Intel Threatens To Revoke AMD's x86 License

theraindog writes "AMD's former manufacturing division opened for business last week as GlobalFoundries, but the spin-off may run afoul of AMD's 2001 cross-licensing agreement with Intel. Indeed, Intel has formally accused AMD of violating the agreement, and threatened to terminate the company's licenses in 60 days if a resolution is not found. Intel contends that GlobalFoundries is not a subsidiary of AMD, and thus is not covered by the licensing agreement. AMD has fired back, insisting that it has done nothing wrong, and that Intel's threat constitutes a violation of the deal. At stake is not only AMD's ability to build processors that use Intel's x86 technology, but also Intel's ability to use AMD's x86-64 tech in its CPUs."

Comments

if they do that

[bugs2squash]

It's the end of the x86 dominance. People will just look harder to find alternatives.

Re:if they do that

[Anonymous Coward]

Bah hah hah, silly idealist.

The two-party system is here to stay in American politics and the x86 stranglehold.

Re:if they do that

[mdm-adph]

I, for one, welcome our new strong ARM'd overlords.

Re:if they do that

[hwyhobo]

People will just look harder to find alternatives

People who? Do you really think that 99% of the computer users even know what x86 means?

Re:if they do that

[Anonymous Coward]

People will just look harder to find alternatives.

People will not find an alternative because Windows doesn't run on it; for the desktop world. Will the server world be any different? Not sure; desktop will be x86 and intel.

Re:if they do that

[ORBAT]

Yeah. And it's the Year of the Linux Desktop(tm) too, right?

Re:if they do that

[ADRA]

-They- being the PC manufacturers that sell most PC's these days. -They- being the OS vendors who would be into a world of hurt trying to support every differing configuration of the x86+ based architectures....

Since x86_64 is a superset of x86, would this mean AMD couldn't even sell x86_64 based chips either?

Re:if they do that

[the linux geek]

If x86 dies, which it is in the process of doing, Microsoft will port Windows to run on SPARC, ARM, PPC, whatever comes next. Microsoft has gotten where it is by being good at business, and being good at business does not consist of pushing a dying platform that it has no vested interested in. Windows has been ported to other architectures before, and is inherently portable.

Re:if they do that

[nametaken]

I heard this argument when it was "people will stop using windows".

It's nice to think about and all, but wake me up when it actually happens.

Re:if they do that

[jellomizer]

I would have said that 4-6 years ago. But now that Intel retook the thrown for x86 chips, I doubt we will see to many people looking for alternatives.

Re:if they do that

[SIR_Taco]

Since x86_64 is a superset of x86, would this mean AMD couldn't even sell x86_64 based chips either?

Funny thing is that AMD licensed/agreed to share their x86_64 arch back to Intel.
So essentially it's:

"I'll let you play with mine if I can play with yours."

Now a 3rd party (loosely affiliated with AMD) is playing with Intel's x86, and that wasn't part of the agreement.

Re:if they do that

[DragonWriter]

People who? Do you really think that 99% of the computer users even know what x86 means?

No, but most users don't need to. Microsoft does, and Microsoft has no reason to want any one other firm to be indispensable to PC vendors the way Microsoft is. So, if the AMD cross-licensing agreement goes away and there isn't serious competition for Intel in the x86 world, I'd expect Microsoft to start supporting alternatives.

Re:if they do that

[ciroknight]

Yes. You said it yourself: x86-64 is a superset of x86; the architecture is reliant on instructions from the legacy set. No license to the legacy set, no dice on building the superset.

But that's okay, because there's no way Intel would ever pull the trigger. This is just corporate posturing to get a better crosslicensing agreement and a slice of the Foundry's pie. They'll fight it out for a couple more weeks and then a "settlement" will be reached behind some closed doors, probably with the Foundry agreeing not to mint over N non-x86 chips and some cash changing hands in whichever direction.

Re:if they do that

[bucky0]

But the big draw of windows is the inertia of 1,000,000 one-off apps that businesses have written. Microsoft would be scared of people moving to another architecture just because if people were making a (painful) switch anyway, they might look at the alternatives.

Re:if they do that

[Anonymous Coward]

Yeah, yeah, we know, and then you'll rule over all of us with your mountain of hoarded Transmeta Crusoe chips, and you'll crush the fools who all laughed — laughed, mind you — and the world will be a happier place and cancer will be cured and then Jesus will stop by to visit.

*sigh* Same thing we always get around here whenever Intel's stock price drops...

Re:if they do that

[Tiber]

HAHAHAHA.

I'm on my way to buy SUN stock right now.

Oops, maybe not.

OK, I'm off to buy stock in HP!

Errr...

I'm going to purchase some DEC stock!

Oh fooey.

Re:if they do that

[WindBourne]

Yeah, you can always be sure that those chines builders of laptops and desktops will ALWAYS push MS. Why, they would never think to build something new and cheap and use Linux to lower their costs.

Re:if they do that

[arth1]

Windows is inherently portable? Thus speaks someone who hasn't tried switching the motherboard and CPU between Intel and AMD on a HP box, and watched Windows failing to boot because the OEM only included a HAL library for one of them.

Yes, there were Alpha and MIPS versions of Windows NT. No, there haven't been any for a long, long time. If it was just a matter of passing a different CPU flag at the top level of the compiler, it would have cost MS next to nothing to continue to provide support for XP, Vista and W7. Windows has become quite married to x86 over the years, and I doubt that switching would be trivial.

And, of course, it wouldn't do much good if Windows would run on a different CPU, if all the apps which people run aren't also recompiled for the new CPU. Windows binaries aren't exactly p-code...

Re:if they do that

[fuzzyfuzzyfungus]

The trouble isn't porting Windows; but in dragging all the x86 legacy stuff with it. With the exception of a modest amount of .NET CLR stuff, which should actually be platform agnostic, virtually all of the windows ecosystem is on x86. And, as is mentioned every single time linux migration is discussed, most of that is never, ever, ever going to get ported. Obsolete software from dead companies, in house stuff, old versions that are uneconomic to upgrade, etc. Not to mention drivers.

Re:if they do that

[jjrockman]

"I'll let you play with mine if I can play with yours."

Man, if I had a nickel...

Re:if they do that

[Anonymous Coward]

sounds RISCy.

Re:if they do that

[MightyYar]

Windows is not the issue... the millions of x86-dependent applications would be the issue. This isn't Linux where you just apt-get the version for your architecture.

They'd have to do something in emulation like Apple did with Rosetta, but then the non-x86 version of Windows would run most applications slowly and so PC magazine and consumer reports and your friendly neighborhood geek would recommend sticking with x86, since MS doesn't have Apple's option of simply making the old architecture go away.

Re:if they do that

[the linux geek]

Windows is still maintained on Itanium; both it and the Alpha port could run x86 binaries at close to native (or, at one point in the Alpha's lifetime, faster.) I'm no Microsoft fanboy, but I don't think they're stupid.

Re:if they do that

[moosesocks]

Back in the day, Microsoft had ports of NT for Alpha, PPC, and MIPS.

Alpha and MIPS are both dead and gone, while SPARC and PPC have corporate overlords who seem to have no interest in catering to the consumer market. (It also seems a bit unclear as to why Sun continue to develop and produce the SPARC, given the huge costs that must be associated with it)

Re:if they do that

[neokushan]

CELLebrate good times, come on!

Re:if they do that

[Daravon]

Windows on PPC... Then Microsoft will tout how much faster PPC is than x86 based processors, and the world ends in an infinite loop.

Re:if they do that

[Anonymous Coward]

If it was just a matter of passing a different CPU flag at the top level of the compiler, it would have cost MS next to nothing to continue to provide support for XP, Vista and W7. Windows has become quite married to x86 over the years, and I doubt that switching would be trivial.

It IS just a matter of passing a different CPU flag. MS discontinued the MIPS, PPC, and Alpha versions because there was not only no demand for it, but the few people who bought it tied up lots of MS customer service time bitching that X86 programs didn't run on MIPs/PPC/Alpha.

Windows is no more married to X86 than Linux or OS X. In fact, I can tell you where to get a fairly modern Windows Kernel running on a PPC chip in pretty much any electronics store: The XBox 360.

The NT kernel was designed from the ground up to be portable. The only real reason it's currently only supporting X86 is because that's the only place there's any sort of demand. If X86 dies (And it won't. AMD and Intel both have lots to lose, though AMD more than Intel here), Microsoft will port over to PPC (Or whatever), throw on an emulation layer, and probably take the opportunity to break a whole bunch of crappy stuff in Windows that's maintained simply for backwards compatibility.

Re:if they do that

[Yvan256]

Don't blame me, I voted for Kodos!

Re:if they do that

[cbiltcliffe]

Yes, there were Alpha and MIPS versions of Windows NT. No, there haven't been any for a long, long time.

And, if I remember rightly, at least the Alpha version of NT was simply the Intel version of NT, packaged with a VM.

Don't know about the MIPS version, though.

Re:if they do that

[Volante3192]

Well, for starters, if you change the mobo and cpu, you're technically changing the system and violating the OEM licence.

The OEM manufacturer, thus, has no reason to include a HAL for the alternate chip since it was never designed or licenced to run on it.

That aside, if x86 does die out, MS WILL find a way to make Windows work on it. Basically their cash flow would depend on it.

Re:if they do that

[RightSaidFred99]

If x86 dies, which it is in the process of doing

Lol. Cite? Or to put it another way, "The reports of my death are greatly exaggerated".

At this point it's just silly and uninformed to make such a claim, as if all you think would need to happen is "compile Windows on [whatever]", "Use linux!", "the web will make ISA irrelevent!" or whatever stunningly ridiculous claptrap people use as an explanation for why this will happen soon.

Re:if they do that

[petermgreen]

People by and large want machines with decent performance that can run thier existing applications (Which means running windows, prefferablly the 32 bit intel versions)

AMD/GlobalFoundries has the most to lose here IMO since right now most people are running 32 bit operating systems and in any case x86-64 is an extention of x86. Intel probablly has the clout to get MS to make desktop editions of ia64 windows again if they have to.

Still intel does potentially have a lot to lose from this too and I strongly suspect this will be settled (and the exact terms of the settlement will be kept secret).

Re:if they do that

[nurb432]

Just like apple was in a world of hurt when they dropped the more elegant PowerPC for ix86?

Yup, really killed them off didn't it.

( I still think it was a bad move, but no sense harping on it now.. )

A complete change over would allow a more controlled HAL standard to be developed too.

Re:if they do that

[Yvan256]

Funny that you should say that, because if the world ends at the infinite loop, we'll all be running Mac OS X.

Re:if they do that

[Fulcrum of Evil]

We have virtualization now - If I can run a legacy app in a dos box, who cares what the actual hardware is?

Re:if they do that

[dow]

I was under the impression that in the developement of Alpha, PPC and Sparc versions, Windows NT was made into a small and efficient micro kernel, with most of the higher functions working on the hardware abstraction layer you mentioned. I'm pretty sure I remember reading that the x86-64 versions of XP was fairly easy for Microsoft to produce, as much of the 64 bit support was already in the Alpha version.

Re:if they do that

[Grave]

Intel has no intention of preventing AMD from making x86 chips, because they know they'll be unable to manufacture any of their own chips as well (with x86-64 licensing coming from AMD). This is purely meant to ensure that anybody who might come along and acquire the foundry business doesn't wind up trying to produce their own x86 chips. Or at least, I'd like to believe such...truthfully, I wouldn't put it past Intel to just be making a money grab here.

Either way, holding AMD in violation of their agreement means they would effectively forfeit 64-bit licensing rights as well, and that makes no sense for them.

Re:if they do that

[slummy]

Hi, meet Mr. GPU.

Re:if they do that

[Tubal-Cain]

Does AMD have any architectures not based on x86 to fall back on?

Re:if they do that

[fm6]

Not necessarily. Just because Intel can charge what the market will bear doesn't mean they will. Less competition means they can charge more, but presumably they're smart enough not to charge so much they destroy the market.

Anyway, won't happen. If AMD loses its right to make x86 chips, then Intel loses its right to make x64 chips. That's a lose-lose situation. Intel is just using the AMD reorg to gain a little leverage. They'll renegotiate the agreement to give slightly more favorable terms to Intel and that will be the end of it.

Re:if they do that

[LandDolphin]

I don't normally do the "Mod Parent Up" comments, but this really deserves it.

But that's okay, because there's no way Intel would ever pull the trigger. This is just corporate posturing to get a better crosslicensing agreement and a slice of the Foundry's pie. They'll fight it out for a couple more weeks and then a "settlement" will be reached behind some closed doors, probably with the Foundry agreeing not to mint over N non-x86 chips and some cash changing hands in whichever direction.

That pretty much sums it all up. Grandstanding.

Re:if they do that

[SebaSOFT]

Agree, ARM has been gaining grounds due to it's low (as none) power consumption when idle. So long backward compatibility tough.

Re:if they do that

[slaker]

I don't believe that's correct. NT4 had alpha-native code for the OS and as I recall Digital offered some software to translate x86 binaries into Alpha-native system calls in roughly the same way that WINE stuff runs Windows now.

I used to support Alphas and Motorola systems on NT4, and I did get to see a beta build of Windows 2000 on an Alpha at least once.

Re:if they do that

[jsrlepage]

[...] but I don't think they're stupid.

You must be new here.

Re:if they do that

[MrCrassic]

You're dreaming.

Intel has 18-wheeler-truckloads more resources for marketing than AMD will ever hope to garner. While there will always be the minority that will seek alternatives, Intel has the power to win them over, whether it'd be through financial incentives, equipment "giveaways" or brute-force, corporate style.

If AMD loses its x86 license, I'll speculate that AMD will have to choose the three obvious paths:

• 1. Sell itself to Intel, thus unilaterally giving Intel ~100% control over mainstream consumer microprocessor fabrication, production and sales, OR
• 2. Throw lots of money and time into developing a new processor spec (which will take forever and has a high risk of failure), OR
• 3. Use an older or less popular spec (The resurrection of PowerPC?).

I don't want to see AMD go down, but it's kind of sad to know that Intel has the power to do exactly that.

Re:if they do that

[hwyhobo]

I think it's a bit of a catch 22. Microsoft might undertake such a porting venture only if the new hardware platform is sufficiently popular (30-40% of the market?). BUT - how do you get there without having Windows run on it in the first place?

Re:if they do that

[Anonymous Coward]

You'd probably swallow it by accident.

Re:if they do that

[MrCrassic]

According to the ever-trustworthy Wikipedia article on x86_64 (http://en.wikipedia.org/wiki/X86-64), AMD designed this spec. I'm not sure if that translates to ownership, though I'd like to think that it does.
Even still, the question I'm wondering is if x86_64 uses licensed bits of x86 instructions in its specification. Because if that is the case, wouldn't that mean that Intel can theoretically have this also prohibited from use as well?

Re:if they do that

[Anonymous Coward]

99% is a really high target man. I doubt that 99% of computer users know where the power switch is.

Maybe you should look to 5% or 10%

Re:if they do that

[mhall119]

AMD does much more than CPUs, especially after having acquired ATI. The would be much smaller without their CPU line, but there is still enough to run a business on.

Re:if they do that

[NotBornYesterday]

It also seems a bit unclear as to why Sun continue to develop and produce the SPARC

Because there is a demand for it, and it does things that x86 doesn't. 8 cores * 8 threads = awesome virtualization abilities. The ability for SPARCs to scale up in a linear fashion to > 100 cores in a single general-purpose SMP box positions it in the high-end datacenter realm, where PPC is, but not x86. Plus, SUN isn't going it alone. Fujitsu is on the SPARC bandwagon with them.

Re:if they do that

[coopaq]

The two-party system is here to stay in American politics and the x86 stranglehold.

And operating systems and phone companies and potato chips and cereals and sexes.

Re:if they do that

[Darkness404]

Yah, unfortunately for most things more advanced than DOS, it fails on anything that isn't currently top-of-the-line. Not to mention that ARM, SPARC, and PPC architectures simply aren't fast compared to x86. Sure, they are more power efficient, but I challenge you to find a single CPU (multi-cores are acceptable, multi-CPUs not) thats based on ARM, SPARC or the PPC architecture that is at least as fast as a low-end Core i7 and are cheap enough to be included in a mid-range PC.

Re:if they do that

[Anonymous Coward]

Man, I'd give you a nickel if I could hear that more often. I suppose that would just make you a pimp...

Re:if they do that

[ADRA]

Virtualization makes the host arch less important, but we generally don't run a virtual instance of a different hardware platform. The reason is that there's a really high penalty of either recompiling (with the possibility of bugs) or full instruction emulation which makes runtime performance horrrrrrrible.

Most modern virtualization systems will run the guest OS almost natively using CPU's virtualization extensions to make the magic happen without much overhead.

Try running Windows on an ARM or PPC x86 emulator and see how long it takes before throwing your hands up in slowness frustration.

Re:if they do that

[socsoc]

Microsoft has gotten where it is by being good at business, and being good at business does not consist of pushing a dying platform that it has no vested interested in.

So now they just push a dying os that they have a vested interest in?

Re:if they do that

[blair1q]

What color is the sand you have your head in?

Re:if they do that

[Supergibbs]

Microsoft has been rewriting many apps in .NET Office 2007, (most of) VS2008, (all of) VS2010, Windows Media Center, and probably many others. I'd guess the small apps like the games and calc. Of course the kernel will never be and I'll bet even some of the above apps link to native code but I wonder if MS is positioning itself to be able to move to other architectures more easily.

Re:if they do that

[EdwinBoyd]

Don't blame me, I bought a Cyrix!

Feel free to mock me, however.

Re:if they do that

[ADRA]

Yeah, it was most likely abandoned, not because if the hardware access ability, but most likely because of the higher end software side of things.

If MS supported Alpha, PPC, etc.. they'd need to support all platforms equally making any time that ASM was used in code platform dependent. This is a lot more work than simply a flag on a compiler.

Plus, depending on how they make their logo requirements, they may have forced vendors to support all Windows platform variants to get certification. I'd hate to buy a PPC Windows to find that there are only 1/8th the supported applications that would be found on the same x86+ variant.

Re:if they do that

[agw]

If x86 dies, which it is in the process of doing, Microsoft will port Windows to run on SPARC, ARM, PPC, whatever comes next.

You must be new. There already was a PPC port and a speculated Sparc port as well. That was, what, 10 years ago?

Re:if they do that

[ivan256]

Based on personal experience, I'd say that less than 99% of CS101 students know where the power switch is... (Less than 200 students in my first CS class, two didn't know how to turn on the lab machines.)

Re:if they do that

[TubeSteak]

But that's okay, because there's no way Intel would ever pull the trigger. This is just corporate posturing to get a better crosslicensing agreement and a slice of the Foundry's pie.

RTFA carefully.
By alleging that AMD is violating the agreement, Intel can pull the trigger on AMD and still use AMD's patents.
Because of Intel's threat, AMD is saying that they can pull the trigger and still use Intel's patents.

It's an interesting game of chicken that they're both playing.

Re:if they do that

[tinkerghost]

AMD/GlobalFoundries has the most to lose here IMO since right now most people are running 32 bit operating systems and in any case x86-64 is an extention of x86

Perhaps, but both AMD & Intel would have to stop all of their current x86 runs. AMD would have to flip to ARM without x86, but Intel would have to redesign every one of their current x86 chips - I believe they all have the -64 extensions encoded on the chips.

In this economy & with the margin on chips, I have to agree this is posturing & it'll be settled without any serious consequences.

Re:if they do that

[Darkness404]

Intel would also certainly be looking at an antitrust case, if not by the US then by the EU certainly.

