte storage?

Oooh, I get it...Bill Gates onces predicted "computers will be free, and people will pay for software"...it's coming true, people will eventually be back to paying per-cpu cycle for the privilege of running software.

Think it won't work? It already works on your "free" cellphones..

Pay-per-use computing, welcome to the future, circa 1970

You're very right here, Mr AC.

Artificial scarcity to keep prices up

[janwedekind]

I mean it's only CPUs (not food for poor people) but still. Artificial scarcity slows down the development of civilisation.
You could also see it as vendor lock-in or bundling. Because you don't have a choice when you decide to buy the upgrade.

Monopoly Power

[MarkvW]

If you have a monopoly, or a near monopoly, you can get away with that.

Like pricing Coke based on the weather

[Ewann]

This is a great idea but a bad implementation. If Intel wanted to build consumer goodwill while still maintaining price flexibility, they should have offered a $50 rebate to anyone willing to 'downgrade' their CPU after they bought it. This is very similar to an experiment I've heard run with soda machines: it's a great idea to be able to dynamically adjust the price of the soda based on the weather, and it's very easy to do: install a temperature sensor, write a little code, and you're ready to go. The trick is how you promote the idea. If you add a "surcharge" when it's hot outside, people get angry and think you're taking advantage of them. However, if you offer a "discount" when it's cold outside, people think they're getting a deal. You can use the same prices and just advertise it differently.

I'm for it.

[Xacid]

I'm actually all for this. I think if this is geared correctly this will actually drop the overall price down on the chips as they'll only have to make one kind. Then if you decide to upgrade later you get a code - no new hardware needed, and no e-waste. How often do you build a machine and only toss in 2gigs of ram with the thought "oh I'll upgrade that later on my own for cheaper" then never do? Maybe I'm just guilty of that, but I'm 1) lazy and 2) hate waste. If I just needed a code to turn my budget pc into a beast over time I'm all for it.

CPU Neutrality? Hardware based CrippleWare?

[stoicfaux]

3rd party software vendors could pay Intel and Microsoft to enable additional CPU resources when their software is running. Advertise and sell incredibly "cheap" CPUs, advertise and sell "cheap" software, and let users pay for a "premium" upgrade to get better app performance. It would take the CrippleWare concept to a whole new level. =P

Bad for the Consumer Market, good for Enterprise.

[yakovlev]

You're getting close, and you're right, this is definitely NOT the same as selling you a crippled chip that is upgradeable.

Many enterprise customers LOVE this functionality. Often times, enterprise software is priced by number of CPUs. As a result, customers don't WANT additional CPUs when they don't need them. However, business needs may change in the future, so they want the option to upgrade. Often, they will pay EXTRA to have that CPU in there lying dormant. If they need it, they pay to turn it on. If they don't, the amount extra they paid is less (often significantly less, due mainly to software costs) than they would have paid if they had bought the upgrade initially.

For the consumer market, I don't think this makes sense. A newly purchased computer is likely to have all the performance a user needs, even on the low end. Once that is no longer the case, one of two things will have happened:

1.) The user has so much junk running in the background that their CPU is always pegged running junk. In this case, a CPU upgrade might help, but more likely that will just get 100% utilized too. The real solution in this case (for Joe Sixpack) is to buy a new PC. As an alternative they can pay someone to reinstall Windows, which would cost about the same as the CPU upgrade and be much more effective.

2.) It's 3-5 years later and there are new applications (like HD video) that require significantly more computing power. In this case, even adding the new cores won't be enough because the cores you'll be adding are 3-5 year old cores, and there's no guarantee that other components (like the GPU) will be up to it. Again, the correct solution is to buy a new PC.

The key here is that for the majority of consumers, low-end is good enough. When they do want an upgrade they either have software problems or they need a BIG upgrade.

What are some of the other incidents?

[Futurepower(R)]

"... Intel's long history of doing things badly..."

What other incidents are on your list of Intel doing things badly?

Re:What are some of the other incidents?

[postbigbang]

They start with math co-processors that can't do math....