Re:if they do that

[kasperd]

Since x86_64 is a superset of x86, would this mean AMD couldn't even sell x86_64 based chips either?

Is it time for AMD to make some more steps in throwing out the heritage from x86? Yes an x86-64 CPU can run in backward compatible 16 and 32 bit modes. In fact they still start up in 16 bit mode from which they have to be switched to 32 and then 64 bit mode. But the 64 bit mode could have done more to remain backwards compatible with old 16 and 32 bit code, however at the time AMD made the bold decision to allow some old software to not work in the 64 bit mode. Maybe it is time to take another step in getting rid of the heritage. How about making it possible for the CPUs to start up directly in 64 bit mode? That would be a natural next step towards completely getting rid of the backward compatible modes. The 64 bit instruction set can hardly be called the same instruction set as the 32 and 16 bit ones, first of all it is 64 bits, and it also has more general purpose registers. Those are clear distinctions from the old instruction set, and it was created by AMD, so I don't think Intel could prevent them from using it. I know you can run 16 and 32 bit code in the 64 bit mode, and a lot of people still use it (at least the 32 bit code), but I still think we are at the point where the 64 bit AMD ISA is more important than the 32 bit Intel ISA.

Re:if they do that

[Anonymous Coward]

It's the end of the x86 dominance

Heil, Itanic! We salute you, even if Intel itself doesn't believe in you, the true underdog! Now, where is my 80W four core desktop Itanium product when I need one for my 800$ Linux build?

Re:if they do that

[Chris Burke]

One-off business app != legacy.

Legacy != performance is irrelevant.

Re:if they do that

[mizzouxc]

Intel will threaten but not do it; They don't want an anti-trust case on their hands.

Re:if they do that

[Darkness404]

Sell itself to Intel, thus unilaterally giving Intel ~100% control over mainstream consumer microprocessor fabrication, production and sales

And this would be approved by the EU? Sure, the US justice system doesn't really understand technology but the EU does and they would know what the effect is. With no competitors Intel could very easily stop innovating, and prices of computers would rise. The EU knows that and would not allow the buyout/merger (and if Intel does cut off AMD, then they can be looking forward to an antitrust case)

Use an older or less popular spec (The resurrection of PowerPC?).

I can't see PPC nor ARM capturing any marketshare because A) They are slower then most x86 chips and B) They have pathetic clock rates that make them nearly impossible to market even against an inferior CPU.

AMD has a lot to lose, but yet I can see Intel being hit by a massive antitrust case if they continue with this.

Re:if they do that

[ducomputergeek]

Maybe Freescale or Motorola or IBM?

Re:if they do that

[Anonymous Coward]

#4 get IBM in on the fight. Remember, the reason AMD exists is because IBM refuses to buy processors that have only one vendor that can make them.

Re:if they do that

[turgid]

So, if the AMD cross-licensing agreement goes away and there isn't serious competition for Intel in the x86 world, I'd expect Microsoft to start supporting alternatives.

Windows already ships by the million on PowerPC hardware: XBox 360.

Before the XBox 360 came out, the development environment that Microsoft was supplying was Windows ported to Power Mac hardware (G5 I believe).

AMD x86 processors aren't going away, though. This is intel just flexing its muscles, spreading FUD to get AMD's share price down and to scare consumers away from its chips.

All of these big technology companies have patent cross-licensing deals with each other. You'd better believe that intel can't survive without AMD's patents, and AMD can't survive without intel's, or Sun's, HP's, Microsoft's, TI's, FreeScale's... and so on.

Re:if they do that

[Anonymous Coward]

"thrown"

The bastard son of a crown and a throne?

Re:if they do that

[Anonymous Coward]

Exactly. Changing architectures will, at worst, break every 3rd-party application in existence, and at best, send them back several years in performance.

Re:if they do that

[prefect42]

Teehee. That's the sort of crap someone who didn't own an Itanium box would say. If only an Itanium could run a typical desktop binary compiled for IA64 as fast as an x86_64 box that costs half as much.

Re:if they do that

[Frosty Piss]

It's the end of the x86 dominance. People will just look harder to find alternatives.

Maybe not. In some ways it's a lot like the Linux vs. MS debate in that the minority have (perhaps valid) issues with the dominant player, but the majority don't care. And there are people like me who for various reasons have always chosen Intel for our Linux boxes.

Re:if they do that

[hattig]

There's a lot at stake.

It appears from the snippet of Intel-AMD agreement posted that I've seen (at Tech Report, in comments) that The Foundry Company is perfectly fine under the agreement, as AMD has a certain share of the company, and it formed from AMD's assets.

So Intel might be playing with fire. They lose this, they've just lost x86-64 - and Itanium is dead due to minimal investment in the past 5 years, and this year is when 64-bit x86 will hit the common desktop with Windows 7. More likely that AMD would get that license really loosened if they won and a bunch of money, but you know, if they're backed by ballsy Arabs...

If AMD lose, Intel could have all sorts of fun.

Re:if they do that

[DragonWriter]

Back in the day, Microsoft had ports of NT for Alpha, PPC, and MIPS.

And Windows Server still supports Itanium.

Alpha and MIPS are both dead and gone, while SPARC and PPC have corporate overlords who seem to have no interest in catering to the consumer market.

Meanwhile, Itanium is still actively developed, and has a corporate overlord that had announced plans to move it to the consumer market, before abandoning those in the face of the popularity of x86_64 which it cross-licensed from AMD, reserving Itanium for the enterprise server/HPC market.

Re:if they do that

[Anonymous Coward]

... Thus speaks someone who doesn't know what the fsk they are talking about.
You are comparing apples to plywood....
Here's another comparison in line with your way of thinking. "Picking up a house and putting it down on a different foundation will not work."
No friggin' guff! If you had one iota of clue of how the HAL is configured after the fact versus porting an OS to another platform MAYBE just MAYBE you'd figure out that it is possible.

Where did you ever read that the idea is to grab the drive from one sytem and put it into another with different architecture?

Now, if the OS is written is such a way where direct communication to the hardware is not necessary then, yes, it is possible to port the foundation (kernel/drivers) to another platform and recompile the higher layers (NetBSD anyone?)

There is no way to grab an image of Windows and expect it to run on another platform.

Passing a flag to the compiler? Lol! Talk about clueless at best there, buddy!

Modded Insightful? For what? Sheesh, others who have less of a clue than this guy!

Sure, mod me a troll, however, I'm tired of people blabbering about something they have now knowledge of AT ALL and getting brownie points.

p-code? Every check some of Vista's components? They're running on a CLR. It's abstracted, that's one of the reasons why Vista runs so slowwwww..... They tried to abstract the snot out of it in order to have an OS that can be ported to other hardware without having to rewrite half of it.

The reason why the Alpha and MIPS versions of NT are not around is simple: support or the lack of it. Hardware vendors didn't want to have to write drivers for the different platforms. If you looks at the driver development kits, Microsoft is trying to abstract that as well to make it easier to support multi platforms. Unfortunately the creates a great bit elephant of an OS with so many layers.

Re:if they do that

[drsmithy]

Windows is inherently portable? Thus speaks someone who hasn't tried switching the motherboard and CPU between Intel and AMD on a HP box, and watched Windows failing to boot because the OEM only included a HAL library for one of them.

Ah. So you agree that because I forgot to include the software RAID components into my initrd and promptly broken my Linux install, Linux isn't portable, then ?

Yes, there were Alpha and MIPS versions of Windows NT. No, there haven't been any for a long, long time. If it was just a matter of passing a different CPU flag at the top level of the compiler, it would have cost MS next to nothing to continue to provide support for XP, Vista and W7.

Yes, it would. It would cost them _massive_ amounts to QA and support multiple hardware platforms.

Windows has become quite married to x86 over the years, and I doubt that switching would be trivial.

Windows is currently maintained on at least three different platforms - and those are just the ones we know about.

Re:if they do that

[drsmithy]

And, if I remember rightly, at least the Alpha version of NT was simply the Intel version of NT, packaged with a VM.

No, it wasn't. It *was* only 32 bit thought.

You are thinking of FX!32, which was the emulator for running x86 applications on Alpha.

Re:if they do that

[DragonWriter]

If only an Itanium could run a typical desktop binary compiled for IA64 as fast as an x86_64 box that costs half as much.

If x86_64 goes away because the cross-licensing agreement is killed, that comparison won't be relevant. What will be relevant is Itanium's performance on x86 (not IA64) apps vs. other competitors performance on x86 apps, for the legacy, and performance on native (IA64) apps vs. the native performance of remaining competing processors (not x86_64, which won't be an option) at the same price, and Itanium is likely to have an initial advantage simply because of Windows (since Server is currently maintained on Itanium, its probably, of the 64-bit chips that would be left without x86_64, the one which would see native versions of the rest of the Windows line first.)

Re:if they do that

[KillerBob]

Even still, the question I'm wondering is if x86_64 uses licensed bits of x86 instructions in its specification. Because if that is the case, wouldn't that mean that Intel can theoretically have this also prohibited from use as well?

Theoretically, yes. Practically, no. While it does develop on prior art, it also implements something new, and so it's considered new technology. It'd be like me trying to copyright the letter e, and then suing you for using the word "the".

There's also no way that Intel would pull that trigger. Sure, they could stop AMD from producing chips which support x86. But they'd lose their rights to use x86-64. Seeing as so many of the computers in the world run Microsoft, that'd be corporate suicide: Microsoft, in its infinite wisdom, has deigned not to implement a 32-bit workaround to the 4GB memory addressing limit. As new computers are routinely being sold with 6GB or more RAM, they'd be writing themselves out of the market for future computer manufacture. While IA64 *is* a supported architecture in 64-bit Windows, nobody's developping for it since Intel decided to scrap the architecture and focus on building x86-64 CPU's. I'm not even sure it's on the supported list for Windows 7, to be honest...

They'd also lose out on a significant chunk of the market, as it would take time for them to start designing and building IA64-based chips again, during which time AMD, which would still be able to use x86-64 in its chip design would be able to dominate the market.

Yes, I'm oversimplifying the intricacies of CPU design. But it's easier to drop a part of your architecture that's really only there for backwards compatibility in the first place than it is to scrap an architecture completely and rework your existing technology to take advantage of an alternative so that your product can remain viable in the marketplace.

Re:if they do that

[Bert64]

It's mutually assured destruction, x86-64 has become ubiquitous enough now alongside x86 that neither side would be stupid enough. In fact, you would end up with VIA being the only manufacturer who has agreements with both parties and thus legally allowed to manufacture x86 compatible processors

Re:if they do that

[Anonymous Coward]

Virtualization makes the host arch less important, but we generally don't run a virtual instance of a different hardware platform. The reason is that there's a really high penalty of either recompiling (with the possibility of bugs) or full instruction emulation which makes runtime performance horrrrrrrible.

Ask yourself. The apps that must be virtualized (because the developer can not or will not port them) must they always be run as fast as possible for them to have any good derived from them?

I doubt if it that many, I really do.

And if it is, then there will be a market for native x86 hardware to run the unportable, unvirtualizable, yet mission critical code at blazing speed, which might be about the size of the market for supercomputers to predict the weather.

Just my opinion.

Re:if they do that

[doublebackslash]

I'd like to see this go a bit father.

Today's chips, at their core, look a lot like RISC chips. They do a lot of work to hide that, translating x86 ops to native ops. I'd like to see a chip that can run in a x86 'translated' mode and a 'native' RISC mode, much like was done with 32bit/64 bit.
this is, admittedly, a much harder task to accomplish, but exposing a more efficient RISC mode would drive OS vendors to migrate to that. With a bit of careful juggling and VM technology the chips would allow legacy code to run while exposing the more efficient native modes to software that took advantage of it.
Such a shift would take time, but so is 64bit.

Oh well, I guess I'll go back to the idea lab and keep on dreaming.

Re:if they do that

[setagllib]

The same superior generic hardware that is now threatened by this licensing war.

Re:if they do that

[hattig]

Every?

Business apps are written in .NET, Visual BASIC and Java.

Maintained applications would get recompiled - Microsoft would surely provide an easy means to compile to the new architecture! Apple managed it with far less resources. Twice.

It would be a risk, but on a five year plan ... maybe not.

Re:if they do that

[irtza]

as long as it doesn't SPARC an idea.

Re:if they do that

[Bert64]

Even more so, Linux is usable with a full suite of applications on PPC, IA64, MIPS, ARM, Alpha and others right now... Windows would have to start from scratch on an alternate architecture building a user and application base. With no apps users wouldn't switch, and with no users very few commercial developers would consider porting their apps.

Remember, Windows used to run on MIPS, PPC and Alpha, and still runs in a limited fashion on IA64... These architectures died or are dying because they had little or no closed source apps ported to them.
Compare that with linux where most apps can simply be recompiled, and the source code for 99% of apps is available so that people with a vested interest (ie the company making and promoting the hardware) can port the apps if noone else will. Look how many of the IA64 contributions in the linux kernel come from HP/Intel, and how many of the PPC contributions come from IBM.

Re:if they do that

[acohen1]

That would be 3rd base, not 2nd. Get with the picture.

Re:if they do that

[Extide]

This is pretty much all correct but you ought to leave out the false MS bashing. They do implement PAE, and only certain versions of windows are 'allowed' to really utilize it to get >4gb in 32-bit, due to their licensing policies. Windows Client basically does not allow >4gb in 32bit, Windows Server does.

Re:if they do that

[Jurily]

Do you really think that 99% of the computer users even know what x86 means?

They don't have to.

"What do you mean you're selling a computer that can't run my old games?"

Re:if they do that

[Anonymous Coward]

2/10

Re:if they do that

[Bert64]

The Cell is PPC based, can be faster than a Core i7 in some respects and is available cheaply in quantity.
If you are willing to order sufficient quantity, IBM can crank out fast and cheap PPC chips quite easily... They did it for both Sony and MS with their respective games consoles.

Emulation on the other hand will always incur a performance hit, sometimes quite a substantial one... Tho it helps if the CPU is designed to handle it. When the Alpha was still fairly new, you could run x86 emulation on it and actually outperform the real x86 hardware that was available at the time, but unfortunately Alpha was never made in sufficient quantities to push the prices down.

Re:if they do that

[Extide]

While windows may be portable, all the rest of your apps probably aren't.

Re:if they do that

[setagllib]

Which is why it lagged behind x86-64 Linux, BSD, etc. for a few years, and even today still doesn't have nearly as much driver support.

If that's "easy to produce", it is very clear why Microsoft runs behind schedule on everything else they do as well.

Re:if they do that

[Extide]

WINE does not do anything remotely close to instruction level emulation, it only emulates an ENVIRONMENT. Quite different.

Re:if they do that

[Rik Rohl]

The scary part is that Microsoft is the one that actually needs to step into this and threaten to bang both of their heads together if they don't behave. Microsoft is bigger then both of them combined and sets the direction that the PC industry is going so if they decide to start moving to something else Intel and AMD are in deep do-do. Hell, they could just drop a couple of billion in cash into a new chip design and stuff the both of them

Re:if they do that

[mmkkbb]

Why? Apple, Microsoft, and Sony (at the least) have all changed architectures without losing all binary backward compatibility.

Re:if they do that

[Xtravar]

Today's chips, at their core, look a lot like RISC chips. They do a lot of work to hide that, translating x86 ops to native ops. I'd like to see a chip that can run in a x86 'translated' mode and a 'native' RISC mode, much like was done with 32bit/64 bit.

Except wouldn't that potentially be slower? More data would need loading off the disk into memory, and from memory into cache, since the RISC translation is (usually) larger.

Don't get me wrong, there's a lot 'wrong' with IA32, but I'm not so sure it would be worth using a 100% RISC ISA as an alternative. Complex instructions can be beneficial.

Re:if they do that

[frieko]

Misconception. RISC may be more elegant but it is less efficient. Translation and register renaming take up tiny amounts of die compared to the instruction cache savings of x86.

Re:if they do that

[Anonymous Coward]

Do the overloads have a SPARCle in their eyes?

Re:if they do that

[kimvette]

Ah, but it's the right of first sale which actually applies, NOT a EULA. Besides, I now agree to amended EULAs. I now put paper up to the screen and agree to the EULA I modify, just as you would modify any contract you have a problem with, so it is immaterial in the first place.

It is a commodity good, not a work for hire. Doctrine of first sale applies, not contract law.

Re:if they do that

[Bert64]

The lack of a HAL library is a fault of how the install was configured rather than windows itself... If you compile a linux kernel yourself and only include support for the hardware on your machine, it won't boot in a different one either.

The MIPS versions of NT ran on hardware specifically designed to run NT, MS even made some such hardware themselves... It did not sell very well and is extremely rare (i've never seen a MIPS machine capable of running NT in the flesh). I doubt this hardware would be powerful enough to run current versions, and no new hardware has been released. The only MIPS hardware designed to be powerful was made by SGI and even that is discontinued, it had completely different firmware to the NT MIPS boxes and ran in big endian mode instead of little endian. NT never had any support for any of the custom hardware in the SGI boxes either.

Alpha was a slightly different story, DEC and later Compaq did most of the porting effort, MS were forced to allow this having lost a lawsuit to DEC, Compaq decided to drop support for NT and MS prompt followed suit. Compaq had decided to drop the architecture completely.

MS could potentially have continued supporting PPC, but it would have been significant work writing drivers for new hardware as it came out (MS don't write drivers for x86, the hardware makers typically provide them drivers for free).

Out of curiosity, i would like to play with NT3 or 4 on MIPS or PPC based hardware, if anyone is aware of what hardware exactly it's capable of running on...

Re:if they do that

[thedak]

*tries to find the "so true it's depressing" mod* now where did that option go....

Re:if they do that

[anss123]

I'd like to see a chip that can run in a x86 'translated' mode and a 'native' RISC mode, much like was done with 32bit/64 bit.

The inner core of modern x86 CPUs are actually more like VLIW CPUs and is also quite CISCY at times (doing stuff like add and store in a single instruction). If you could code in microcode you'd quickly discover that it lacks the sanity checks of the outer instruction set and some other "features" like a supervisor mode.

Oh well, I guess I'll go back to the idea lab and keep on dreaming.

Do that :-)

Re:if they do that

[Bert64]

The Alpha compiler had a 32bit compatibility mode called "taso", in which it restricts all pointers to being 32bit, because unlike other 64bit processors the Alpha has never had a 32bit mode. NT was compiled with this, which is why it was only ever able to support 2GB of ram on the Alpha while Linux, VMS and Tru64 had no such limits.
That said, Windows 2000 or possibly the next version were due to have a full 64bit release for Alpha. I have the Beta2/Alpha cd somewhere so i might take a look, so they had come some of the way.
Linux on the other hand had been running on 64bit Alpha, Sparc and i believe MIPS and PPC (ie both 64/32 and big/little endian architectures) long before x86-64 even existed.

Re:if they do that

[Bert64]

Writing portable drivers should not be too difficult, Linux and NetBSD prove it can be done, i used to run all kinds of PCI cards on an Alpha where i doubt anyone ever tested the driver on such hardware before. The problem is closed source driver companies don't want to go to the effort of compiling and testing - if they provide a precompiled driver people will expect it to work... And they don't want to release the source code to let other people do it (people playing with exotic architectures are likely to be technically capable and perfectly willing to compile and debug things themselves).

Re:if they do that

[Bert64]

Yes, and had any of these companies followed any kind of sensible procurement plan in the first place they'd never have got locked in to these proprietary applications from now-dead companies. And they just don't learn, despite being screwed in the past companies are still buying new proprietary crap that may well suffer the same fate (look how many companies are going bust lately).