Re:Try another vendor

[m.dillon]

Personally speaking I try to support AMD as much as possible. Their price/performance is almost universally better, particularly when I factor in the power bill and the cost of the MB and power supply. They are considerably less political than Intel.

AMD has fallen behind a bit on PCIe and AHCI support but their new chipsets (e.g. 880G) are finally catching up, thought they are certainly rougher around the edges. It took a bit of massaging to get DragonFly's AHCI driver to probe them properly due to firmware breakage in the newer AMD chipsets (not handling IFS and PCS interrupts properly). I suspect it is partially due to the longer training/negotiation times required on a 6GBit SATA port, even when the device is only 3GBit. Still, AMD's AHCI/SATA firmware still doesn't do FBS (Fis-based-Switching) for devices behind a PM and that is annoying to say the least.

All modern chipset and MB configurations these days are measured by how many concurrent PCIe lanes they can support. AMD is doing quite well on that front.

In terms of performance Intel has the edge on the high-end, but those Intel MB/chip configurations are monsters in terms of power consumption, heat, and noise when the cpu is being run full-out, meaning one has to spend more money on cooling. I can run the AMD systems full out with cheaper case internals. The costs add up. I do like the fact that the high-end AMD consumer cpus are unlocked, and people regularly OC them past 4GHz. So far I haven't seen a need to do that to mine.

That all said, we have to ask ourselves whether the minor difference in performance even matters any more. I stuffed the new PhenomII x 6 along with the cheapest PCIe dual-port video card I could find (HD 4650) and even without real hardware acceleration in the X driver my X display is ten times faster than the one on the machine I bought just two years ago. Performance has far outstripped my needs even when I'm doing bulk package source builds that utilize all available cpu horse-power. The new MBs can take up to 16G of ram (I have 8G stuffed for now)... it's difficult to find a use for all of that ram.

In fact, the issue for me now in terms of getting the most out of my systems has devolved down to just storage bandwidth. I've taken to adding a small SSD as an intermediate meta-data cache which greatly improves find/ls/file-lookup performance on my multi-TB filesystems. Even a system with lots of ram can use a 40G mid-level SSD meta-data cache. The SSD cache has been far more effective than adding more spindles (RAID).

-Matt

Re:Every silver lining has a cloud!

[Fantastic Lad]

How does one contort their brains to the extent where that seems the reasonable outcome?

Just use the logic you offered but in reverse, then instead of calling it "contorted" call it "sane".

You sound like a character from a Douglas Adams novel.

Artificially gimping an otherwise perfectly good product in order to create a market is screwed up. Rationally, you charge extra for the extra effort required to provide higher quality and higher performance. That makes sense and it is morally acceptable. In this case, since it takes no extra effort to provide the highest quality, any price differential is now based on falsehoods and psychological manipulations rather than honest pay for honest work.

Just because this is common practice does not make it sane. It makes it disrespectful. It's disrespectful to manipulate people as though they were lab rats.

The fact that you're okay with that, means you've either forgotten or don't have the capacity to understand what it means to be truly human. There are a lot of human-like entities freely walking around in the world, but it doesn't mean they should be.

-FL

Pull back!

[Fantastic Lad]

Jeezuz. I think you need to pull your eyes back a few more inches to look at the big picture.

This is the same issue as Net Neutrality. Creating artificial shortages in order to stratify the public into haves and have-nots when everybody could happily have the same highest level of quality for the same reasonable price.

If this system comes into effect in its fullest expression, do you honestly believe that you would even be able to afford a "first class" processor? If when buying hardware, you even entertain the IDEA that you might like to upgrade later, (as you indicate), then you don't have enough money to ever be a first class computer user.

The only reason I'd be for this idea is that artificial limitations can be hacked, and usually are. But then it would be just one more intensely stupid way that people would be criminalized.

Fuck Intel on this. It's a bad call.

-FL

Re:Pull back!

[Xacid]

"first class computer user" Orly? I don't need to be some die-hard super-gamer. I just need my shit to run. As requirements grow so can my hardware. Look at it in terms of mass-manufacturing. Things can only get so cheap to produce and the more you're making of a product the more overhead there is with each item added. If as consumers we only wanted the best of the best and price wasn't an object then everyone would be driving luxury cars and have giant mansions.

Re:Pull back!

[Fantastic Lad]

If as consumers we only wanted the best of the best and price wasn't an object then everyone would be driving luxury cars and have giant mansions.

The problem is that in this particular case the class division is entirely artificial; price ISN'T an object. What's wrong with everybody having luxury cars and giant mansions if it doesn't stress the system, which in this case, it doesn't?

I reject the notion that most people should live econo-class simply to make rich people feel superior. If extended to health care and food and water resources, this becomes an intolerable situation when the misery is deliberately manufactured for the express purpose of creating a visible wealth divide.

-FL

Re:Mucking with depreciation of IT costs?

[turtleshadow]

This could really muck up the depreciation cost of IT data centers.
Intel has to have run some financial models on this to go this direction.

Is the $50 unlock going to depreciated or be full cost 3 years after the initial sale?

If I got a racks that we don't have to have a pull and replace with current CPUs but could get another 1 -2 years by unlocking them I'm going to get a note from the comptroller for not choosing to spend the really low unlock cost but instead going with upgrades which will be higher.

Next will the unlock transfer?
That would really bite if it was non-transferable unlock / license.
This would also be important when a CPU does go casters up and is replaced with a like unit. Would the unlock follow the specific CPU package or the customer installation? Doing any kind of "credit" tracking is a nightmare financially and for license compliance.
Second hand sales are also a potential problem.

Anyhow I also see Viking and/or SDD makers also doing this stuff with the wacky great Sata DIMM. How many more circuits needed to unlock 1TB RAM drawn from SDD rather than the base 32 GB they sold you a license for.

Non mainframe datacenters have had "unlockable" storage upgrades for almost over decade (IE tape libs), its almost time for unlockable, SDD, CPU & SDD/SATA memory upgrades.

Missing how capitalism works

[DocJohn]

I think many commenters here don't understand how a free-market system and capitalism works.

A company can charge ANY price it wants for its products, regardless of the cost to make them. Do you really believe, for instance, that a BMW 5 series costs $20-30k more to make than a similarly equipped 3 series? Heck no -- in fact, the two cars share a lot of the same major parts (like engines in certain versions of them).

A manufacturer can set any price it feels the market can bear. And from a marketing and computer upgrade standpoint, this move by Intel makes perfect sense for the normal consumer market. Instead of having to upgrade your whole computer in 2 years, pay Intel the $50 and your computer may be good for another 2 years.

This "you're being ripped off" bit is ridiculous and would only be a logical argument in a socialist system where all prices were pre-determined by a government entity and set according to exactly what it cost to make the product + X percent markup.

Luckily, we do not live in such a system.

--
Mental health and psychology at Psych Central

It's about right of property

[kiwix]

The problem is not the new model in itself. If you consider price discrimination to be a good thing, then the new model is indeed a win-win. (Actually, if you account for the fact that Intel has to put some extra silicon in the chip to support this new model, it's not going to be a definite win-win, but let's forget about that)

The problem is that in order for this new model to work, you have to resign you property rights on the piece of harware that you buy. You will no longer be allowed to do whatever you want with it. You will only have a license to use it in some restricted way. The software industry and the entertainment industry went down that road some time ago, and many of us feel it as a bad thing. We don't want the same to happen with our hardware.

monkey mechanics

[epine]

Personally, I think that shows that the 'invisible hand' doesn't deliver more and more prosperity to the human race, something that the more enlightened economists such as Stiglitz are beginning to say aloud.

Of all the economists I've read recently, my own views are closest to Stiglitz. But I don't think he's saying what you seem to imply.