Re:if they do that

[sjames]

People who? Do you really think that 99% of the computer users even know what x86 means?

Of course not, but they DO know what "they don't make that anymore" means. They also know what "this thing is 8 years behind the times and turtles laugh at it when it runs" means.

What they don't get (and nobody seems to be telling them) is that light browsing and email after work works fine on an old PIII.

Re:if they do that

[Darkness404]

The Cell is PPC based, can be faster than a Core i7 in some respects and is available cheaply in quantity.

Fast as in, will run desktop applications fast, or can do some obscure math calculations fast? Because, according to most reports, Core i7 architectures (or scaled down versions) will be what Intel is going to be pushing for the next few years, and in every report I have read, they totally demolish the competition (x86) in "real" speed. Whereas the Cell was more or less built to run supercomputers, render HD video and do other CPU intensive processes compared to the Core i7 which was designed more for a desktop machine.

Tho it helps if the CPU is designed to handle it.

Yes, but emulation of x86 in the CPU level (or technologies made to make emulation easier) might run afoul of Intel's agreements with AMD

Re:if they do that

[Anonymous Coward]

Woe! Ewe ectojelly spilt a wurd in their!

Re:if they do that

[FST777]

"Fast" being somewhat of a vague notion here. If you look purely at clock frequency, you are probably right (especially when we forget about speciality CPU's). If you look at FLOPS, you are wrong (consider the different GPUs on the market today). But processing speed and processing power are more than that.

I have absolutely no doubt at all that IBM can roll out a wild roaring beast of a POWER chip pretty soon after Microsoft announces plans to do a full fledged (not XBox-fledged) port of Windows for the things. Also, ponder what would happen when Microsoft goes to nVidia with a nice proposal.

Oh God! What if Ballmer has a productive game of golf with Schwarz?!?

Re:if they do that

[noidentity]

I, for one, welcome our new strong ARM'd overlords.

You'll be happy to know that Intel can provide you with those chips too.

Re:if they do that

[Anonymous Coward]

If x86 dies, which it is in the process of doing, Microsoft will port Windows to run on SPARC, ARM, PPC, whatever comes next. Microsoft has gotten where it is by being good at business, and being good at business does not consist of pushing a dying platform that it has no vested interested in. Windows has been ported to other architectures before, and is inherently portable.

Windows is portable, but the huge base of already-sold licenses for running x86 binary-only-executable applications, and the existing copies of Windows on CDs (and elsewhere), is not.

So if Microsoft brings out Windows 7 for SPARC or ARM or whatever ... people will not have any existing closed-source software at all that can run on it.

Meanwhile, there is at least one other OS, with a large and growing base of applications, that already runs on architectures like SPARC and ARM. Included in this software base is the Office suite that commands about 20% of the business market. People could move to that for no cost.

Re:if they do that

[Dogtanian]

Aside from the problems mentioned here, there's also the other issue that as soon as they exposed/permitted the native core instructions to be used, people would use them and they'd have to support them forever .

(Disclaimer: I don't know that much about chip design, and some bits of the next paragraph may be misleading, half-remembered or misrepresent Intel's motives.)

IIRC, during the x86's long development (or mutation), Intel added some features because they could, and because it seemed like a good idea at the time. Some of those features were never much used, or turned out to be not such a good idea, or were rendered mostly irrelevant by design changes in the next generation. Still, once they'd been included they had to support them for evermore, because they couldn't risk breaking compatibility with code that *did* use them.

Now, I assume that the current chips' RISC cores was designed because it suits Intel's current way of implementing the x86-emulation/execution (and as the other guy said, wasn't designed for end-user/program use). If Intel come up with a smart, new and totally different design/architecture, the way things stand, they could simply replace it with a different core that used different microcode instructions, and change the x86 "wrapper".

If Intel had exposed the microcode of the previous generation, they'd either have to stick with the old core architecture, or include it as emulation. Except because it was emulated, it probably wouldn't run as fast, and old (i.e. *existing*) programs that used the old microcode would probably run slower on the new chips. So they'd be forced to stick with the old architecture.

(Essentially it's the hybrid software/hardware equivalent of (e.g.) someone exposing the implementation details of a Java class simply because they "can" or "someone might want to use it". If in future they want to redesign that class in a more efficient manner, they have to worry about code that used the old implementation's internal workings.)

All because they exposed some microcode functionality which wasn't even meant to be anything more than a "black box" implementation detail- unsuitable for general use- in the first place.

Re:if they do that

[sheddd]

IMO the only reason X86 still dominates is AMD. Without competition Intel would've forced migration to Itanium.

Re:if they do that

[athlon02]

Just don't upset the ALPHA overload and you should be fine. Otherwise, he met send you on a one way trip on the Itanic.

Re:if they do that

[Khashishi]

oops, cancel that redundant mod. mouse slip.

Re:if they do that

[et764]

Not all virtualization requires hardware extensions. In fact, VMware was doing it long before Intel and AMD added virtualization support to their processors. VMware pulled this off by doing dynamic translation, where the virtual machine monitor would transparently rewrite native x86 into virtualized x86 code. For the most part this was just doing a straight copy, and perhaps rewriting some jump addresses. Privileged code that runs in the OS kernel had to be rewritten as something equivalent that would run fine in an unprivileged process.

This really isn't so different from running .NET or Java code. The code starts out compiled to a virtual instruction set, and the JIT compiler translates this on the fly to something that can run natively on the CPU.

This is also how Rosetta worked in Mac OS X to run PPC apps on an x86 processor. XBox 360 does a similar thing to run old XBox games, since the 360 uses a PPC processor but the old XBox was x86.

Sure, you take a performance hit in doing this, but the apps generally get rewritten to run natively eventually, and the ones that don't end up being old enough that they run faster on modern hardware even with the extra translation layer.

Re:if they do that

[Abuzar]

I, for one, welcome our new strong ARM'd overlords.
sounds RISCy.
as long as it doesn't SPARC an idea.

naw... C64[-bit] all the way!!
By 2011: C128[-bit]

Re:if they do that

[tgd]

Or someone who never owned an Alpha system.

Strictly speaking you COULD run x86 apps on them, but the performance was abysmal.

Re:if they do that

[Anonymous Coward]

Does AMD have any architectures not based on x86 to fall back on?

SPARC is open source.

http://en.wikipedia.org/wiki/Sparc
"As a result of SPARC International, the SPARC architecture is fully open and non-proprietary."

http://en.wikipedia.org/wiki/Sparc#Open_source_implementations
"OpenSPARC T2, released in 2008, a 64-bit, 64-thread implementation conforming to the UltraSPARC Architecture 2007 and to SPARC Version 9 (Level 1). Source code is written in Verilog, and licensed under many licenses. Most OpenSPARC T2 source code is licensed under the GPL. Source based on existent open source projects will continue to be licensed under their current licenses. Binary programs are licensed under a binary Software License Agreement."

Re:if they do that

[electrosoccertux]

Fortunately, the 2nd Amendment still holds.

Re:if they do that

[TheRaven64]

Not to mention the fact that it often changes a lot between processor versions. It might be nice for things like a Java VM to be able to generate the microcode directly (although, not really; you'd then need two instruction decoders in silicon), most other people would not want binaries that only worked on the P4, or the Athlon, or the Core 2, and no other x86 chips. If you're going to do that, you may as well just use ARM, SPARC, or PowerPC.

Re:if they do that

[DarkOx]

wasn't part of the agreement...as per intels understanding. I am sure these contracts are quite complex it its going to be a small army of lawers and judges to figure this one out. Something tells me AMD would not have organized its manufaturing arm in this way if they thought it was a violation of the agreements. We shall see.

Re:if they do that

[TheRaven64]

Microsoft is bigger then both of them combined

Ignoring your inability to form basic English sentences, I didn't quite believe this, so I checked. In terms of market capitalization:

Microsoft: $79.26B
Intel: $144.47B
AMD: $1.51B
Microsoft is about twice as big as Intel, in terms of current valuation, but I was amazed at how small AMD is compared to either. Microsoft could buy AMD without even noticing (well, apart from the antitrust complaints). The EU has fined Microsoft around 50% of the value of AMD; they could have bought a controlling interest in AMD for the same amount as the fines imposed by the EU. For reference, ARM has a market cap of $1.32B, slightly less than AMD, and controls around 80% of the CPU market. It's really amazing how little Intel has accomplished in comparison to their competitors.

Re:if they do that

[dkh2]

I'd like to see a chip that can run in a x86 'translated' mode and a 'native' RISC mode, much like was done with 32bit/64 bit.

Already ready to use. The Transmeta Crusoe processor does this on the fly. Of course, now they're owned (or is that pwned?) by Novafora so your guess is as good as mine whether this will survive.

Re:if they do that

[Tuoqui]

I *HIGHLY* doubt that if Intel dissolves the agreement that they'd be permitted to continue using AMD's patents. I'm sure any agreement would include something along the lines of 'if we are unable to use your patents then you are unable to use ours'

Re:if they do that

[TheRaven64]

Yes, if only Microsoft had bought a company that produced a decent x86 emulator that they could use to run legacy applications on the new platform in the same way that Apple did when they switched to x86...

Re:if they do that

[FuturePastNow]

Fast as in, will run desktop applications fast, or can do some obscure math calculations fast?

The latter. Cell has a single crappy in-order PPC processor core attached to eight fast as hell math co-processors. It is not a good general-purpose processor.

Cell rules at scientific applications written specifically for it, and happily for Sony, video game stuff (physics calculations, AI, etc) is close enough to that.

Re:if they do that

[treeves]

No, it's not really all that fun running a race with no competitors.

Re:if they do that

[Anonymous Coward]

Black, from his point of view.

Re:if they do that

[hairyfeet]

Not to mention this would hurt AMD a hell of a lot more than old Chipzilla Intel. Why? Because the majority of the world is still running 32bit OS, so Intel would still be able to sell chips. Having a 64bit CPU that was NOT backwards compatible with 32bit would be as worthless to the Windows buying public as a car that could only go in reverse. Folks don't buy CPUs for the arch, they buy it to run their programs on. Doesn't run their programs? No sale.

If I was AMD I would be trying to straighten this out but quick. But after the crazy amount of money they spent buying ATI(How much did they have to write down again?) I don't know how much sense there is in that company. I know that from the looks of things they are getting slaughtered compared to Core, and trying to out lowball Intel on price is suicide. I hate to say this, but one or two more mismanagement maneuvers by AMD and we may end up with just Intel and Via. But this major blunder could cost them serious money in lawyer bills alone if Intel pushes this, and that money is something AMD just can't afford right now. Dumb move AMD.

Re:if they do that

[Anonymous Coward]

I *HIGHLY* doubt that if Intel dissolves the agreement that they'd be permitted to continue using AMD's patents. I'm sure any agreement would include something along the lines of 'if we are unable to use your patents then you are unable to use ours'

A) For the sake of all that is literate, read the fucking article.

TFA specifically says that the cross licensing agreement allows the offended party to simultaneously revoke the offender's license and continue to use the offender's IP.

B) Your logic is completely backwards.

The agreement, as it stands, is designed as a poison pill to make it ruinous for either company to break the terms of the agreement, because their competitor gets to keep using both companies' IP. This prevents anyone from strategically pulling out of the agreement in order to dick over the other guy.

Re:if they do that

[Anonymous Coward]

That would be 3rd base, not 2nd. Get with the picture.

It would help if you would link to the picture.

Re:if they do that

[beav007]

I hate it when that happens. It's not hard to prevent either. It's just that those in charge can't be bothered :(

Re:if they do that

[beav007]

RTFA carefully.

You must be new here...

Re:if they do that

[BenLeeImp]

I can't comment on Apple, but Microsoft and Sony both have only kept partial backwards compatibility in their new consoles. I am, of course, assuming this is what you are speaking of. The new PS3s actually have no PS2 backwards compatibility. The 360 has only limited backwards compatibility - basically only the popular titles from the classic XBOX lineup.

Re:if they do that

[pleappleappleap]

Your numbers indicate that Microsoft is substantially smaller than Intel. Are your numbers reversed?

Re:if they do that

[Anonymous Coward]

But now that Intel retook the thrown for x86 chips, I doubt we will see to many people looking for alternatives.

I see you've really thought this threw.

Re:if they do that

[maxume]

You inverted Microsoft and Intel (Microsoft is the $140 billion company).

Intel basically drives investment in fab technology, I wouldn't harsh on them too much.

Re:if they do that

[Joe U]

But the big draw of windows is the inertia of 1,000,000 one-off apps that businesses have written. Microsoft would be scared of people moving to another architecture just because if people were making a (painful) switch anyway, they might look at the alternatives.

Remember how .NET compiles? Microsoft started hedging their bets a while ago.

Re:if they do that

[RudeIota]

Some good information on the switch: http://arstechnica.com/old/content/2005/10/5486.ars

Re:if they do that

[Aphoxema]

I, for one, refuse to subscribe to the belief of binary social ideals in operating systems, phone companies, potato chips, cereals and sexes. Instead, I simply accept that all things are ruled by a single, megalithic overlord that can be described as all things at once, much like a superposition of all things slashdotters have welcomed and more.

Re:if they do that

[hobbit]

Why would you want to run a shitty OS on superior hardware? Methinks the lady doth protest too much.

Re:if they do that

[irieken]

We have virtualization now - If I can run a legacy app in a dos box, who cares what the actual hardware is?

Unfortunately most of the software that the general public wants to run is only available in x86 binary form. This means that emulation, as opposed to virtualization, is required when running them on a new architecture.

On the other hand, open source software apparently loves to jump architectures... Firefox on ARM--weeeeeee!:)

Re:if they do that

[Anonymous Coward]

Fool

Re:if they do that

[nsheppar]

I wish I could believe in that, but I think it's the beginning of a monopoly. Then, in about two decades, people will finally wise up and start to do something. Then, another decade will pass, and a better successor will finally take hold.

The same thing is happening currently in the software industry.

Re:if they do that

[im_thatoneguy]

Yeah but PowerPC sucked.

And Apple users are a delusional bunch who seem to have no problem losing all backwards compatibility with their software every 4 years.

Re:if they do that

[tobiasly]

Either way, holding AMD in violation of their agreement means they would effectively forfeit 64-bit licensing rights as well, and that makes no sense for them.

Says you! That Itanium line is going to just take off any day now, and then AMD and their crazy backwards-compatible technology will be left out in the cold!

Re:if they do that

[rockNme2349]

I'd like to see this go a bit father.

I'd like to see this go 32 bits farther.

Re:if they do that

[Anonymous Coward]

I think you missed something here. The revocation of the 64 bit licenses would mean that intel would not be able to sell any of their current CPU's (except for maybe some models of the atom). There isn't some magic "remove infringing IP" from all of our designs wand. A stalemate at this point would be disastrous for both companies. This is nothing more than intel grandstanding. It will be settled privately long before it ever reaches a courtroom.

Re:if they do that

[Anonymous Coward]

I haven't moderated in months, but when I did I would make a point of clicking the white space after selecting a moderation to take focus away from the drop-down list. If you don't, using the scroll wheel will change your selection.

I also right-click on the number of the comment I intend to moderate and open it in a new tab first. Mostly because if I didn't, I'd forget that I've been moderating and close the browser before I've clicked the Moderate button at the bottom of the page.

Re:if they do that

[Asic Eng]

Saying this as a frontend chip designer - unfortunately what's really important about Intel and AMD is not the processor architecture, but their fabs and their manufacturing technology. That's the major factor controlling on-chip caches, clock speed, power consumption etc. If x86 would die, it would only be replaced by SPARC or PPC if Intel were to decide to go that route.

Re:if they do that

[Anonymous Coward]

Anyone whose written device drivers for both Pentium 4 and ARM knows making OS code portable across CPU architectures mostly amounts to #ifdef directives everywhere.

Re:if they do that

[mihalis]

Windows is still maintained on Itanium; both it and the Alpha port could run x86 binaries at close to native (or, at one point in the Alpha's lifetime, faster.) I'm no Microsoft fanboy, but I don't think they're stupid.

Alpha/NT could run x86 applications, but not x86 device drivers. One of the big problems with that version of NT was the drivers. They couldn't make it compelling the first time round. Now there are no chips available with the horsepower advantage to emulate x86 faster than native x86. Alpha is dead. Itanium is dying. Power is back to being an in-order design and even with all that cache, memory bandwidth and raw clock cycles probably still couldn't quite match a Nehalam running native x86 binaries.

Re:if they do that

[catmistake]

exactly. why do we even need chips and hardware anymore? we can just virtualize all of it.

Re:if they do that

[irieken]

Maybe Freescale or Motorola or IBM?

My money's on TI; OMAP3 is kicking ass and taking names:)

Re:if they do that

[thebjorn]

+2 Troll is Slashdot's way of saying groupthink is confused

.. or maybe just a good troll...?

Re:if they do that

[Hes Nikke]

I, for one, welcome our new strong ARM'd overlords.

You'll be happy to know that Intel can provide you with those chips too.

I wouldn't be so sure about that.

Re:if they do that

[Kalriath]

We'd essentially be shoved back to the Core Duo (NOT Core 2 Duo, the Core Duo) and earlier.

Of course, we'd just end up with a bunch of Itanium processors which are incapable of running anything instead I guess.

Re:if they do that

[angelwolf71885]

time to go kiss IBM's ass and bring back power pc chips.. by wont apple have egg on its face xD

Re:if they do that

[Michael Hunt]

That's a complete load of crap and you know it.

You've purchased (i.e. now own) one copy of an OS, which is yours to do with as you see fit (copyright law notwithstanding.) You could run it on your toaster if your toaster had a 32-bit HAL.

This 'it's illegal to run OEM windows on anything but the CPU it was sold with' line of reasoning is something that microsoft haven't exactly gone out of their way to discourage (and, in fact, their EULA says exactly that), but the force of EULAs is legally questionable at best in most jurisdictions, being trumped quite squarely by thousands of years worth of precedent for first-sale rights.

Conversely, of course, the OEM has no reason to include any given HAL, and is obviously not required to by any law or regulation. If you can't legally get your hands on the HAL you need, that's your concern. That said, a straight-down-the-line reinstall would most likely work; OEM install media isn't appreciably different (volume ID notwithstanding) from retail or volume media.

Re:if they do that

[earthforce_1]

Linux users would be less screwed since everything is open source - same apps, just new binaries to work with your old data. Windows users would be screwed royally since they would almost certainly have to relicense/repurchase all their apps, or run their old ones in emulation. But apple got away with it...

Re:if they do that

[Anonymous Coward]

Says you! That Itanium line is going to just take off any day now, and then AMD and their crazy backwards-compatible technology will be left out in the cold!

Itanium is taking off from 2001, is older than the "War on Terror".

Re:if they do that

[ishobo]

...and MIPS are both dead and gone...

I highly recommend knowing a bit about the subject matter before making asinine comments.

Sse http://www.mips.com/ for more information.

Re:if they do that

[Anonymous Coward]

IANAL, but

If A violates a contract between A & B, all it means is that B is no longer beholden to the contract.

It does not magically mean that all of a sudden Intel gets to use AMD's IP all it wants without consequence. The only way what you describe would happen is if Intel sued AMD for breach of contract, requested continued (one-way) access to AMDs IP & the judge agreed.

Not only would this be unlikely (& not just because Intel would then essentially become a monopoly with not such a good historical record of behaving well in this position), but it would take several expensive years, and Intel knows the likely outcome (even if AMD did breach the license) wouldn't really give it anything.

No the GP is correct. This smells of posturing.

Re:if they do that

[petermgreen]

emulation works to a point but

1: it's very hard to get it perfect
2: you generally lose a lot of performance. This is not an issue when emulating really old stuff but trying to emulate x86 on the comparitively slow arm is going to give terrible performance.

Sometimes you can get away with it. Apple did a pretty good job all considered. Sony screwed up pretty badly imo (even thier PSone emulation has bugs and thier partially software PS2 bc on the european PS3 was pretty terrible at least with ratchet and clank, the american PS3 with bc had pretty much all the PS2 hardware inside sidesteping the emulation problem). I haven't tried the XBOX 360 myself so I can't comment on that.

I doubt some small netbook vendor would have the resources to do this well even if there were suitable (as in faster than the intel chips they are trying to replace) CPUs on the market.

Re:if they do that

[moosesocks]

In the desktop/server market, yes. MIPS is dead and gone. Currently, they're an ARM-wannabe.

Re:if they do that

[fireheadca]

Isn't it Intel's Pig squinting and sticking out its tongue?

x86

Re:if they do that

[Plutonite]

Feel free to mock me, however.

If you think we need permission to mock you in this terrible place, you must indeed be new here.

There you go.

Wait did you say Cyrix? Never mind. Clearly my mocking you is the least of your concerns right now.

Re:if they do that

[scientus]

the Xbox 360 runs on PPC. They made one of the largest purchaces of Apple computers in order to get use to the PPC architecture.

Re:if they do that

[abelb]

Cyrix make x86 processors.

Re:if they do that

[Darby]

much like a superposition of all things slashdotters have welcomed and more.

Speaking of which, how is your mother? ;-)

Re:if they do that

[skaet]

Windows already ships by the million on PowerPC hardware: XBox 360.

I don't know where you're getting this from but it's been said that while the original Xbox may have its roots in Windows 2000, it's now so heavily stripped and altered that it is its own OS. Likewise, the 360 is again a complete custom rebuild of the OS and is also independent.

Re:if they do that

[digitalchinky]

Maybe they'll just say screw it, we're doing 128 bits? Or are we already patented up to the megabits already?

Re:if they do that

[gd2shoe]

As an aside:

Microsoft is planning a 32 bit version of Windows 7. I know because I'm running the 32 bit beta on my laptop. Don't most people running Vista already run the 64 bit version?

I don't think Windows 7 is any significant milestone for 64 bit processing. The transition is gradual and ongoing.

Re:if they do that

[artsrc]

ARM and IBM.

Re:if they do that

[ishobo]

It depends what you classify as server. My company develops custom solutions for server applicances using silicon from comapnies like RMI. And no, they are not trying to be like ARM.

Re:if they do that

[gd2shoe]

An interesting and worthy point. Still, you should give your false MS defending a rest. After all, it is their policy. Why sell a broken version of your software?... So you can up-sell your more expensive products (yes, this is exactly how MS works).

Re:if they do that

[davolfman]

It makes the Anti-trust folks nervous.

Re:if they do that

[Eskarel]

Wrong.

Nothing could have forced consumers to Itanium because the loss of legacy 32 bit code in 2001 would have been completely and totally unacceptable. The migration to 64 bit still isn't done, and for the vast majority of applications still isn't necessary.

The only way that Intel could have forced people over to Itanium would have been to stop making x86 chips and refuse to allow anyone else to make them either. This would have left them with a developer base who already had to rewrite all their applications for a new architecture and who were seriously pissed off with Intel. That would basically have been the kiss of death for Intel, and probably the biggest bonanza in the history of some other hardware manufacturer(probably Sun).

Itanium is what you base your product lines off what your engineers think is the right thing to do as opposed to what your customers actually want. Unless you're the only competitor, losing backwards compatibility before the majority of your customers are ready to give it up is suicide. If you force people to change more than their ready to change and prepared to change, they're more than likely to just change to someone else who won't do that to them.

Re:if they do that

[LackThereof]

Only NEW business apps are written in .NET, Visual BASIC and Java.

Most business apps found in the wild are old and unmaintained, written in a variety of elderly languages. They're still functional either in modern Windows installs, or on dedicated boxes running some free/cheap DOS variety.

Business offices don't upgrade their specialized, custom apps. They keep buying more licenses as they expand, but avoid upgrades at all costs while the original is still functional. Tour any large office with a specialized "business app", you will see lots of very antiquated user interfaces; some of them won't have mouse support.

Re:if they do that

[Hal_Porter]

Consoles are a special case though. They are sold at a loss and subsidised by the cost of the games.

Usually when a new generation console is introduced they have back compatibility by essentially including bits of the old console - i.e. the first PS3s, or by some hokey emulation code - the XBox360. The PS3 dropped the legacy hardware to cut prices and become profitable. The XBox360 could only play 13% of XBox games, which is actually quite an achievment considering how real time consoles are and that the CPU and GPU were totally different.

So manufacturers talk about back compatibility as a marketing bullet point. It's not really true though. I don't think it matters - people that care about old games will have the old console anyway and they can play them on that.

PCs are different to this - people have loads of software which they absolutely want to use when they buy a new machine. And compatibility break will cripple sales of a new OS.

Re:if they do that

[Hal_Porter]

Windows users would be screwed royally since they would almost certainly have to relicense/repurchase all their apps, or run their old ones in emulation. But apple got away with it...

Apple gets away with a lot of things that would kill other companies.

Re:if they do that

[Darby]

Windows is not the issue... the millions of x86-dependent applications would be the issue. This isn't Linux where you just apt-get the version for your architecture.

Which means Windows is part of the issue. That's the environment they've worked so hard to create, and an example of an alternative environment that does work well to address just such issues.

It doesn't mean that MS is "guilty", but it does mean that if you use Windows, you might have an issue therefore Windows is indeed (a big part of) the issue.

 

Re:if they do that

[debatem1]

Indeed. When I was working on putting Ubuntu on the Alpha 400 I was quite surprised by how dependent it was on closed-source drivers, but almost everything worked within about two months. By the end, the only things remaining were dependent on hand-coded ASM or entirely closed source. If there were to be a mass migration to ARM, a pretty big chunk of the Windows software ecosystem would go poof overnight.

Re:if they do that

[Hal_Porter]

That's not true. Windows NT has run on Mips, i860, Alpha, PPC and Itanium. None of them ever had even 1% of the market.

Re:if they do that

[hwyhobo]

Interesting. Was it ever workstation or only server?

Re:if they do that

[Anonymous Coward]

Anon for moderation.

You're almost right. This is a closed-source issue. I can't imagine any more buying an app without the source...it just seems like bending over and asking for it.

Of course, the place I'm working for now didn't even get the passwords for their network, so I guess there are worse things.

Re:if they do that

[Anonymous Coward]

MS bashing?

Tosser.

Re:if they do that

[ishobo]

Where are you getting the 80% number? I can believe the 80% market share for a 32-bit CPU in mobile devices. I am not buying into that number for total volume of cores shipped, and certainly not based on revenue.

Re:if they do that

[Fulcrum of Evil]

I'm sure the admins would love it if we could get rid of all the stuff that breaks.

Re:if they do that

[Hal_Porter]

NT 4.0 workstation supported x86, Mips, PPC and Alpha, just like the NT 4.0 server. If you look on the CD it included binaries for all supported architectures.

Re:if they do that

[bonch]

Except that they didn't lose backwards compatibility. Do you even know about Universal Binaries, Rosetta, or OpenStep's legacy of hardware (and even operating system) independence?

Re:if they do that

[erKURITA]

And no MIPSterious accident happens, you know...

Re:if they do that

[Ihlosi]

This is intel just flexing its muscles, spreading FUD to get AMD's share price down and to scare consumers away from its chips.

Eh, what? Have you looked at AMDs share price recently, or at their market share?

If Intel was really trying to do that, they're trying to put AMD out of business. That wouldn't be in their interest, since it'd attract even more attention from antitrust agencies.

Now, if you're a conspiracy buff, maybe Intel is trying to lose this on purpose, so AMD can stick around for a while longer and the antitrust guys keep looking the other way.

Re:if they do that

[master_p]

And by providing a JIT compiler from x86 to whatever CPU they will target next, x86 binaries could run unaltered in this new architecture.

Re:if they do that

[master_p]

Even with JIT compilers? the compiled code should not be thrown away. It should be reused in the same or next sessions.

Re:if they do that

[Anonymous Coward]

Running x86 binaries on arm is tricky enough feat. Doing the same and getting performance that's even close to the original is crossing the line to heroic deeds.

Re:if they do that

[Anonymous Coward]

don't burn all your MIPS waiting.

Re:if they do that

[Anonymous Coward]

This gets my thumbs up.

Re:if they do that

[Anonymous Coward]

If X86 dies (And it won't. AMD and Intel both have lots to lose, though AMD more than Intel here), Microsoft will port over to PPC (Or whatever), throw on an emulation layer, and probably take the opportunity to break a whole bunch of crappy stuff in Windows that's maintained simply for backwards compatibility.

I don't think you understand Microsofts situation. Their only ace at the moment is that being a Microsoft House is (seen as) the safe choice: things generally work...

If MS customers need to start worrying whether that 3rd party app works on your platform or if the emulation is fast or complete enough or if MS decided to break some backwards compatibility that their in house apps relied on... what then is the point of paying for Windows?

Microsoft do not want to end up in a situation where people draft comparisons between Linux+Wine and Windows 2010, trying to see which one emulates Vista better.

My opinion: Microsoft needs x86 more than Intel or AMD.

Re:if they do that

[D4rk Fx]

What's with you people and your MIPSspellings?

Re:if they do that

[TheRaven64]

Yes, sorry. The rest of the comment was written with the correct values in mind.

Re:if they do that

[Iloinen Lohikrme]

Few notes about the options in hand...

1. Sell itself to Intel, doesn't work as the competition authorities, either from the US or EU will prevent it. There is also an possibility that if Intel would become the only sole provider of x86 chips and have a monopoly on the markets, competition authorities could brake up the company.
2. Won't work if they don't get Microsoft behind their backs. Essentially Microsoft would need to make Windows hardware independent, which would benefit them as it would decrease the relative power of Intel, but then again transforming Windows and making all the work would take too long and take up too much money.
3. Again, without Windows they are dead in the water.

My own estimate is that they will make a complaint to the competition authorities of both the EU and the US on Intel levering its intellectual property to gain a complete market dominance. The EU has previously already ordered Microsoft to openly licence its protocols thus it could be probable that they could order Intel to openly license their intellectual property concerning x86 with a fair price. That is something that probably both Intel and AMD won't want, having a duopoly is much more profitable.

Re:if they do that

[marcosdumay]

"...the majority of the world is still running 32bit OS..."

Yeah, but Intel would lose the entire server market.

Re:if they do that

[marcosdumay]

"Don't most people running Vista already run the 64 bit version?"

No. 32 bits Vista is more stable (DRM stuf and the old closed source impatibilities), so many people prefer it. My bet is that MS won't be able to transition to 64 bits, ever.

Re:if they do that

[mdm-adph]

And so can all the other 200 or so ARM licensees, I think? ;)

Re:if they do that

[TheRaven64]

The MIPS systems that ran Windows NT were around 50MHz. Modern MIPS chips, even the kind found in very cheap laptops, run at around ten times this speed. They aren't the fastest chips on the market, but they should be fast enough for a lot of general use.

Re:if they do that

[TheRaven64]

Maybe, but the i.MX515 (from Freescale) looks like a very strong contender, giving better performance at a lower price than the OMAP3. It really depends on how quickly TI can get the next generation of OMAP3 and then the OMAP4 out of the door.

Re:if they do that

[MightyYar]

While I obviously prefer open source, I buy closed source for a few reasons:
1. There's no real open-source competitor (Photoshop, iMovie)
2. If the closed-source application uses open standards
3. The nature of the project is transient and it doesn't matter if the parent application becomes unusable.

Plus, in reality emulation fills in the gaps. I have VMs that can run my old Mac stuff if I have to recover some old MacWrite document that I wrote in college.

Re:if they do that

[mmkkbb]

Sure, but this where .NET comes in handy. ARM doesn't have to translate .NET binaries since they run on a VM.

Re:if they do that

[Anonymous Coward]

Before the XBox 360 came out, the development environment that Microsoft was supplying was Windows ported to Power Mac hardware (G5 I believe).

For the common users Windows is its applications, a powerpc Windows would be like Windows CE: a non-Windows (user-wise).

Re:if they do that

[MightyYar]

We agree. If you read the post I was replying to, he was making the claim that MS would simply port Windows and move on with life. I was countering that porting Windows isn't the hard part, getting everyone moved over with all of their old applications is.

You took me out of context :)

Re:if they do that

[maxwell demon]

Anything of 16 bit real mode is so old that I doubt that there's still any patent on it. So there wouldn't be any problem in keeping that (this might be important in order to use existing BIOS, which usually is entirely 16 bit real mode). Just add the possibility to switch to 64 bit protected mode directly from 16 bit real mode, without going through 32 bit first.

I think what could safely be removed in very short time is 16 bit protected mode (does anyone still use that?). Also 16 bit virtual mode could probably go quite soon. With 32 bit mode, I don't think it could already be removed. Possibly in one or two years.

Removing support for all those old modes should also free die space for new stuff (extended precision sse maybe, then also the x87 stuff could finally be removed).

Re:if they do that

[Volante3192]

Until someone challenges it in court, OEM copy = system it was sold with, cause if you get raided by the BSA that's how they'll see it.

Re:if they do that

[Chaset]

From what I've read of history, AMD had just such a feature in the 5x86 (late 486~early pentium days). Of course, nobody used that mode because it wasn't x86. I don't know whether people would be more willing to use such a feature now. Linux would probably be ported to it.

Re:if they do that

[mzs]

"IIRC, during the x86's long development (or mutation), Intel added some features because they could, and because it seemed like a good idea at the time. Some of those features were never much used, or turned out to be not such a good idea, or were rendered mostly irrelevant by design changes in the next generation."

See the bound instruction. The only reason I know about it is because I jumped to it once after some stack corruption. That was definitely a pie in the sky instruction created by the computer scientists instead of the engineers.

Re:if they do that

[downix]

been 8 years so far, haven't thrown my hands up yet.  Of course, being able to throw a dozen CPU's at emulating a single one does help, a lot.

Re:if they do that

[cbiltcliffe]

Ah. Ok.
My first computer only came with DOS 6.0, so I'm not quite as "getoffmylawn" as you, by the looks of your UID.

Re:if they do that

[home-electro.com]

Crusoe AFAIK did not support native RISC (it was not even a RISC processor, it was VLIW inside), only IA32. And a crappy one, for that matter.

Re:if they do that

[DrSkwid]

Strange, I use these every day :

XP 32 & 64bit, Linux, Symbian, OSX ppc & Intel, FreeBSD, Plan 9, whatever is on my mp3 player, Win2k, OpenBSD, whatever powers my HP Laserjet :)

Re:if they do that

[bhtooefr]

Except Itanium also had serious performance issues in the real world.

Re:if they do that

[bhtooefr]

IIRC, Intel still sells the XScale IOPs... and it was only ~6 months ago that a desktop based on an Intel XScale IOP was discontinued.

Re:if they do that

[bhtooefr]

But if you use the new Slashdot comments page, it automoderates. As in, as soon as you select.

(I don't use the comments page, though.)

Re:if they do that

[gd2shoe]

You say "many people" but not "most people". Most people will use what the computer ships with, and I'd be surprised if OEMs don't ship the 64 bit version almost exclusively.

MS will make the transition to full 64 bit. It's only a matter of time.

Re:if they do that

[toddestan]

They lost backwards compatibility with anything about 5 years or older at the time of the switch, as Classic never made it to Intel.

Re:if they do that

[home-electro.com]

The migration to 64 bit still isn't done, and for the vast majority of applications still isn't necessary.

It probably will never be necessary for most applications. 64 bit is unacceptable on laptops market to due to much higher power requirements.

Re:if they do that

[home-electro.com]

But who really needs a speed of i7? PC's performance exceed needs of an average user by at least an order of magnitude....

Re:if they do that

[SebaSOFT]

In addition to that, Linux distros for ARM like Maemo (http://www.maemo.org) have successfully made gcc paths and sanboxes (scratchbox) to compile in an X86 to arm, so you can develop in x86 and safelly compile it in ARM just changing the (I think it's called) path.

How far is Microsoft of that? I really don't know how integrated are the tools for WIndows mobile, but dude, Maemo is a full blown Linux and it's X server, you don't have to change anything from destop to palmtop (not really).

Re:if they do that

[kasperd]

Anything of 16 bit real mode is so old that I doubt that there's still any patent on it. So there wouldn't be any problem in keeping that (this might be important in order to use existing BIOS, which usually is entirely 16 bit real mode). Just add the possibility to switch to 64 bit protected mode directly from 16 bit real mode, without going through 32 bit first.

Good point, that might actually be an easier path forward.

I think what could safely be removed in very short time is 16 bit protected mode (does anyone still use that?).

What exactly do you mean with that? The first protected mode Intel introduced was on the 286 and was a strange hybrid of 16 and 24 bits. With the 386 they introduced an extension of that, which allowed for 32 bit code. They extended the existing segment descriptors by using free bits in various places to support larger address space. The result was a bit messy, but completely backward compatible with 286 protected mode. In other words, the 386 only had one protected mode in which you could have some 16 bit segments and some 32 bit segments. With the 386 they also introduced paging. For a large part it was independent from the segmentation model. The result was, that you could run protected mode without making use of paging at all, but the other way around was not supported.

Also 16 bit virtual mode could probably go quite soon. With 32 bit mode, I don't think it could already be removed. Possibly in one or two years.

Do you mean virtual 86 mode could go? It doesn't exist in the 64 bit protected mode, only in the 32 bit protected mode. Notice that unlike the extension from 16 to 32 bit, the 64 bit was a completely new mode, that allowed them to throw away things they didn't like such as virtual 86 mode.

Removing support for all those old modes should also free die space for new stuff (extended precision sse maybe, then also the x87 stuff could finally be removed).

AFAIK they don't spend a lot of die space on older modes. I don't remember who told me this, but at some point I was told that the real mode is essentially implemented by just taking a very old and proven design and sticking in a small corner of the CPU. Since modern CPUs are made with much denser technology, it doesn't take up any significant die space, but it isn't optimized as much for performance as the newer modes that people use for everything important nowadays. Besides, most of the die space is actually used for L2/L3 caches, which are independent from the mode the CPU is operating in.

Re:if they do that

[kasperd]

Translation and register renaming take up tiny amounts of die compared to the instruction cache savings of x86.

Only L1 cache actually make a distinction between instruction cache and data cache. And AFAIK the instruction cache actually stores the translated instructions, so there isn't going to be any savings from x86 code being more dense. But maybe you meant that the space you save for x86 instructions stored in L2 and L3 cache are worth more than the overhead of translation.

MAD

[Anonymous Coward]

At stake is not only AMD's ability to build processors that use Intel's x86 technology, but also Intel's ability to use AMD's x86-64 tech in its CPUs.

So, surely this is a case of mutually assured destruction for both isn't?

Re:MAD

[DragonWriter]

So, surely this is a case of mutually assured destruction for both isn't?

Something like that, but not perfectly symmetrically. While x86_64 is well-enough established that it would be inconvenient for Intel to have to go back to x86 and build a new, non-derivative extension with similar capabilities, it would be less of a problem for them than AMD losing the rights to use x86-anything.

Given that Intel and AMD don't have serious competition for desktop PCs right now, its possible that the a result that hurts Intel a lot but AMD more in the short term could benefit Intel in the long term, though really the intent here is almost certainly to get concessions from AMD on the basis that Intel may be able to prevail in court, and AMD stands to lose more if the agreement is terminated; it is extremely unlikely that Intel's goal is to terminate the agreement.