One problem with the invisible hand metaphor is that an invisible hand never takes a sick day. Everyone just assumes it shows up for work. It works under certain conditions, but those conditions are not guaranteed to exist without a steady hand at the switch. One of the services that government can provide to the economy is arranging for those conditions to exist more often than not. Sure, this is expensive, but the last time the invisible hand spread a fever and then took sick leave, the invisible hand collected a trillion dollar bonus payment. That doesn't seem right to me.

It also continues a worrisome trend in America of widening income disparity. The entire economy is shifting to drug lord structure: only the guy at the top has money to burn and chicks for free, everyone else functions with an aspirational motivation, to have a life that sucks less by moving another rung up the ladder.

It works the same in professional sports. The vast majority of athletes who try to break into the pro leagues are lucky to break minimum wage for the time and energy invested, if they don't actually lose money. You might say that the weak aspirants should know better. Try that argument on a pro scout. There's no obvious formula for picking the gems. There's a few dead ringers in every draft year, which is exactly my point. The vast majority have uncertain prospects, even athletes drafted after the top ten from the first round.

In MLB I once watched a show on the draft process which stated that 50% of the prospects with enough talent and drive to make it fall by the wayside on injuries, esp. rotator cuff. Six months off at a key point in your development is a terrible set back, even with full recovery. Athletes, especially young men, have a lot of ego. Few believe in statistics. The setback will happen to the other guy. I'm better than him. But in reality, it's mostly a coin flip.

There's a scene in Days of Thunder which I recall because Tom Cruise, posing as a racer in real life on the publicity tour, stated that he really believed his dialog when his character said that avoiding an accident in front of you is more skill than luck. And why wouldn't he? In his own profession, he's one of the chosen few. No matter what he believes in his spare time. IIRC, Tom actually said that the opposing dialog made him furious when filming the scene.

One of the problems with systems more like Somali and less like Sweden is promoting ruthlessness and cruelty and placing a low value on life at the bottom. This translates into less education for young girls (which drives global population growth), and more boys willing their way to glory with high explosives (which shrouds freedom with the Patriot Act).

So what kind of society does the invisible hand prefer? More like Sweden or more like Somalia? Or is it somehow value neutral by the virgin birth and the miracle of small government?

A better question is this: What roles must government play to ensure that privatizing profit comes along with privatizing loss (no more "too big to fail")? And what is the least expensive way for government to provide this function? And what is our rational at the end of the day that Gorden Gecko won't find yet another way to steer us over a cliff? Greed is good, but so is ensuring that the greedy are playing with their own bankrolls, and not cleverly mortgaging the system around them.

It surprises me that this thing with Intel inflames passions. The practice has been around for a long time. Circa 1980 there were expensive washing-machine disk drives where the vendor would enable half the platters at the time

Re:You are falling into their trap...

[Your.Master]

What do you expect companies to do? Not try to maximize the money they're getting for you?

In reality, if you were to cap the price difference at $10, then i5 prices would move up as well as i7 prices move down. Because even with artificial restrictions on artificial market segmentation, you try to maximize profits, not necessarily minimize consumer prices.

isp had lots of cost in ruining news groups

[Joe The Dragon]

isp had lots of cost in ruining news groups needing lot's of big raid setups in just one part.

Thought experiments with flawed assumptions

[reiisi]

Here's one for you:

Would INTEL go out of business if they failed to sell i7s as i5s?

Overpricing is not strictly evil if the company doing it doesn't effectively control the market.

Let's try another thought experiment with you.

What happens when AMD goes bankrupt and ceases operations because INTEL has sucked all the profits up with all their market engineering?

Two points I was trying to make in one post:

(1) This is different from disabling functions because the test sample produced parts that failed to meet certain tests.

(2) This is not something a market leader should do.

It's probably not something non-leaders should do, either, your too-sophisticated thought experiment not-withstanding, but that would be their choice.

Market leaders have more limited options in what they do. Otherwise, the market cannot be kept free. (Another of your false assumptions.)

Utterly not new

[argStyopa]

Anyone remember the 386SX chip? It was the 386DX with the math coprocessor deliberately disabled to sell to a different market niche.