Re:MAD

[powerlord]

Something like that, but not perfectly symmetrically. While x86_64 is well-enough established that it would be inconvenient for Intel to have to go back to x86 and build a new, non-derivative extension with similar capabilities, it would be less of a problem for them than AMD losing the rights to use x86-anything.

Like heck.

It would force both AMD and Intel to pull their chips temporarily.

The only thing Intel could sell is the Atom (32 bit only), and the original version of the Core (again, 32 bit only).

Yeah ... they'd still have a product to sell on the market, but a staggering amount of their products (most of the Core line) would simply stop.

Likewise AMD would still have the Geode and other chips to sell, but their desktop/server line would have to stop.

MS would probably continue okay (I hear Win7 runs okay on the Atom and old Core processors), but it would mean that we'd be back at 32bit limits for things like memory.

The groups that would be hurt the most (beyond AMD and Intel)? Computer retailers like Dell and Apple (whose products would have to be redesigned), and the American computer economy as a whole (I'm sure you'll be able to find AMD and Intel chips made in China that would keep shuffling off the assembly line just fine).

Re:MAD

[Mr_Magick]

If Intel pulls AMD's license AMD is all but destroyed. However, I see a huge monopoly case being brought to court very quickly after any such turn of events.

Intel has no true competition in the x86 market other then AMD. A point that no small amount of layers are more then willing to point out to the right judge.

Re:MAD

[edmudama]

What would a monopoly case have to do with it?

Just because you're the little guy who's facing a much larger competitor, doesn't give you the freedom to break licensing agreements.

Re:MAD

[DragonWriter]

Like heck.

Nothing that you say contradicts what I stated, so I'd like to understand where your disagreement is. As I said, this would be very much like mutually assured destruction, but not precisely symmetrical, as AMD would be hurt more than Intel.

It would force both AMD and Intel to pull their chips temporarily.

Most of their CPUs, and even moreso most of those in the main consumer PC market, certainly.

The only thing Intel could sell is the Atom (32 bit only), and the original version of the Core (again, 32 bit only).

And Itanium.

Likewise AMD would still have the Geode and other chips to sell, but their desktop/server line would have to stop.

While Intel would still have a desktop/server line to sell, one that is losing ground to x86-64 more than its other competition.

MS would probably continue okay (I hear Win7 runs okay on the Atom and old Core processors), but it would mean that we'd be back at 32bit limits for things like memory.

Windows Server already supports Itanium, so if x86-64 was cleared out of the way and Intel returned to its original hope of moving Itanium into consumer products, it probably wouldn't take much for Microsoft to have its consumer-focussed Windows products supported on Itanium, too.

The groups that would be hurt the most (beyond AMD and Intel)? Computer retailers like Dell and Apple (whose products would have to be redesigned)

OTOH, those computer manufacturers that currently produce Itanium systems, while they would lose any x86-64 systems they have, would suddenly face a less crowded market.

Re:MAD

[DragonWriter]

Intel has no true competition in the x86 market other then AMD.

That's rather the point of a patent: no patent holder has competition in the market for things directly within the coverage of the patent. Patents are intended as government granted monopolies. That's the whole point. Intel has a monopoly there whether or not they have licensed other people to use the technology on which they have a monopoly.

But "x86" isn't the kind of market anti-trust analysis would usually focus on. Desktop PC CPUs might be, but without x86-64s, which are a big chunk of the current mainstream market there, I'm not sure that Intel would have a monopoly there. Sure, they've got Itanium and their (currently low-end) 32-bit x86 offerings, and AMD has its low-end non-x86 offerings. Taking out x86-64 would be disruptive, and AMD would almost certainly lose marketshare from it, but its not at all clear that Intel would win, or even have as dominant a position as it has now where, depending on exactly where you draw the line, you arguably have Intel and Intel-licensed products dominating the entire market.

Re:MAD

[Anonymous Coward]

Intel won't just lose their x86-64 agreement. All those neat features like 3dnow and SSE are cross licensed too. Intel would have to go back to their Pentium 1 or Pentium 2 lines to continue using stuff that's missing AMD tech.

Re:MAD

[DigiShaman]

In short, the entire industry would be F.U.C.K.E.D!

I'm willing to bet Washington DC would pull both CEO's behind close doors and slap some sense into them. As far as they are concerned, keeping this high-tech industry from imploding is for the sake of national security.

Re:MAD

[seeker_1us]

The only thing Intel could sell is the Atom (32 bit only), and the original version of the Core (again, 32 bit only).

Or they could bring out a 45 nm version of Itanium2.

Re:MAD

[Trepidity]

There are other aspects to the asymmetry, though, such as Intel's problematic antitrust position. If they're seen as using their patents to attempt to clear the market of their only major competitor, those patents may be seized from them.

Re:MAD

[powerlord]

I'm sorry, I reread your message and you're right, nothing I said contradicted you, but I feel like Intel losing x86_64 would set them back much more than saying "its not exactly symmetrical" implies.

I think I just felt that even if it is not perfectly symmetrical destruction to AMD and Intel (and we haven't talked about AMDs embedded chip work which probably brings in a $$$ or two and would also go unhindered), I think the fallout to the whole industry would make that lack of symmetry irrelevant.

The ONLY upshot I could see if something like this happened would be an immediate uptick in the adoption of ARM based Linux platforms since Linux would already be in a position to exploit an ARM laptop/desktop market, while everyone else (OSX, Windows, in this context), would have much more work on cross-compilation to do.

Re:MAD

[powerlord]

They could, but that would take design time, and development time. The designs I listed should be "ready to go" with only minor modification (if needed), before they could be put into production to fill the hypothetical "processor gap".

Well this should be fun to watch

[Anonymous Coward]

This has been brewing for years. AMD with it's anti-competitive lawsuit against Intel, and now this. AMD's suit is mostly won, but Intels new suit could really make things interesting.

AMD's next line of Phenom II are coming out soon and AMD doing better in terms of sales. Intels feeling the pinch from Netbook sales pulling out the rug from the I7's anticipated sales. The market is changing and it favors AMD in the terms that people are spending less. Intel has a lot more to loose then AMD and that's why this is going to be so good.

If Intel wins the consumer will lose, if AMD holds its ground Intel will suffer a large drop in sales and the giant of the company will fall. Any sort of drop in sales from Intel and it will have to make major cutbacks and Intel will loose all sorts of momentum just to save it's cash. The middle ground is what we all will hope for but even that could really hurt the Intel giant.

Re:Well this should be fun to watch

[TimothyDavis]

Intel has a lot more to loose then AMD and that's why....

If Intel wins the consumer will lose, if AMD holds its ground....

....make major cutbacks and Intel will loose all sorts of momentum just to save....

At least you got one of them right. 33% is still a failing grade, though.

Re:Well this should be fun to watch

[Eskarel]

Intel is unlikely to fall. They're certainly hurting in that their serious dominance is in the top end of the market and the economy is shrinking that market segment. AMD seems to be surviving, at least in the short term, and they're still reasonably competetive at the lower price points(I haven't looked in a while, but they had nothing that even came close to i7).

On the other hand, unlike AMD, Intel still has loads cash, loads of market share, is an equal competitor at the lower price points and isn't losing any money on their high end systems(today's high end is tomorrow's consumer grade).

Intel indeed has more to lose than AMD, but that's because if AMD loses much more they won't exist anymore. They've already had to spin off their fabrication line(which caused the lawsuit) so that they can try to escape toxic debt levels, and Phenom, while a reasonable chipset didn't even make it into the high end range. They're also worth substantially less than they paid for ATI, and I haven't seen them picking up all that much market share in that sector either.

Re:Well this should be fun to watch

[Klintus Fang]

on what basis do you think people wanting to spend less in the current economy hurts Intel more than AMD? Intel's chips do tend to be more expensive because people are still willing, for the most part, to pay more for them. Intel charges higher prices on their CPUs because they can. Not because they need to.

But you need to keep in mind: Intel is still a process generation a head of AMD and AMD's recent spinoff of their foundry is not going to help AMD in that regard. And in the end, the cost of manufacturing a CPU decreases with each advance in process generation.

If the amount of $ customers are willing to spend on their CPU's drops as the economy grinds down, both companies will take a hit on their profit margins, but Intel can afford to lower prices further than AMD can afford to do, so long as AMD continues to be a lagging behind Intel by a process generation.

That scenario hurts AMD more.

and the answer is...

[je ne sais quoi]

And the answer is... powerPC! But only if someone takes an interest in working on the chip to lower power consumption and heat output. My dual G5 runs great but the sucker sounds like a jet engine taking off when it starts doing something computationally intensive.

Re:and the answer is...

[the linux geek]

I recently acquired a PowerStation made by FixStar, the same people who make Yellow Dog Linux. Since Apple and IBM gave up on Power-based workstations, this is among the last you can get, and it's quite nifty, and fairly reasonably priced ($1250 for quad PPC, 2GB RAM). If you have $6000 extra to throw at it there's also a Cell expansion board. My only real issue with it is lack of compatible 3d graphics hardware.

Re:and the answer is...

[Darkness404]

The problem with PPC, is, they just aren't fast, I'm quite certain that your PPC workstation would end up seeming slower than a sub-$1000 x86 Core i7 box (http://www.newegg.com/Product/Product.aspx?Item=N82E16883229072). Yes, PPC does scale quite nicely with multiple CPU configurations, but it is still very expensive for the speed. It similarly didn't help that Apple added the usual "Mac markup" with their hardware, and there are (comparatively) very few PPC motherboards and other hardware support. But I just can't see PPC ending up making a dent in x86 even if AMD starts making PPC chips, it didn't work for Apple and I don't think it will work for AMD because they simply are not marketable.

Re:and the answer is...

[je ne sais quoi]

Actually, they're not bad for just clock speed. The quote you gave was for a quad-core 2.66 GHz chip. My dual G5 has 2.7 GHz chips in it and it's 3.5 years old, I'm sure you could cram four G5s onto a motherboard and compete with a quad-core chip just fine. 2.7 GHz is still a relatively fast chip, just a quick cursory glance at newegg shows that there's not much over 3.33 GHz for sale right now.

Granted, nobody has made a multi-core powerPC chip as far as I know, but I argue that's not the issue, nor is the clock speed though, it's the power consumption. You just can't put a 2.7 GHz G5 in a laptop and expect it to not melt the thing into a gooey flaming mess, or even if you did manage to cool it, you'd never get any kind of decent battery life. I suspect that the laptop problem more than anything caused apple to switch to intel, but the fan noise necessary to keep the G5s cool even on the desktop was becoming an issue as well. However, power consumption and heat have nothing to do with the architecture of the chip, even making x86 chips AMD has been running into problems trying to make them low power and cool.

Re:and the answer is...

[je ne sais quoi]

Sorry to respond to myself, but I just learned about a power architecture chip being made by PA Semi. Apparently, it IS a low power consumption 64 bit chip that could be put into a laptop and it already runs at 2 GHz (or did two years ago). It consumes 7 watts whereas the core duo consumed 21-25 watts. According to this, PA Semi thought it was all but in the bag for apple to use their chips to replace the G5, but apple switched to intel for some unknown reasons.

I think a switch from the x86 would be pretty minimal in the long run, there's plenty of alternative chips out there, they just don't have the resources put into them like the x86 and x86_64.

Re:and the answer is...

[Anonymous Coward]

a Cell expansion board. My only real issue with it is lack of compatible 3d graphics hardware.

This is a joke but...The answer is obvious!

Re:and the answer is...

[Doctor Faustus]

nobody has made a multi-core powerPC chip as far as I know
Isn't that what's in a PlayStation 3?

What an opportunity . . . for lawyers!

[MarkvW]

Complicated factual and legal scenarios, deep pockets, core business problems . . . This is a dream scenario--for the lawyers.

The lawyers will be able to earn so much out of this mess!

. . . that's why this tempest ought to blow over pretty soon. T

Business as usual

[AKAImBatman]

Intel and AMD like to squabble about licensing every few years. Probably in an attempt to broker a deal that is even more favorable than the last. They usually spend some time posturing in court, bare their claws a little, then settle with a new cross-licensing agreement. If Intel gets too pushy, the feds start staring at them REALLY hard. Which tends to make Intel fall in line.

Strictly speaking, Intel's argument is pointless. Yes, their deal is with AMD. But AMD's foundry only manufactures the chips, it does not design them. (Unless I somehow misunderstood their fabless plan.) Since the fab creates the chips on behalf of AMD, the licensing is not violated.

That's my 2 cents worth, anyway. I'm not a lawyer, but I doubt one would make many more comments without viewing the legalese between the two companies.

Re:Business as usual

[Chris Burke]

Strictly speaking, Intel's argument is pointless. Yes, their deal is with AMD. But AMD's foundry only manufactures the chips, it does not design them. (Unless I somehow misunderstood their fabless plan.) Since the fab creates the chips on behalf of AMD, the licensing is not violated.

It may not be that easy. The Intel/AMD license agreement, for all its notoriety, is completely confidential and thus nobody knows exactly what is in it except for a small number of people at both companies. Despite that, it has long been suspected that part of the agreement is that AMD would not manufacture more than a certain % of its chips at a 3rd party fab, which FoundryCo -- wait, it's GlobalFoundries now, slightly less stupid name -- would almost certainly count as once fully spun off.

Strictly speaking, though, nobody outside the upper echelons knows. The only thing I'm 100% certain of is that AMD thought about the cross-licensing agreement when they came up with the idea for spinning off the fabs, and would not have done it if they thought it would cost them their license. But of course companies can differ in their self-serving legal reasoning, and who knows maybe they knew they were taking a chance and felt that the global anti-trust inquiries and the threat of losing AMD64 licensing would keep Intel playing ball?

Re:Business as usual

[Anna Merikin]

Do you remember the old 286 and 386 math co-processor chips makers like AMD and C&T marketed? If push should come to shove over this (which I wholeheartedly doubt b'cause both have too much to lose) you might see motherboard makers including a socket for an AMD 64-bit/GPU chip, which will be an add-on to an Intel x86 32-bit CPU.

But neither company is so stupid as to muddy the water exactly when Nvidia is looking for a way to make chips with or without Intel's permission.

Re:Business as usual

[Anonymous Coward]

And, if Intel continues to behave as it has in the past, it will continue to plagarize ideas from AMD and anyone else it can, like it did to DEC (e.g., the branch prediction stuff originally designed for the VAX 8600 that Intel swiped and then licensed from DEC later on when it got caught).

Re:Business as usual

[0xABADC0DA]

If Intel gets too pushy, the feds start staring at them REALLY hard. Which tends to make Intel fall in line.

One remedy used in the past for monopolies is to take it's patents and trade secrets and place them in the public domain. Even if Intel were to win a complete victory, they could end up losing it all.

Re:Business as usual

[rsmith-mac]

The only thing I'm 100% certain of is that AMD thought about the cross-licensing agreement when they came up with the idea for spinning off the fabs, and would not have done it if they thought it would cost them their license

I'm not so sure that was the case. The existing debt was crushing AMD, and the prospects of the costs needed to do yet another fab build-out would have further compounded that problem. Unless the x86 market were to pick up tomorrow and everyone bought AMD's highest margin chips, their options would have been bankruptcy or ceasing to invest in fabs (and there-by completely losing to Intel over the next few years).

AMD didn't have a choice, spinning off their fabs in order to secure billions of USD from ATIC was the only way they were going to stay afloat.

Re:Business as usual

[Johnny Mnemonic]

Or, move to a more friendly country. Think about Intel having a monopoly, and then moving to Japan, for instance.

Re:Business as usual

[Klintus Fang]

I personally suspect that what Intel wants to get out of this is some clear understanding that non-AMD customers who might start using AMD's spun-off FABs (which is certainly going to start happening at some point) do not get to benefit from the cross licensing agreement that Intel has with AMD. I'm not a lawyer, and haven't seen the licensing agreements either, but I suspect that what Intel is trying to avoid is having AMD's old fabs thinking they are a party in the licence agreements and thereby becoming a back door for other companies to get in on the agreement.

Re:Business as usual

[Kalriath]

Seriously, when has this ever happened? Sounds more like a pipe dream to me.

Re:Business as usual

[Chris Burke]

AMD didn't have a choice, spinning off their fabs in order to secure billions of USD from ATIC was the only way they were going to stay afloat.

You make good point. I will only add that AMD could operate for quite some time at the current burn rate, but that this leaves them only able to compete with Intel at all for a couple years without finding a way to finance new fabs and spinning of the fabs was an excellent way to do this. With the credit crisis that occurred not long after they sealed the deal, this seems doubly wise. Yet on other hand, losing the ability to make x86 chips would also doom their ability to compete at all and they'd last for as long as they could prevent a legal injunction. So unless they have a decent legal argument and can hash out a deal with Intel, it's going to hurt either way.

Fuzzy on x86 IP

[MSTCrow5429]

Maybe I'm missing something, but how can the x86 architecture itself be subject to copyright? Isn't the protected property not the publicly documented instruction set, but the implementation thereof?

Re:Fuzzy on x86 IP

[gEvil (beta)]

Maybe I'm missing something, but how can the x86 architecture itself be subject to copyright? Isn't the protected property not the publicly documented instruction set, but the implementation thereof?

I believe it's not the core x86 instructions, but rather all the various MMX and SSE extensions that have been tacked on in the past 10-15 years. And as mentioned in the summary, AMD's x64 extensions are at stake, too.

Re:Fuzzy on x86 IP

[Anonymous Coward]

Ever heard of patents?

Re:Fuzzy on x86 IP

[MikeBabcock]

Intel has a lot of patents that essentially cover the implementation of x86 in any meaningful way. The way x86-64 was strapped onto x86 is on the other hand owned by AMD.

They play nicely, and they both get to use each others' innovations.

Re:Fuzzy on x86 IP

[idontgno]

Not copyright. Patents..

In other words, Intel claims patents over much of the technology that makes an x86 an x86, and AMD agreed (back in 2001--the patent cross-licensing agreement that's in dispute in this issue). AMD could hardly walk away from the agreement now* and continue to manufacture x86-descended CPUs--their previous acceptance of the patents would be evidence against them in Intel's inevitable patent infringement suit.

No, I Am Not A Lawyer. And I'm sure it's nuanced much more finely than this. But that's kinda the Sesame Street version of how this is shaping up.

Patents.

* Yes, I know, AMD isn't disclaiming the license agreement; they're saying the new Globalfoundaries fab has rights to those licenses because it's an AMD subsidiary; Intel's saying they aren't and therefore don't inherit the licenses. If it becomes a full-out patent lawsuit nuclear exchange, AMD might be in a position of manufacturing x86s without license, which would be bad, or not manufacturing x86s at all, which would be worse, or not allowing Globalfoundaries to manufacture x86s, which would be stupid.

Re:Fuzzy on x86 IP

[forkazoo]

Maybe I'm missing something, but how can the x86 architecture itself be subject to copyright? Isn't the protected property not the publicly documented instruction set, but the implementation thereof?

My understanding is that if you wanted to make a 286 clone designed from scratch, you would probably be in the clear. OTOH, if you want modern extensions like MMX, or even SSE/AMD64, then you need a license for the more modern variations. That said, the whole field is deeply complicated and unclear. Some parts of the situation have never really been tested in court so anybody claiming to have a 100% understanding of the legal issues is almost certainly mistaken.

Re:Fuzzy on x86 IP

[blair1q]

The original license was for the original microcodes that AMD was given as part of AMD and Intel's original second-source partnership, and for all derivatives AMD made from that.