Re:Utterly not new

[swordgeek]

Actually, no.

The 386SX was a 386DX with the math coprocessor non-functioning. Some CPUs came off the line with bad bits, and if it was the coprocessor, they'd just sell it as an SX.

The 486SX was a deliberately disabled 486DX. That was the scandal of the day.

Re:Utterly not new

[argStyopa]

Aha, I went and dug out my old magazines and yep, I was wrong but differently than you mention.

The 386DX was a full 32-bit chip, the 386SX was the 32bit internal architecture, but with a 16-bit bus. With the 386 the coprocessor was a separate part (80387). Further, because of the bus differences, the 80387 coprocessor had to purchased in the right variant (387DX to go with a 80386DX processor, 387SX to go with the 80386SX processor).

But you're right in the sense that it was the 486 series where the (now internal) coprocessor was disabled deliberately to provide the 486SX at a lower price point.

Speaking of mainframes, Control Data did this

[Graabein]

Back in the days when dinosaurs were still walking the streets, I worked for an oil company that owned a large Control Data Corp. Cyber mainframe (ah, those were the days..)

At one point my employer got an offer from CDC to upgrade the memory of their Cyber, to double the size (I think from 32 MB to 64 MB). The only problem was the price, it was so horribly expensive that even an oil company had to think about it. And debate it endlessly internally, as I recall. In the end, CDC offered to lease the upgrade to them, so they accepted.

The next day our resident CDC technician (yes, resident. These computers came with an on-site technician) walked into the computer room with a pair of wire cutters in hand. He shut down the Cyber, opened a cabinet door and cut a single wire on the backplane. "There, all done."

Needless to say, my employer was not amused, but they got the memory upgrade they paid for.

In many ways this issue is the same as downloading crippleware or time limited try-before-you-buy software. The full functionality is already there, but you haven't yet paid a license for the right to use it.

In the end this comes down to a question of virtual vs. physical ownership. If we accept that we don't own a piece of software, just the right to use it, why not the same with hardware? You paid for a processor, yes. It's physical, you carried it home from the store. The manufacturer promised you N cores, spinning at whatever GHz, and with a certain amount of cache. They delivered on that promise.

Now they're offering an upgrade without the need for a new trip to the store. Why is this bad?

Re:Speaking of mainframes, Control Data did this

[Wumpus]

The reason this is bad is that they're trying to extend the idea that you don't own the stuff you paid for from the realm of software, where as you point out this is already a pretty common practice, to physical objects that you paid for. People's expectation is that they own something they paid for, and that they can do whatever they want with it. For Intel's scheme to work it would have to be illegal for me to try and unlock the features they want me to pay for, even though I paid for the CPU.

This is all hypothetical. I won't be buying one of those. Actually, if Intel gives starts giving them away for free I might.

Besides, if your employer wasn't amused by CDC doing this, why should anyone else be amused?

Any others?

[Futurepower(R)]

That was a serious, but honest mistake. That mistake was due to mismanagement, but not extreme mismanagement.

Re:Any others?

[postbigbang]

Ok, how about taking five years to do the PA-RISC=>Itanium and developing something that only a mother could love?

The list could go on. Instead, this discussion regards blatant usury for CPUs. There is shame when value propositions are done this way, and the shame is on Intel.

They might be doing it on purpose.

[Joce640k]

The whole idea of this might be that it's 'broken'. What would you rather buy? An Intel chip that you can hack (eg. i5 to i7) or an AMD chip that you can't?

You can bet that each chip will only go up one stepping, not all the way to the top and there will be other limitations (maybe only mid-range chips will allow it).

It's a bit like AMD's black editions - they don't garantee overclocking but hint that it might work. They never harmed AMD sales. I bet broken DRM won't harm Intel sales either.

Who's holding a gun to your head?

[Joce640k]

Please be more specific...I'm not seeing anybody.

And Intel is in bed with Israel

[akayani]

DON'T BUY INTEL PRODUCTS...