Since all AMD x86 designs are tainted by that original copyright, even if AMD no longer uses any of the original circuits or microcode, AMD can never make an x86 of any kind unless they pay Intel and get Intel's permission.

More...

[Jaysyn]

... stupid intellectual property bullshit.

Re:More...

[Anonymous Coward]

Man, such a crude comment.

yet so true.

posting AC for mod points.

What's really at stake

[davidwr]

At stake is not only AMD's ability to build processors that use Intel's x86 technology, but also Intel's ability to use AMD's x86-64 tech in its CPUs."

At stake is money and corporate posturing.

This is just another day of corporate King Of The Hill.

The ideal scenario would be if...

[Anonymous Coward]

...the problem of inefficiency that is the x86 ISA just solved itself from this. Hello, PowerPC. Hello, ARM.

Re:The ideal scenario would be if...

[RightSaidFred99]

A better scenario, even, is if people would quit pretending there's anything wrong with x86.

Re:The ideal scenario would be if...

[Anonymous Coward]

Try that with someone else than those who have actually done assembler on these architectures and know what they are about in machine code and compiled form from any high level language, will you?

Intel will license it

[truthsearch]

Intel will definitely work this out. They're almost forced to license x86 to prevent being labeled a monopoly. Many believe the only reason they licensed it in the first place was to prevent legal action by the justice department. With a competitor making similar chips it's hard to claim they strong-arm computer manufacturers into using their products.

Re:Intel will license it

[hemp]

The real reason for the licensing had nothing to do with the Judicial system.

In order to bid on certain government/DOD contracts you are required to have a second source for most items. This to prevent all of the usual issues you normally get when dealing with a single source, namely they go out of business and you can't find them any more.

By allowing AMD to license and manufacture, Intel was able to bid on more government contracts. This all occurred back in the 80's prior to Intel dominating the CPU field.

Re:Intel will license it

[Lost Race]

Many believe the only reason they licensed it in the first place was to prevent legal action by the justice department.

They licensed long before there was any danger of monopoly accusations. The reason was that certain large customers would not commit to any long term purchase contracts without an available second source (in case the primary source went out of business or decided to stop production of the item). AMD et al were basically a checkbox on the requirements list to make a few big sales.

Re:Intel will license it

[blair1q]

You are apparently ignorant of history.

Intel licensed x86 to AMD originally because Intel was unable to keep up with demand.

When Intel decided to end that relationship, AMD refused to stop making x86's, and sued Intel to keep the right to do so.

AMD actually LOST that case, but AMD and Intel were told by the courts to make a license that worked, and AMD was forced to pay Intel for court costs. They renewed the license in 2001.

AMD has now breached the license. Intel has no responsibility to keep AMD in business. Intel can get another foundry to make x86 CPUs. There's no law against being a monopoly.

Natural law is against being a failure like AMD.

Re:Intel will license it

[DragonWriter]

AMD has now breached the license.

The license is confidential, so what we know is that Intel claims AMD has breached the license, and AMD claims they haven't. Since we have two interested parties each making claims that both reflect their own natural interest, and don't have the essential facts (like, the contents of the agreement) against which to evaluated those self-interested claims, we really can't say much about whether or not the license has been breached.

Re:Intel will license it

[eggz128]

Intel licensed x86 to AMD originally because Intel was unable to keep up with demand.

It wasn't so much that Intel couldn't keep up with demand, more that IBM's policy required that a second source be available just in case they couldn't.

AMD has now breached the license. Intel has no responsibility to keep AMD in business. Intel can get another foundry to make x86 CPUs. There's no law against being a monopoly.

No, there is no law against being a monopoly. There are laws against being an abusive monopoly however. Intel has been convicted of abusing it's monopoly status in Japan, has at least been accused of doing so in the EU. Maybe AMD could file a complaint in the USA also and have it successfully investigated. Once convicted of being an abusive monopoly the rules change.

Natural law is against being a failure like AMD.

In theory the UK monarch can veto any law parliament puts before him or her. In practice, vetoing rarely happens as it can lead to the removal of the monarchs head. Intel should be careful just how far they push this as states could just decide they are abusing their position and remove their right to x86 all together.

Re:Intel will license it

[Anonymous Coward]

"There's no law against being a monopoly."

True. However, there are laws against leveraging your position as a monopoly to prevent competition.

That's because the "natural law" would not lead to outcomes society as a whole would like and society makes its own laws (in democracies at least).

For example if my company has shitloads of money in reserve and my smaller, or upstart, competitors don't, and I sell my product at a heavy loss until I have driven the other companies out of business, then in Germany I would have violated anti trust law.

This is kind of the situation here. Ending the licensing deal would severly hurt both but AMD would go out of business long before Intel. Whether this would violate the law or not is entirely unclear and potentially differs between the US and Europe (where GlobalFoundry has several fabs).

Re:Intel will license it

[rts008]

In order to bid on certain government/DOD contracts you are required to have a second source for most items.

This is an important factor here...excellent point!*

The only way I can see this becoming more than just 'typical competitors maneuvering' would be if Via's chips (like the Nano) satisfied that requirement.

I, for one, just do not have a clue.
I also think it's important to look at the big picture here, instead of just two of the dogs circling a pile of bones, growling at each other.(*thus my first sentence)

Re:Intel will license it

[sjames]

AMD is alleged to have breeched the license. Of course, Intel would REALLY rather not cancel it since AMD will otherwise cancel the license of x86_64 to Intel.

Re:Intel will license it

[blair1q]

The license is confidential, so what we know is that Intel claims AMD has breached the license, and AMD claims they haven't.

Intel asked AMD a couple of months ago to agree to make the license public so that this would not be ambiguous to anyone.

AMD refused.

Intel knows what is in the license and wants you to know. AMD knows what is in the license and does not want you to know.

There is enough that is public to show that GlobalFoundries is not a subsidiary of AMD. The confidential portion would have to do with the percentage of x86 CPUs AMD can outsource to companies that are not subsidiaries. That is why Intel says the breached portion is confidential.

AMD has been a bad actor for almost its entire relationship with Intel and x86. The GlobalFoundries spinoff exists only to (a) placate the Arabs AMD screwed a year before in a bad stock issue that tanked; and (b) keep the foundries from being seized by creditors when the rest of AMD goes bankrupt.

You can bet when the creditors show up AMD will be showing them documents stating that GlobalFoundries is not a subsidiary but a fully severed company in which AMD is merely a minority shareholder.

Re:Intel will license it

[blair1q]

Intel wasn't actually convicted of abusing its monopoly status. It wasn't a monopoly. And it wasn't convicted.

It settled with an economic commission (none of these things are courts) and at that point decided it was cheaper to pay the fine (less than $50 million; about an hour's pay to Intel) than to fight it in a court.

In the settlement Intel admits no wrongdoing, and the Japanese assert none.

Re:Intel will license it

[Plasmoid2000ad]

Well your almost right on the original license. In order for the new to cpu's and mostly unheard of Intel to supply its fancy new design chip to IBM, IBM wanted a backup supplier (like the airline industry demands two engine suppliers for airplanes). Intel Contracted AMD and gave them the license. This was grand from 1982 - 1986 ... or from 8088/8086 up to the 286. When the 386 rolled around IBM was more or less gone, and Intel was looking to the Dell and HP style OEM's. It tried to cut AMD but AMD sued AND WON. but it didn't happen until 1994, so how did AMD keep going? While x86 use was in limbo, and Intel weren't sharing AMD reverse engineered the 386, the 486 and even the original Pentium. At this time they managed, with no further help from Intel, to reverse engineer their chips to the point where their CPU's were considered MORE compatible with Intel chipsets then Intel's own (quite and achievement i think) No one knows exactly what was agreed in 1994 when Intel was forced by court order to make an agreement with AMD, but since then AMD hasn't been making drop in replacments for Intel chips but has been using its own chip design requiring its own motherboards. Assuming the Intel - AMD agreement isn't one sided... how is what AMD is doing any different from what Intel is doing with TSMC to produce Atom chips for it (is it just that the current low end Atom has no x86-64 extensions?) That aside... do the same circumstances exist that existed in 1994 that Intel has to, under court order, make an agreement with AMD which allows AMD have a license to x86. I reckon it must. Intel wasn't half the monopoly it is now back then.

Re:Intel will license it

[WozNZ]

AMD has now breached the license. Intel has no responsibility to keep AMD in business. Intel can get another foundry to make x86 CPUs. There's no law against being a monopoly.

Natural law is against being a failure like AMD.

Might not be BUT there are MANY MANY laws about abuse of monopoly position as MS and Intel have found out many time. If they do not renew the licence I am sure the EU will have a very EXPENSIVE reason for them to renew :)

Re:Intel will license it

[Dun Malg]

You are apparently ignorant of history.

You apparently can't even be bothered to read the wikipedia entry on AMD.

Intel licensed x86 to AMD originally because Intel was unable to keep up with demand.

AMD was a second source for the 8086 and 8088 because IBM demanded two sources, not because Intel couldn't make enough.

AMD refused to stop making x86's, and sued Intel to keep the right to do so. AMD actually LOST that case,

AMD was the one who challenged the x86 license cancellation and won the case in arbitration, and after numerous appeals it was upheld b the California Supreme Court.

They renewed the license in 2001. AMD has now breached the license.

Given that the licensing agreement isn't public, your analysis is clearly pulled straight from your rectum.

Intel has no responsibility to keep AMD in business.

The amusing thing about cross licensing agreements is that they cross. You can't really cancel half a contract. If Intel forces AMD out of the x86 CPU market... then Intel is out of it too, unless they intend to use something other than EMT64, which is a licensed implementation of AMD's proprietary AMD64.

Natural law is against being a failure like AMD.

Oh, I see. Your an Intel fanboy. That explains it.

who moderated this fool up so high?

Re:Intel will license it

[RzUpAnmsCwrds]

Well as much as you might label AMD a failure, and as much as you would like to believe that Intel can just pull the plug on AMD, there's one answer that proves you wrong on both counts:

AMD64

You can thank the "failure like AMD" for x86-64 and for integrated memory controllers. This at a time when Intel was dicking around with the failure that was Pentium 4.

You can call it IA-32e, EM64T, or Intel 64 if it makes you feel better. But the fact is that AMD64 is here to stay, and it's patented by AMD. Which means that Intel either keeps cross-licensing, or throws out all of its current designs and its current ISA.

So, yeah, let's pretend that Intel and Intel alone is responsible for their current designs. Or we could live in the real world. One where AMD depends on Intel patents and Intel depends on AMD patents. That's why the license was renewed in 2001.

Re:Intel will license it

[Anonymous Coward]

Actually there is a law against being a monopoly. The anti-combine legislation can kick in and force people to split. It didn't happen to mickeysoft, sadly, but it did happen to AT&T. Pretend you are a big-wig at AIG. Pretend you are expecting millions in bonuses for fucking up. Pretend Obama is talking about you and pulling the plug in the same speech.

Re:Intel will license it

[MtViewGuy]

I believe that in the end, Intel and AMD will settle because if Intel tries to force AMD out of the x86 business Intel will face the same government lawsuit that brought down the United Shoe Machinery Company in the 1960's--namely abusing patent and intellectual property laws to protect its monopoly or near-monopoly position.

Re:Intel will license it

[Anonymous Coward]

> You apparently can't even be bothered to read the wikipedia entry on AMD.

I *wrote* the wikipedia entry on AMD.

Yours truly,

Anonymous Coward

Old and busted = Mhz

[QuantumRiff]

New hotness = Lawyers on retainer!

I for one, will miss the Megahertz Myth race.. But hey, it might go crazy when AMD has a GPU as the Vector CPU in the computer, and Intel has to sell a 63-bit processor.

I guess it will be exciting to watch new developments again.. Seems they've gotten a little to comfortable with each others positions lately..

Re:Old and busted = Mhz

[blair1q]

AMD had to push out its GPU-as-CPU plans by two years because apparently $5.4 billion didn't buy them enough technology and talent to get it done.

Re:Old and busted = Mhz

[Anonymous Coward]

Uggh - this is about the instruction set & surrounding patents, not the precise width of the registers. So Intel losing access to the x86-64 patents will be a huge problem (although not anywhere near as big if AMD loses access to the x86 patents).

Simply decreasing the register width won't get you anywhere (except probably incompatibility with existing software & give a headache to all the software developers who now have to recheck their work to see what kind of impact this has). Not to mention that computer engineers like to design things in powers of 2 (just like people who use base 10 like powers of 10). Things come out cleaner.

This won't be exciting - within 6 months you'll hear (or not) about some renewal of the licensing agreement. Things will get exciting if one or the other actually do decide to sue the other.

Poor Microsoft

[zigfreed]

Long live StrongSparcPC_x64! Poor Microsoft, how on earth would they sell Windows 7?

Re:Poor Microsoft

[nonicknameavailable]

with FUD

Re:Poor Microsoft

[Forty Two Tenfold]

Wif fud? Liek a Win7 license in a Happy Meal? I think MacDonald would rather keep the Apple pie license.

It won't succeed

[girlintraining]

If Intel becomes the exclusive provider of x86 chips, they'll be smacked by the government with anti-trust litigation (Note: I did not say WHICH government, my fellow silly Americans). It was the same with Apple being the company Microsoft pointed to when it was hit with anti-trust. Intel is simply hoping that AMD is too fearful to engage in litigation, or risk folding the business, simply to expose Intel to government action -- they are betting that AMD simply accepts whatever monthly tribute is required by Intel, thus assuring it's continued irrelevance without being wholly dismissed out of the market. If AMD still had its balls, they'd call the bluff and tell Intel to go to hell -- because Intel needs AMD a lot more than they're letting on.

Re:It won't succeed

[DragonWriter]

If Intel becomes the exclusive provider of x86 chips, they'll be smacked by the government with anti-trust litigation (Note: I did not say WHICH government, my fellow silly Americans).

Whether or not they are the sole manufacturer, since any manufacturer depends on a license from Intel, they are still a monopoly insofar as x86 chips are a valid market for anti-trust analysis; and, more relevantly, are arguably a monopoly, therefore, where desktop and traditional laptop PC CPUs are concerned, which is somewhat more likely to be an appropriate market. OTOH, in most jurisdictions, as I understand it, (and certainly this is the case in the US) just having a de facto monopoly isn't enough for anti-trust litigation, you have to leverage the market improperly. Enforcing an existing agreement that is, in itself, not an unfair leveraging of the market probably won't usually qualify, so just ending the AMD agreement because of a breach by AMD (if that has actually occurred) would probably not be a problem.

Re:It won't succeed

[MtViewGuy]

Intel better watch out because there is a precedent on this: the case against the United Shoe Machinery Company by the US government, namely abusing patent and other intellectual property rights laws to drive out competitors (USM had a huge patent portfolio on shoe-making equipment that they used to keep competitors at bay).

Given Intel's near-total domination of the x86 market and the property rights over several Intel CPU features, they risk being hit by the Feds like what the Feds to USM several decades ago.

Spansion vs. GlobalFoundries

[Anonymous Coward]

Would someone care to enlighten me about the difference? I Spansion was AMD's former manufacturing division because all AMD Austin fab employees have been Spansion employees for several years now.

Re:Spansion vs. GlobalFoundries

[artor3]

Spansion only ever made Flash memory, AFAIK. GlobalFoundries makes processors, which is why Intel can make these threats.

Another difference: GlobalFoundries still exists, and is making a new fab in Saratoga. Spansion, on the other hand, is trading for two cents.

Re:Spansion vs. GlobalFoundries

[Anonymous Coward]

You may be right about Spansion. That change may have occurred after Fab 25 quit making x86 processors and focused on flash products.

Patents are bad

[Anonymous Coward]

This is just another example.

For other examples on how intellectual monopoly is bad go here:

http://www.dklevine.com/general/intellectual/againstnew.htm

GF may not be a "subsidiary" according to Intel

[mr_mischief]

...but isn't that generally what a company that is in majority controlled by another company called?

Also, would AMD really have been so short-sighted as to sign a cross-licensing agreement with Intel that wouldn't allow AMD to contract an unlicensed third party to fabricate AMD's designs under AMD's licenses as an agent of AMD?

Re:GF may not be a "subsidiary" according to Intel

[DragonWriter]

but isn't that generally what a company that is in majority controlled by another company called?

What something is generally called is often entirely irrelevant to the definitions of terms in a particular contract. (There are a number of "general" definitions of "subsidiary", and they conflict: one is a firm wholly owned or controlled by another, another is one in which the parent has majority control, or in which an entities which are themselves subsidiaries of the parent collectively have majority control. Because of conflicts and ambiguities in general usage, terms that are important in contracts are usually defined precisely in the contract.)

Re:GF may not be a "subsidiary" according to Intel

[mr_mischief]

Yes, yes. I'm sure AMD's raft of lawyers is as specifically aware of the definitions that matter in this specific instance as are Intel's raft of lawyers. If there is a specific definition to reduce ambiguity, then they both have copies. If not, then they can argue either way and the judge can choose his or her own definition.

Re:GF may not be a "subsidiary" according to Intel

[DragonWriter]

I'm sure AMD's raft of lawyers is as specifically aware of the definitions that matter in this specific instance as are Intel's raft of lawyers. If there is a specific definition to reduce ambiguity, then they both have copies.

Right. My point was simply that common use has no bearing. No one not in AMD or Intel knows what is in the cross-licensing agreement, and so no one knows whether this is a clear violation, clearly not a violation, or in some murky space in between. And there are all kinds of reasons why it might be either, and why the companies might posture exactly the way they are now no matter what the reality is. Until the text of the agreement is known, there is no real intelligent debate to be had on whether or not it is a violation.

Re:GF may not be a "subsidiary" according to Intel

[mr_mischief]

You're right. We can't know if it is a violation. All we can say is that if it's a violation, there are only three really plausible explanations.

First, AMD's lawyers might not have considered the GF rollout when the cross-licensing deal was struck, and forgot to consider (or were ignored by management) the cross-licensing deal when the rollout was pending. Second, AMD's lawyers could be just useless, but that I doubt. Third, AMD knows it's a violation but they feel secure enough in the other legal wrangling against Intel that they're willing to renegotiate everything between the two and are using this as a bargaining point.

In other news

[tunafreedolphin]

Cats and dogs still hate each other. Oh wait, who cares.

x86 was a hack anyway

[nurb432]

Not trying to sound like a troll here, but x86 should have been retired decades ago. It designed in a totally different era and was never intended to scale well and its been a series of hacks to get it to do so. ( it was impossible to predict where we were going back then, the cpu industry was far too immature )

Sure, they have done wonders keeping it moving, but its long since time to start over with a clean architecture.

My preference would be MIPS or SPARC inspired, but thats just me, either way its time to move on/up.

Re:x86 was a hack anyway

[Anonymous Coward]

Starting over has been tried at least two ways I know of... Itanium and transmeta. Both have failed miserably. Care to bet on what would happen if some other one tried to come out?

Re:x86 was a hack anyway

[Jamie's Nightmare]

You should be OK. I'd wager you'd only be marked as a troll if you replaced "x86" with Unix. Though it would be an abuse of power in both cases.

Re:x86 was a hack anyway

[cdrguru]

Problem is, you would lose 50-70% of the existing computer market. Apple tried switching CPUs and in both cases needed massive hand-holding of customers, emulation and dedicated support from vendors. Sorry, but the Windows market doesn't have the same level of vendor committment.

Sure, lots of major software vendors (think Symantec) would help out their customers and would have a new chip architecture supported from day 1. And if there were no other vendors out there, it would be a pain-free transiation. The problem is for Windows that people are running software that was created in 1990. And the vendor may not exist any longer. Microsoft has seen recently how much fun it is to tell these people they have to find a different solution because the world has moved on. It didn't go well.