Intel is 'in bed' with Israel a country who is involved in some of the worst human rights abuses on its own civilian population today. Including the use of phosphorus and DU weapons. Use your money to send a message to the world that Human and Civil Rights matter a LOT!

What is the next most important issue?

[Futurepower(R)]

I'm interested in exploring any evidence of poor or good management at Intel. Intel's failures seem to be not technological, but sociological.

Are the management shortcomings at Intel severe enough that the CEO should be replaced? I wrote this Slashdot comment that considers some of the issues: Intel buy nVidia? Replace Intel CEO Otellini?

You said, "The list could go on. Instead, this discussion regards blatant usury for CPUs." What is the next most important issue? Yes, this Slashdot story is about a practice that would result in Intel damaging its own reputation, but what is the larger picture? Is there some overlying insufficiency of management that connects all the management failures together?

Re:What is the next most important issue?

[postbigbang]

You want a consulting job?

If one considers the litigation that Intel's been under, the DOJ investigations, the settlements, the QA problems, the internal darwinian struggles, you might give a casual observation and say: oh, growing pains.

The problems with Intel amount to near-monopolistic zeal after having an "oil well in the basement" and failure to really spawn many other acquisitions into markets, the way that say, EMC can.

It doesn't help that one of their biggest allies has been sued and settled in 63 international jurisdictions-- Microsoft.

To answer directly: larger picture is a crook having trouble maintaining value, as a result of leadership and management practices. You heard it here: short the stock.

how about this ?

[average slashdotter]

WHAT THIS MEANS

PRICE DISCRIMINATION

• what Intel is doing with crippling the CPUs is called 2nd DEGREE PRICE DISCRIMINATION + 3rd DEGREE PRICE DISCRIMINATION – it is charging different prices for supposedly different versions of the same underneath silicon chip. Plus, by providing such CPUs, Intel has divided consumers into two groups with separate demand curves & charges different prices to each group – at each group's reservation price ( the price each from each group is willing to pay ), CAPTURING ALL THE SURPLUS from CONSUMER.

ECONOMIES OF SCALE, MATURITY OF LITHOGRAPHY

• As everyone said here, their chip baking processes are so mature that they have very low chips with defects from the platter. Even with ever shrinking technologies.

• If they have very similar products in terms of architecture & features, priced differently, it probably means that underneath the chips are identical & have the same number of transistors. They are just branded differently, or “crippled” to PRICE DISCRIMINATE.

COLLUSION

• Majority of the CPU market is CISC architecture ( i386, x86, etc ) share is divided between Intel + AMD – This is OLIGOPOLY, or in case 1 vs 1 a DUOPOLY. Others' share is almost negligible in this MASS market. I even go to say that other markets are niche ones - RISCs ( Power, Itanium, etc ).

  • Oligopoly, it's features & barriers of entry:

    • economies of scale ( Intel can get resources much cheaper than AMD. Or it can push retailers. Intel did this with DELL, had to settle this issue with AMD )

    • patents(Intel licensed AMD to produce i386 compatible chips, licensing SSE to AMD, AMD cross licensing x86-64 to Intel, etc)

    • technology(chip production, R&D, lithography, etc)

    • name recognition( hello, anybody knows, “intel inside”, “pentium” )

    • strategic actions by incumbent ( Intel can afford to sell their chips so cheap that AMD goes bankrupt in a short time, that is Intel can wipe off AMD, for example )

    • REMARK : because there are only a few firms, each must consider how its actions will affect its rivals and in turn how their rivals will react ( Price – P, Quantity – Q, advertising, investing in new production capacity, etc )

• Eventually, Intel & AMD will come to an EQUILIBRIUM, a state where, each is doing its best given what its competitor is doing. They both want to produce when Marginal Revenue = Marginal Cost, where each maximizes its own profit. Plus, they can collude ( which is illegal ) they can both go for collusion equilibrium. If that doesn't work, they go for COURNOT equilibrium, else – COMPETITIVE(which will never happen, as the CPU market is dominated by Intel & the CPU market is not a perfectly competitive market).

• Because Intel is dominant, we have THE DOMINANT FIRM MODEL.

  • In this scenario, the dominant firm – Intel, determines its Demand curve, as a difference between market Demand and Supply of fringe firms. Then it maximizes its profits where its Marginal Revenues = Marginal Costs.

LEMONS & WHAT ALL THIS MEANS TO CONSUMERS:

Intel is even more efficient that if it was a monopoly. Because, if there is a monopoly, there is DEADWEIGHT LOSS to SOCIETY, where neither Intel nor Consumers can enjoy their surplus. However, in this case, Intel captures all the surplus, slicing & dicing the customers & setting prices that none is left to economic inefficiency. Hence, we, as consumers, can't enjoy cheap & powerful CPUs anymore, instead we HAVE to pay what they offer.

Now, with less competition from AMD, we as consumers loose out in a long run as well.

AMD all the way

[Yaddoshi]

I have only purchased one Intel processor for a system build: the PGA370 Celeron 400MHz with a motherboard that was originally "supposed" to support the PGA370 Pentium III processor. Intel then screwed everyone by introducing the "flip chip", changed the electrical configuration on the Pentium III processors (FC-PGA370), requiring anyone who wanted to upgrade their CPU to replace their motherboard. I've never purchased an Intel processor for a system build ever since. This "Intel Upgrade Service" con is only one more grain of sand on the pile.

I do agree the software will be pirated shortly after the CPU is released, however there will still be many people foolish enough to fork over the $$.

I was indifferent to Intel vs. AMD before...

[GameboyRMH]

...but now I'm solidly in the AMD camp. Fuck you, Intel.

Same reason I now lean towards ATI video cards, even though their cards are clearly inferior at the moment. ATI open sources drivers while nVidia invents new proprietary tech like CUDA and tries to lock users in with PhysX.

Not really all that new an idea

[DarthVain]

As many posters have pointed out, this has been done with hardware for sometime now, only difference is a fix for the disable being implemented on software. Both CPU and GPU have had instances of disabling certain features through hardware, and hobbyists have figured out ways to reconnect them, if you really want to. Usually there is a chance of destroying whatever you are altering, but that's the risk you take.

At one point Graphic Cards did the EXACT same thing, where pipes, etc would be disabled, but using their driver software, through I don't think companies would trying to sell you an upgrade, just binning cards for the market an easy way, a more easy way than actually having to design a whole bunch of crippled cards. This CPU crippling is the same, it would be much easier/cheaper to simply make one chip and disable some for cheaper markets than to design and build crippled chips.

The problem with this model which the graphic card industry found out VERY quick (I think it only lasted a year or two), is that hobbyists found out REALLY quick that all the cards are exactly the same so all you had to do was crack the driver and you could turn your 100$ card into a 500$ card. This was done in no time, and everyone installed the crack software and the cheap card, as why would you ever buy the expensive one?

Intel will learn of this if they follow through with this plan, I guarantee it.

Intel has just been slashdotted

[Galestar]

We are unable to process your request at this time. Please try again in a few minutes. If you continue to get an error please contact Intel Customer Support and provide the following error code: 44F9482A

Looks like Intel needs to unlock their own chip's full capabilities.

Clock down

[phorm]

So how long until there's a virus that disables the cores on your "unlocked" CPU's?

Intel learns from Car dealers

[lsatenstein]

Isn't it a fact that car dealers charge for prep-charges and for whatever else they can get away with? So why should Intel not do it. They are already the most expensive vendor of computers on the market. I can see Intels action as opening the door to competition, and I hope the competition does gear up for it. Here is a saying for companies that gouge. You can always find better, but you can't pay more. Think about your last car purchase from a dealer and keep on smiling. The next step is to rent you the right to consume computer cycles. Greed hath no limits.

Re:I think many forget that this is a good thing

[Culture20]

Great. So the only CPU left is the extreme edition. I can afford it. Can you?