It would not go well changing hardware architectures, either. What we have is an investment in the billions of dollars in the x86 instruction set. While the source might be somewhat portable (ha!), the object code in user hands is not. Emulation and virtualization could help some, but it isn't the final answer and would not help everyone out. Just as it wasn't the solution for Apple either time.

Re:x86 was a hack anyway

[nurb432]

Like ARM? It seems to be slowly gaining market share.

Itainum had marketing issues, transmeta, while a cool idea i don't think really translated ( pun intended ) into useful product.

PowerPC as an alternative did well until apple pulled the plug, and is still there in the IBM world.

SPARC isn't dead either.

Re:x86 was a hack anyway

[sjames]

It was never even intended to be a general purpose CPU when designed. It was the I/O coprocessor that turned out to be much faster than the CPU it supported.

Re:x86 was a hack anyway

[Anonymous Coward]

x86 has been a RISC design for a long time now. It just translates in hardware from x86 to the internal RISC engine's language. So it's getting all of those wondrous benefits already.

In fact, an unexpected side effect of this is nice instruction set compression. With RISC you have to issue several ops to do what one x86 op could accomplish. Each instruction takes bandwidth that the x86 has saved. Therefore the x86 actually gets better performance.

I still say ARM is in a position to dominate as a platform. Despite all Intel has done it just can't even come close on energy efficiency.

Soon the general purpose computer will leave the desk and make its way into our pockets. Very few people will need or even want a full sized computer. Just like now most people now choose laptops over desktops. Then laptops to netbooks. And in the not too distant future, cell phones will do everything most people would want to do short of writing complex software, full length novels, or anything else requiring lots of typing. Even then maybe that's possible, too. But we're going to get there sooner on ARM than x86.

Re:x86 was a hack anyway

[DragonWriter]

My preference would be MIPS or SPARC inspired, but thats just me, either way its time to move on/up.

Given the fact that current Windows (Server, but its a start) already supports it, I suspect if x86_64 goes away, Itanium would have a leg up on being the 64-bit mass market chip of the near future, if Intel put the effort into delivering suitable versions for the consumer market.

Re:x86 was a hack anyway

[Anonymous Coward]

Internally modern CPUs have very efficient instruction sets nothing like x86. The x86 translation layer only is a small fraction of the transistor count of a modern machine.

Simply put, the cost of x86 on the hardware implementation side is small and decreasing while the cost of getting rid of it on the software side is enormous and increasing. It's here to stay.

It's feasible though, as the cost of the translation layer further decreases, that we will eventually see CPUs supporting several instruction sets though.

Re:x86 was a hack anyway

[nashv]

Not trying to sound like a troll here,

There times when you _are_ trying to sound like a troll?

Re:x86 was a hack anyway

[jgtg32a]

Its threads like this is why I keep coming to this site.

Re:x86 was a hack anyway

[Doctor Faustus]

My preference would be MIPS or SPARC inspired
I'd like to see the old AMD 29000 architecture come back. It was originally killed because AMD needed to put all of their engineers to work on x86, but it was doing pretty well for itself at the time. It uses the register window idea from the original Berkeley RISC project (and SPARC), but the window size can be different for each call, so it doesn't waster registers.

Re:x86 was a hack anyway

[Anonymous Coward]

I agree. x86 is aging, and I believe a new architecture is necessary.

good

[Bizzeh]

i hope they do, and AMD start producing cheap reliable PowerPC chips so we can all move over to a decent platform.

Re:good

[pak9rabid]

i hope they do, and AMD start producing cheap reliable PowerPC chips so we can all move over to a decent platform.

Right...so instead of having Intel's patents to worry about, they'd have Apple, IBM, and Motorola's patents to worry about...

Re:good

[Bizzeh]

ok then, pure amd64 cpu's? would most likly be a the best idea....

Re:good

[compro01]

Utterly impossible. x86-64 (aka AMD64) is a superset of x86, so the same patents still apply.

Re:good

[Bizzeh]

they only apply if you run x86-64, ie, dual mode cpus... strip away all ability to run 32bit instructions, amd are set.

Re:good

[TheRaven64]

Please, download the x86-64 ISA reference. Almost all of the instructions in 'pure 64-bit mode' are x86 instructions. Some things were removed, and a very small number were added, but around 90% of the instructions overlap with x86. You can take x86 assembly and use it on x86-64 without modification in a lot of cases.

The idea that x86-64 is some brand new AMD thing is just wrong. There were far more differences between the 286 and the 386 than there are between the Pentium and the Athlon64, and most of the changes in 64-bit mode just remove legacy things (segmented addressing, rings 1 and 2, etc), rather than adding new features.

At one level

[mutantSushi]

At one level, wouldn't it have been a smarter, lower-litigation-cost approach if AMD had spun off their NON-FOUNDRY (design) operations but kept all the x86 rights under the same house as the foundry? (if the design company wants to make x86 parts with other foundries, as they have done previously if I'm correct, they simply designate it as "design contracting" FOR the Foundry Company which holds the x86 rights (profit stream going to AMD, but that's a contract detail irrelevant to x86 licence).

At another level, what IF Intel ends AMD's x86 licence?
Isn't the point of the licence in the first place that AMD also has their own signifigant patents they could sue Intel for violating? I just don't see the logic in this, especially given that Intel seems to be doing GREAT compared to AMD, and AMD's continued existence gives Intel an anti-monopoly defense as long as they continue to compete in the x86 market.

At another level, this certainly seems big enough an issue to bring up the legitimacy of patent monopolies with regards to anti-trust law. US law doesn't generally hold (business) monopolies to be illegal per se, but I believe EU law *DOES*, and if Intel would gain a mainstream CPU monopoly by kicking AMD out of the x86 business, there would be repurcussions. If there was no x86 competition (VIA of course "exists"), the chances of EU nullifying x86 patents (or establishing "open" standardized licencing ala MP3) would seem to rise dramatically, which seems counter to Intel's interests.

Maybe Intel is scared of globalfoundaries?

[foxalopex]

Hmm, I wonder if the reason for this is Intel is scared of Globalfoundaries? If I'm not mistaken, the folks who bought the foundry from AMD are the same folks who are building in Dubai. You know the place where money flows like water and they're willing to waste billions to build custom islands? If that's the case, it is possible that AMD could be ramping up their production and process dramatically which would negate any gains Intel has. AMD also seems to have a more market friendly history with other companies than Intel has. Perhaps this is Intel's attempt to gain a monopoly before their ship sinks?

Re:Maybe Intel is scared of globalfoundaries?

[TinheadNed]

The tiny problem with Dubai is its money isn't made in oil, but in banking and tourism. One good indicator of how fucked its economy is, is that they're passing a law banning journalistic discussion of the economy.

And all the new building projects are being shelved as well.

http://www.kippreport.com/kipp/2009/01/21/what-freedom-of-speech/

Re:Maybe Intel is scared of globalfoundaries?

[Anonymous Coward]

I just heard from an Arab buddy yesterday that the economy in Dubai is crashing super hard right now, far more than the US one is.

Re:Maybe Intel is scared of globalfoundaries?

[Xtravar]

Dubai is sinking like a rock lately. Overpriced real estate, overpriced living, with nothing of substance behind it.

http://smashingtelly.com/2009/02/15/bye-bye-dubai/

Re:Maybe Intel is scared of globalfoundaries?

[DragonWriter]

If I'm not mistaken, the folks who bought the foundry from AMD are the same folks who are building in Dubai.

The other partner in the joint venture, besides AMD, is the Advanced Technology Investments Company (ATIC), which is controlled by the government of Abu Dhabi. Abu Dhabi is one of the United Arab Emirates, Dubai is a different one of the United Arab Emirates. The "folks who are building in Dubai" is kind of a murky label, but the high-profile developments in Dubai that get lots of attention seem to be each backed largely by one of Emaar Properties or Nakheel Properties, both which are "private" developers that are joined at the hip to the Dubai government; as far as I know, none of the high profile developments in Dubai are backed by ATIC, or other enterprises controlled by the Abu Dhabi government (who you would expect would be more interested in promoting developments in Abu Dhabi than in Dubai.)

Re:Maybe Intel is scared of globalfoundaries?

[DragonWriter]

The tiny problem with Dubai is its money isn't made in oil, but in banking and tourism.

Not that Dubai, except by analogy, is really relevant to AMD, since ATIC, AMD's partner in the Globalfoundries joint venture, is controlled by the government of Abu Dhabi, not Dubai. Two different emirates in the UAE.

x86 is dying

[droopycom]

Netcraft confirms ? ...

Okay, i'm not even going to bother going further with this stupid joke....

"open" patent licencing as remedy to monopoly

[mutantSushi]

Monopolies *ARE* illegal (irrespective of other issues, e.g. fraud), contray to US.
EU cannot be ignored by Intel.
If EU forced Intel to licence x86 on an "open"/equal basis (ala MP3), that would be AMD's ideal scenario, since as it would be court ordered, it would be a reasonable licence fee to start, AND it would not be subject to "give and take" over AMD's OWN patents implicated in x86-64: They would be completely free to sue Intel for whatever they can get over the x64 patents, and would have no fear of losing their licence/countersuits re: the core x86 patents.

Own goal, Intel.

Re:"open" patent licencing as remedy to monopoly

[tinkerghost]

Monopolies *ARE* illegal

No, their not. Abusing a monopoly position is.

I can certainly patent sexwidget and have a perfectly legal monopoly as the only company in the world producing them. Only if I try to force people to do other things not directly related to my sexwidget in order to get access to them is it considered abusing my monopoly status. In other words, if I try to force retailers to purchase other products like sexfoo & sexbar as a requirement for being able to sell sexwidgets, I'm abusing my monopoly.

Re:"open" patent licencing as remedy to monopoly

[Anonymous Coward]

sexfoo & sexbar
uuu, sounds kinda kinky

Re:"open" patent licencing as remedy to monopoly

[Anonymous Coward]

Somewhat unrelated, but why do cable/satellite providers force me to purchase 50+ worthless channels in order to get the 2-3 that I really want?

Re:"open" patent licencing as remedy to monopoly

[Late Adopter]

It's a bit more than that, isn't it? Otherwise Apple would be abusing a monopoly right now.

It's also important if your sexwidgets constitute a market, that is, how available close substitutes are. Which is itself a gray area (are other mp3 players substitutes for iPods? Are Satellite radios?)

Re:"open" patent licencing as remedy to monopoly

[Anonymous Coward]

its for protection of the prostitutewidgets

Stupid.

[Shads]

I personally think that's a damn stupid threat for Intel to make. AMD is arguably the only company that is preventing Intel from being broken up as a monopoly... you don't threaten to bury your only competition when you're nearly a monopoly. The various governments around the world aren't appreciative of that type of behavior. Unless they would like to be broken into dozens of pieces.

Don't be part of the problem

[A beautiful mind]

There isn't anything in x86 or it's various extensions that could be a licensing issue.

Intel, AMD, the patent lawyers and the government are the crazies here for trying to claim ownership of not only the most widespread processor architecture that is decades old, but of improvements over that architecture that are trivial to anyone who would want to improve it.

The patent craziness is causing real damage to the economy and innovation and it has to stop.

Dont see why they need a licence

[Eravnrekaree]

First of all, AMDs foundry probably is considered to have inherited the licence so I dont know if Intels claims really hold up.

Its been a long time since the chip architecture and schematic of AMDs chips have been directly based on Intels, if they ever have been. The only thing they share is the instruction set. Instruction sets are basically a language or communication protocol and these should not be copyrightable, just as someone could not copyright HTML, IM protocols or English. Only an implementation of software of these can be copyrighted not the language itself.

In my opinion, AMD does not need any licence to implement the ISA in the first place, just as a licence is not required to implement an SQL server or a computer language. Languages are simply not copywritable.

Re:Dont see why they need a licence

[cdrguru]

No, but all of these things you mention are patentable.

Specifically, I believe the issue is with the patents that Intel holds on the newer parts of the instruction set. The original 8086 instruction set patents have expired.

Re:Dont see why they need a licence

[Eravnrekaree]

Interacting with this is the fact that without AMD, Intel will have a monopoly and is using patents to maintain an unhealthy anticompetitive position. Really patent law should be modified to be require non discriminative licencing of the patent under reasonable terms to encourage competition in the market. I am surprised that the ISA is patentable at all, there is nothing at all innovative about it. Its really just an implementation of technology that has already been around for a while, just a reimplementation of an existing idea. ISAs like x86 have been implemented dozens of times. its not unique. Its like FAT, the idea is natural, its like patenting a fork.

The idea that GlobalFoundry does not have the licence rightly is mute anyway. Basically foundry is a piece of AMD that has broken off, the spirit of AMD lives in it, its different than if AMD has transferred the patent to a completely different company. GLobalFoundry is basically a broken off part of AMD and therefore rightly inherits the patent.

Re:Dont see why they need a licence

[joeyblades]

> Instruction sets are basically a language or communication protocol and these should not be copyrightable

Do you mean "should not" as in "it doesn't seem right to you" or as in "as recognized in the courts"?

Most instruction sets are considered key components of their respective computer architectures. As such, you will typically see copyright notices attached to them.

You may not think this is right, but the court of public opinion, is not the one that counts...

Re:Dont see why they need a licence

[DragonWriter]

First of all, AMDs foundry probably is considered to have inherited the licence so I dont know if Intels claims really hold up.

IP license agreements very frequently written so as not to be transferable without the permission of the owner of the licensed IP. Its not AMD's foundry, its a separate company which AMD formed with ATIC (the latter being, at least according to most reports I've seen, the majority interest holder) and attempted to transfer some rights under the license (while also retaining them in AMD, as I understand it.) The license might allow that, it might not, but its not the case of a successor in interest like you would have if the new company was a separate firm that had bought AMD outright, and even with those kind of situations ISTR seeing cases where the terms of a license agreement caused it to be terminated (with licensed merchandise based on TV and movie franchises) by that kind of transfer.

Its been a long time since the chip architecture and schematic of AMDs chips have been directly based on Intels, if they ever have been. The only thing they share is the instruction set. Instruction sets are basically a language or communication protocol and these should not be copyrightable, just as someone could not copyright HTML, IM protocols or English. Only an implementation of software of these can be copyrighted not the language itself.

I'm pretty sure the main issue with the license are patents on elements of the technology not copyright on the schematics and instruction set.

Re:Dont see why they need a licence

[Fulcrum of Evil]

Intel will have a monopoly and is using patents to maintain an unhealthy anticompetitive position.

Which isn't illegal. It's only illegal if you leverage your monopoly into another area.

Really patent law should be modified to be require non discriminative licencing of the patent under reasonable terms to encourage competition in the market.

So start a company, get a patent, get steamrolled by an incumbent who licenses your tech for a year or two until they come up with a way to not pay patent fees.

Re:Dont see why they need a licence

[Darby]

So start a company, get a patent, get steamrolled by an incumbent who licenses your tech for a year or two until they come up with a way to not pay patent fees.

So, by "get steamrolled" I have to assume you mean, license their new patent or whatever allowed them to outcompete you and then get back in the game thereby causing a more rapid improvement?

Oh, wait...you meant you fear competition and hope to profit for life off of one halfway decent idea?

It's hard to tell what you really mean.

Re:Dont see why they need a licence

[ameyer17]

DISCLAIMER: not a lawyer
And the patents on the 386 instruction set "should" be expired or unenforcable too.
The first 386 chips came out in 1986, plus a year after the commercial implementation to file plus 20 years for the patent itself.

Also, instruction sets shouldn't be patentable, IMO.

Re:Dont see why they need a licence

[Fulcrum of Evil]

So, by "get steamrolled" I have to assume you mean, license their new patent or whatever allowed them to outcompete you and then get back in the game thereby causing a more rapid improvement?

No, I mean use their massive power of scale to incorporate your innovation (which is licensed at some unknown price set by persons unknown), improve their brand while yours is nonexistent, and figure out how to make the product without paying you fees, all in less time than it takes you to reocup R&D. Or they could steal your idea anyway, but that happens already.

Oh, wait...you meant you fear competition and hope to profit for life off of one halfway decent idea?

Patents are protection from competition for a period of time in exchange for documenting your work. Taking that away hardly makes it appealing.

It's hard to tell what you really mean.

And you seem to have overlooked follow on effects in favor of feel good sentiments.

Re:Dont see why they need a licence

[Ihlosi]

Oh, wait...you meant you fear competition and hope to profit for life off of one halfway decent idea?

Are we talking about patents or pop music here?

Re:Dont see why they need a licence

[TheRaven64]

When people talk about patents on instructions, what they really mean is patents on the only sensible way of implementing certain instructions. This was the case, for example, in MIPS Technology's patent on unaligned load and store instructions in a RISC processor. There were ways around this, but they are slow and complicated. People implementing MIPS cores had three options:

1. License the patents from MIPS Technology.
2. Create a core which didn't implement these instructions.
3. Come up with some unrelated way of implementing them.

The patent (U.S. Patent No. 4,814,976) covered a specific way of implementing four instructions. If you implemented them in a different way, then you would not be infringing, but I don't know of any companies that did this (by the way, the patent expires this year, so expect to see more complete MIPS implementations start appearing). A few licensed the patents. Most implemented almost-MIPS architectures, where the four relevant instructions were omitted. GCC and a number of other compilers have switches that allow you to generate code which doesn't use these instructions and an operating system can (very slowly) catch the illegal instruction exception and emulate them for legacy code.

Re:Dont see why they need a licence

[Fulcrum of Evil]

Patents last 20 years from filing. Do try to keep up.

More than instructions (I think)

[WoTG]

I believe the cross licensing goes way beyond just the instruction set - that's public knowledge, you wouldn't be able to write a compiler if you didn't know what instruction did what!

I think there are all kinds of implementation details to actually processing the instructions that are involved in the patents (on both sides).

Now that I think about it, a lot of these processing details are probably infringed upon by other non-x86 chips, there are only so many ways to do things... there must be other patent licensing deals out there that don't make the news as often as the x86 lawsuits.

Re:More than instructions (I think)

[maxume]

The whole point of patents is that they become public knowledge.

AMD is a dyslexic MAD

[mangu]

So, surely this is a case of mutually assured destruction for both isn't?

Only if it comes to a final shootout, which will never happen.

But considering what usually happens when a show of force happens, the bigger guy wins. In this case, Intel. It's Intel who forced a confrontation, and they probably will gain better conditions in small print in a contract somewhere.

Otherwise, nothing to see here, please move along.

Bait and swtich

[olddotter]

This is probably just high stakes gambling. AMD has little to lose. (I say that as an AMD share holder looking at my $2.49 stock price.) Intel has more to lose if they have to redo the 64Bit code. According to the reading, if Intel wins, they get rid of AMD, and become a defacto monopoly having to face US and EU anti-trust regulators. If AMD wins, they get to go along as before and Intel can't sell 64-bit CPUs that people want.

Basically I bet AMD's lawyers are saying "Go ahead make my day." Given the above even if Intel wins in court, they lose.

Re:Bait and swtich

[Anonymous Coward]

Definition of subsidiary: As long as the parent company has more than 50% of the voting stock in the subsidiary, it has control. In the case of a foreign subsidiary, the company under which the subsidiary is incorporated must adhere to the laws of the country in which the subsidiary operates, although the parent company still carries the foreign subsidiary's financials on its books (consolidated financial statements).

AMD doesn't own 50% of the foundry.

I don't know about the incorporation clause here? Anyone?