If Intel is forced by the gubmint to not use weasely business practices, then other people will be able to afford it too; the CPUs will reach a lower price point because all of the "upgrade" overhead on the chip, in the sales-teams, in the marketing groups, won't be needed, and they'll need to charge a little less just to sell more (more profit) anyway.

The point is that the "extreme edition" isn't extreme. It's what everyone's chip is, but some have just been (reparably) damaged. Overall, it's a waste of resources, fabricating intentionally defective parts.

Hope it's not...

[slapout]

...protected using HDCP encryption!

Re:I think many forget that this is a good thing

[holophrastic]

it's not a waste of resources, it's a use of resources. It's the logic that no fabrication process is going to be perfect. So you can throw away what isn't perfect, and charge more for the ones that are -- to cover the waste costs -- or you can sell the imperfect ones and have them cover their own cost, thus keeping the perfect ones at a lower price.

You seem to be under the impression that if everyone in the industry is forced to charge more, then prices will eventually come down. That's not the way it works. Prices come down because companies find innovative ways to sell smaller things for smaller dollars.

Look at the computer industry over-all. You'll notice that top-of-the-line computers cost the same ammount of dollars as they have for the last forty years! There's just a whole lot more filler at the lower ends. Look at graphics cards. AMD has what, six pricing levels? That's insane. But the top-end is still the same $300 it always was, and double-top-end gaming/workstation stuff is the same $500 it always was. But now there's a separation between $20, $40, $60, $80, $120, $200 as well.

Pedro1948

[pedro1948]

Am I allowed to say that this blows chunks? An extra $50 to unlock the full potential of the chip? Why not just raise the price, or is the extra stuff just for geeks and normal clods don't need it?

Intel seems to be an interesting, extreme case.

[Futurepower(R)]

"You want a consulting job?"

Not in this case. I'm interested in Intel because I'm writing a book partly about the sociology of technological companies.

"You heard it here: short the stock."

I'm guessing Intel will do well in the next 2 years. Intel has recently released the Q57 chipset which, when paired with new CPUs that have integrated video, provide Intel's first video that is fully sufficient for office use. The Q57 includes a VNC client that provides remote maintenance, even for BIOS settings. It seems to me that combination will be successful with companies that are interested in upgrading to new systems.

However, I'm interested in discussing any of the issues. I notice that there is only one motherboard with the Q57 chipset available on the market, from Asus. Gigabyte lists one on its web site, but a salesman told me that none are available.

Intel's success has been independent of its sometimes foolish marketing schemes, in my opinion.

Re:Intel seems to be an interesting, extreme case.

[postbigbang]

That's an interesting book. First, strip away the testosterone, the megalomania, sociopathic leadership, dripping narcissism, and bold quest to become rich out of some perversion of the American Dream, and you might have something: social darwinism and engineers bordering on autism. Not that there's anything wrong with that...

My guess is that Intel will do well for two quarters, then will die by a thousand cuts.... a languid interregnum.

ZOMG!!! Cheap overclocking!!!

[HappyPsycho]

Now I can overclock my CPU just like the big boys!!! And its cheaper too

My big Q....

[hazydave]

Ok, so Intel's reserving hyper-threading and 1MB of L2 cache as the unlocked "upgrade". Let's get past the other issues -- does the unlock get buried deep into the motherboard/CPU, or do we have to re-buy per OS, stuff like that. Intel's FAQ suggests its a Windows 7 thing...

My simple question: hyper-threading, while cheap, doesn't always improve performance. Sometimes it hurts performance. The scenario is simple: you're running multiple threads on the same L1 and L2 caches, so the caches are missing more often than before. That's of course why the 1MB of extra cache is part of the deal.

This was a quirk before this move: if hyper-threading didn't work for you, you'd turn it off, if it did, you'd turn it on; no harm, no foul. But now, they're asking for extra money for a performance boost. If I pay that, and don't get the performance boost (or worse yet, cache thrashing has me losing performance), I think there's every chance for a big old class action suit against Intel. Even if they're setting aside 1MB L2 cache on a 2MB L2 cache system, the performance could be worse, due to thrashing of the L1 cache.