Re:AMD is full of shit

[QX-Mat]

you're not all troll. very consice and profane i'd argue. as a failing lawyer, i agree with what you said. it all depends on the contracting company and the conditions set in the agreement. more likely than not AMD have a leg to stand on, but a flimsy one at the least. i dont think intel can afford to loose the 64bit crosslicence however.

Back to PPC?

[Cro Magnon]

Does that mean Apple will go back to the PowerPC platform?

Re:Back to PPC?

[Anonymous Coward]

Why?

Lead, follow, or get the hell out of my way.

[geekmux]

But the big draw of windows is the inertia of 1,000,000 one-off apps that businesses have written. Microsoft would be scared of people moving to another architecture just because if people were making a (painful) switch anyway, they might look at the alternatives.

If you wrote the damn app, then learn to recompile it and move on to whatever/whomever is going to be pimping procs next month or next year. If you're that worried about your legacy apps, then learn to use virtualization.

Moores law didn't get to be a "law" by playing nice and waiting around. Lead, follow, or get the hell out of my way.

Inertia is as fast and powerful as the people behind it. Adapt or die. It's that simple.

Re:Lead, follow, or get the hell out of my way.

[hasdikarlsam]

Virtualization only works when you're on the same hardware as the virtualized environment, or approximately the same. You can't virtualize an x86 OS on PPC hardware; that's emulation, and it's dog-slow, buggy, or both.

Also, why isn't there a "-1, wrong" moderation?

Re:Lead, follow, or get the hell out of my way.

[Anonymous Coward]

No it became a law through observation. The law does not state that "new stuff" comes out every month be prepared (read:learn to recompile), in fact it is the opposite.

Re:Lead, follow, or get the hell out of my way.

[halltk1983]

Also, why isn't there a "-1, wrong" moderation?

Religious and / or political threads.

not the architecture --- it's the platform

[Anonymous Coward]

x86 is dominant not so much because of the
cpu, but because of the architecture. by
this i mean, a standard set of base devices
for console, add-in cards, etc. wake
me up when the arm world has a standard
architecture. until then, it has no chance.

by the way, arm performance sucks compared to x86.
that problem would need solving, too.

Good bye

[Anonymous Coward]

Good bye Core 2 and Core i7... we hardly knew ye

Bring back Alpha...

[argent]

If I recall correctly, both Intel and AMD have licensed Alpha technology from DEC-I-mean-Compaq-I-mean-HP. Maybe they could get together with a 64-bit architecture that actually works well.

Re:Bring back Alpha...

[Fulcrum of Evil]

what don't you like about x86-64?

Re:Bring back Alpha...

[Saffaya]

The x86 part

Atom

[koldewaj]

By Intels logic, Wouldn't Intel be in violation if they go forward with plans to outsource Atom Production? http://hardware.slashdot.org/article.pl?sid=09/03/02/2111221&from=rss

Re:Atom

[Xtifr]

Wouldn't Intel be in violation if they go forward with plans to outsource Atom Production?

In violation of what? Their license with themselves? Perhaps you're picturing them suing themselves? I think you need to go look for SCO to find that level of insanity (everyone keeps saying that SCO should sue Caldera). :)

I don't think Intel's license with itself is any more public than it's license with AMD, so speculation is pretty much pointless, but I do suspect that Intel's license with itself is a whole lot shorter! :)

Re:Atom

[koldewaj]

Umm. The some Atoms uses AMD64 (x86-64) tech. And I am pretty sure they can use it becouse of cross licensing with AMD. See http://en.wikipedia.org/wiki/X86-64 for referances.

x86 patents are expired

[pavon]

it would be less of a problem for them than AMD losing the rights to use x86-anything.

That is not going to happen. Patents only last 20 years, which means that any patents filed on 486 or older technology is no longer covered by patents. If anything AMD would lose the extensions that have been added since then like MMX, SSE, 3dNow.

In my opinion, the real threat here isn't loosing their license to the instruction set, but rather other licenses with regards to process (fabrication) and design details (like patents on instruction set to microcode translation, or branch prediction algorithms, etc).

Re:x86 patents are expired

[Anonymous Coward]

model name : AMD Phenom(tm) 8650 Triple-Core Processor
flags : fpu de tsc msr pae cx8 apic mtrr cmov pat clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm 3dnowext 3dnow constant_tsc rep_good pni cx16 popcnt lahf_lm cmp_legacy extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch

model name : Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall lm constant_tsc arch_perfmon pebs bts rep_good nopl pni monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr lahf_lm

I'm pretty sure 3dnow belongs to AMD. Looks like a lot of other extensions (including the others you mentioned) would be lost though.

Re:x86 patents are expired

[Anonymous Coward]

Yes, 3dNow was AMDs response to MMX

So if microsoft...

[ducomputergeek]

....starts supporting the PPC architecture on windows, and Apple supports the minority x86 from Intel....

I'm starting to think that just maybe 2012 IS the end of the world....

Bring back the 6510!!

[Dysan2k]

Just update the sucker for more modern hardware, chunk it to 64-bit ops, and let's relive the C=64 days!

Re:Bring back the 6510!!

[Orion Blastar]

Nah bring back the 68000 series and chunk it up to 64-bit ops. Forget the C=64 days, let's create the Super Amiga with the 6806400 Processor!

Re:Bring back the 6510!!

[Anonymous Coward]

It is already a running task

http://www.natami.net/

A 100% compatible AGA amiga on an FPGA@~160mhz
but with modern extension, chunky mode 16bit & 24bits, higher resolutions, 16x16bit audio channels and a new 68k softcore codename N68070 wich has the same instruction set as a 68020 but with pipelines and branch predications (keeping compatibility with auto-modifier code).
256mo chip (zero wait state)/256mo fast

So your dream is not so dreamy

Re:Bring back the 6510!!

[downix]

They have, it's called the ARM.

Why now?

[stei7766]

IANAL, but wouldn't it look bad in court if after AMD went through all the trouble to spin off the whateverfoundry with nary a peep from intel?

Now all of a sudden Intel cries foul? That sounds awfully anti-competitive, to sit on something like that until the split it complete.

Re:Why now?

[DragonWriter]

IANAL, but wouldn't it look bad in court if after AMD went through all the trouble to spin off the whateverfoundry with nary a peep from intel?

Well, if that was the case, it might. Of course, Intel says they made several attempts to discuss the issue with AMD when they became aware of AMD's plan for the spin off, and AMD refused to discuss the issue, and once the spin-off was done (which is, according to Intel's claim, when the breach occurred), Intel sent notice that it was viewing AMD's action as a breach and providing the required notice that if the breach wasn't cured within 60 days, Intel would terminate the cross-licensing agreement.

Surprisingly enough, AMD's story paints AMD as perfectly in the right, and Intel as completely wrong, and Intel's does the same for AMD. The facts could be anywhere from one pole to the other, but since we don't know what communication actually occurred, before or after the spin-off, and we don't know what the actual licensing agreement says, we really don't know who is right and who is wrong.

Virtual machines

[Anonymous Coward]

Virtual machines.

Oh thank you java. I don't have to care.

In Soviet Russia

[ZPWeeks]

it's hard to claim they strong-arm computer manufacturers into using their products.

...Computer manufacturers using StrongARM into their products!

Nvidia must be rubbing it's hands and

[w0mprat]

finally lithp architecture may have a chance. Yeth, Lithp.

MAD

[Anonymous Coward]

Mutually Assured Disassociation

Nvidia

[Yarhj]

Nvidia has a lot to gain here. Currently, Intel is trying to keep them from combining their low-power graphics chips with Intel's Atom on a single package (which would be fantastic for netbooks and low-power notebooks).

If AMD wins, they might be more likely to grant Nvidia the x86 license they've been seeking for so long. If Intel wins, AMD is effectively hamstrung, and they may very well have to grant Nvidia (and others) an x86 license to stave off antitrust regulators.

Let the games begin

Re:Nvidia

[DragonWriter]

If AMD wins, they might be more likely to grant Nvidia the x86 license they've been seeking for so long. If Intel wins, AMD is effectively hamstrung, and they may very well have to grant Nvidia (and others) an x86 license to stave off antitrust regulators.

Neither company's preferred, winning position is that the cross-licensing agreement is terminated. Both, of course, are accusing the other of breach (though AMD's "by claiming that we breached the agreement, you breached the agreement" line is pretty...well, I'd be surprised if the agreement actually justified that, especially given that Intel didn't publicly announce the claim of a breach, AMD did in their SEC filing, which is where the news came from.)

Intel most likely wants AMD to give concessions in return for a modification to the agreement that would allow the spin-off, and AMD wants to be allowed to spin-off without consequence. If there is any other result, neither Intel nor AMD has really "won".

Virtualization/Translation

[nurb432]

Sure, its not *the* answer since the goal is native code, but it would extend the transition period out long enough to make it stick, if everyone was onboard with whatever alternative was decided on.

Re:Virtualization/Translation

[maxume]

It is fairly likely to happen when cheap smartphones start shipping with 64 GB of flash and decent peripheral support.

Re:AMD is full of shit

[Abuzar]

What kind of dumb-ass moderator would tag the parent-post as troll?
I don't agree with the poster. I think the poster is wrong... but troll? Seriously, the moderators around here need an IQ check.

Give it a rest already!

[Abuzar]

These people are arguing over decades-old technology needed only for backwards compatibility while the rest of us here have to put up with sooooooo much BS trying to implement their outdated architecture.... come on! I'm sick of being stuck with having to x86 code every low level project I work on... it's just there to weigh down desperately needed progress!

I say break the x86 arch, let's a few of us hack windows, linux will be a piece of cake and move on to a whole new modern arch. Let the two dinosaur buffoons fight it out while we give newer more agile companies a chance and solve a whole bunch of IT problems at the same time.

No.

[AcidPenguin9873]

I wish Slashdotters would stop with the incessant "x86 sucks" mantra. You're all fools.

There's plenty of crufty old instructions in the x86 ISA; no modern compilers generate them though, so no one cares that they're there. They take up a couple pages in the ISA manual I guess. The die area it takes to implement them is totally, completely insignificant. They're either in microcode (along with a bunch of other really useful instructions) or the hardware already exists for some other reason.

There's plenty of crufty segmentation and weird ways of laying out memory and whatnot; no modern OS uses that though, so no one cares that it's there. And again with the ISA manuals and some transistors. And there's plenty of modern paging and flat memory models and whatnot too.

AMD and Intel both know how to make good, fast, and (relatively) small hardware to decode variable-length x86 instructions. Yes, of course an x86 decoder is bigger (i.e. more expensive, more difficult to implement, etc.) than a RISC fixed-length decoder, but again, no one cares because we already know how to do it fast enough and cheap enough. Check out an x86 die photo sometime; most of it is cache. Probably about 1/50th is decoder.

And CISC-style+variable-length instructions get you a smaller code footprint and thus better instruction cache utilization vs. what you'd get with a fixed-length instruction stream. Examples: common ops get shorter instructions, there are more flexible addressing modes, more flexible sources/dests within a single instruction, you get one x86 instruction (no more than 15 bytes) to do what would take multiple RISC-style instructions (probably more than 15 bytes).

Sure there's the crufty x87 floating point stack. But there's also the shiny new SSE/SSE2/SSE3/whatever instructions, and modern compilers can exclusively use SSE/SSE2 to do the exact same thing (-mfpmath=sse does it in gcc). And again, die area for x87 FP stuff isn't a big deal since a lot of the hardware is shared with SSE.

ISA extensions have been added to cover all the newfangled SIMD stuff and virtualization you can want. AMD64 covers 64-bit stuff. And 64-bit stuff gives you extra registers too (8 extra integer, 8 extra SSE for a total of 16 each), which is great and a nod to the large number of registers that RISC machines give you.

In short, what the hell is everyone bitching about?

Re:No.

[Anonymous Coward]

AMD and Intel both know how to make good, fast, and (relatively) small hardware to decode variable-length x86 instructions. Yes, of course an x86 decoder is bigger (i.e. more expensive, more difficult to implement, etc.) than a RISC fixed-length decoder, but again, no one cares because we already know how to do it fast enough and cheap enough. Check out an x86 die photo sometime; most of it is cache. Probably about 1/50th is decoder.

If this were all true, how is it that the ARM Cortex-A9 chip is able to get four superscalar RISC cores on a 65nm chip running at over 1GHz ... yet using only about 1 watt of power? Intel's single-core Nano may run at a faster clock, but it uses maybe twice the power and it can't compete with a quad core superscalar RISC CPU.

Re:No.

[Anonymous Coward]

Power Consumption and heat dissipation.

x86 is fine for the server.

x86 is fine for the desktop.

x86 is not suitable for portables.

Re:No.

[Anonymous Coward]

The extra decode complexity of x86 chips adds more pipeline stages to the execution path, increasing latencies for executing uncached instructions and increasing branch mispredict penalties.

All of the numerous add-ons to the instruction set (SSE, x64) require more instruction prefixes and counter the small code size benefit of a well designed CISC architecture.

I don't design chips so I'm sure there are many other benefits of having a clean instruction set architecture. Backwards compatibility with x86 certainly hurts performance. Just because Intel's billions of dollars overcome the problems does not mean that the ISA is irrelevant.

Re:No.

[Wrath0fb0b]

If this were all true, how is it that the ARM Cortex-A9 chip is able to get four superscalar RISC cores on a 65nm chip running at over 1GHz

And how many actual instruction does that retire per second? Does its FPU perform at even a fraction of the equivalent x86 (I was surprised to see that it even had one -- I wrote a small program for my ARMv4 phone and spent a solid hour wondering why a program that runs in a few seconds on my desktop took many minutes on there. Punchline: don't do double-precision floating point math on your phone)? What about pagetables, interrupts, DMA and coherent cache? Memory controller? FSB/QPI?

ARM is a wonderful architecture for my mobile phone but it's never going to be enough for actual work. Maybe a netbook running a stripper version of 'nix, but I'd have to see it first to believe it.

Re:No.

[Anonymous Coward]

Seriously, who cares? My computer works. Let the companies play their stupid legal games. Nerds.

Re:No.

[AcidPenguin9873]

The extra decode complexity of x86 chips adds more pipeline stages to the execution path, increasing latencies for executing uncached instructions and increasing branch mispredict penalties.

If your instruction stream is uncachable, that's pretty much a don't-care for performance; it's going to be super slow regardless of pipeline depth because you have to go out to DRAM or IO (hundreds of clocks) every time you want to fetch a new block of instructions. Saving the few clocks in decode isn't going to make a big difference.

Bracnh mispredict penalty I'll give you, x86 decode does increase that by a few clocks in some cases. But there are a number of arch techniques which help reduce or hide that front-end latency (fetch/decode) from branch mispredict penalties: checkpointing the back-end so the front-end can be steered to the branch target early, once the branch mispredict is discovered (MIPS R10000, I don't know if x86 processors do it), using a trace cache to remove decode stages from the front-end pipe (P4, yes we know it was a failure but it had some interesting uarch ideas), storing decode hints in the instruction cache and L2 cache to speed up the variable-length instruction decode (AMD processors from Athlon to present).

All of the numerous add-ons to the instruction set (SSE, x64) require more instruction prefixes and counter the small code size benefit of a well designed CISC architecture.

I'm no RISC ISA expert, but aren't the SIMD instructions also extensions to their instruction set as well? Maybe they had enough free opcodes to do it without expanding any code size, I don't know. If so, good for them. If not, that's the same as x86.

In any case, it sounds like a wash, then. An x86 machine spends its code size benefit on useful ISA extensions and ends up with about the same instruction cache utilization as a RISC machine. A redesigned CISC could beat x86 in that one metric but would necessarily break ISA compatibility. That's possibly the one area where x86 compatibility costs it in performance. I say, BFD.

I do know that multiple implementations of ARM's NEON instructions basically tack a SIMD FP unit on to the *back* end of its pipeline. So doing SIMD FP on an ARM is likely to be slower than x86.

Re:No.

[Brane2]

I wish Slashdotters would stop with the incessant "x86 sucks" mantra. You're all fools.

There's plenty of crufty old instructions in the x86 ISA; no modern compilers generate them though, so no one cares that they're there. They take up a couple pages in the ISA manual I guess. The die area it takes to implement them is totally, completely insignificant.

Take a look at what it takes to decode x86 instruction in parallel and then we'll talk.

Re:No.

[ion.simon.c]

Seriously, who cares? My computer works. Let the companies play their stupid legal games. Lawyers.

FTFY.

Re:No.

[TheRaven64]

Considering that Acorn were running a multitasking graphical operating system on 25MHz ARM chips, I think you might need revisit your definition of 'real work'. You should also take a closer look at the interrupt architecture of ARM if that's one of the things you believe is superior on x86.

The Cortex A9 scales to four cores on a single die and runs at over 1GHz, while keeping a power envelope of under 1W. If you've only used ARMv4 then you're probably unfamiliar with the NEON instruction set included in most recent ARM chips. This provides 128-bit vector support and both integer and floating-point instructions. The Cortex A8 comes with NEON as standard (it's an optional extension on other ARM7 architecture cores). If you're really doing anything FPU-intensive, however, you should take a look at the on-die DSP that comes on most Cortex A8 chips, such as the OMAP3 or i.MX5 lines.

A modern ARM chips is faster in every respect than the CPU in the old ThinkPad that I still use for real work. It's not as fast as a Core i7 or a POWER6 by any means, but in terms of performance per watt - which translates directly to battery life - it does very well.

Re:No.

[marcosdumay]

"I wish Slashdotters would stop with the incessant "x86 sucks" mantra. You're all fools."

Well, ok, I'll bite. For an architecture that doesn't suck, the following list is way too big:

"There's plenty of crufty old instructions in the x86 ISA; no modern compilers generate them though, so no one cares that they're there."

"There's plenty of crufty segmentation and weird ways of laying out memory and whatnot; no modern OS uses that though, so no one cares that it's there."

"Yes, of course an x86 decoder is bigger (i.e. more expensive, more difficult to implement, etc.) than a RISC fixed-length decoder, but again, no one cares..."

They still use die area, they still can't be turned off (to save power), and they still introduce defects (what cache almost doesn't do). So, they still lead to a more expensive and power hungry chips.

"Sure there's the crufty x87 floating point stack. But there's also the shiny new SSE/SSE2/SSE3/whatever instructions, and modern compilers can exclusively use SSE/SSE2 to do the exact same thing..."

It doesn't matter (except for the die area, number of defects and not being able to turn offf), the compiler won't use it, unless you are compiling your own binaries to your own machine. Those instructions are too varied for pre-packaged software.

"And CISC-style+variable-length instructions get you a smaller code footprint and thus better instruction cache utilization..."

And that is the one advantaje of CISC. Ok, but RISC has no problem compressing the code stream, making even better use of cache (CISC could also compress, of course, making the point moot). And, now, to the important topic:

"And 64-bit stuff gives you extra registers too (8 extra integer, 8 extra SSE for a total of 16 each), which is great and a nod to the large number of registers that RISC machines give you."

RISC machines used to give you 32 register of each kind, by the time they were 32 bits. That was a long time ago... Now they usualy come with 64 or more registers of each kind. One of the problems the compact instruction set gives you is that once you use all the instructions, you can't address more registers, that is why x86 is stuck at 16 of them.

Re:No.

[AcidPenguin9873]

They still use die area,

Yes, I said that. I also said that compared to die area spent on lots of other things in a modern microprocessor/SoC (cache, memory controllers, point-to-point coherent and IO links, buffers everywhere), it's probably minor.

they still can't be turned off (to save power),

You would be surprised what intelligent clock gating can do. Clock gating can't reduce leakage but it does reduce dynamic switching power.

and they still introduce defects (what cache almost doesn't do).

That's another side effect of them taking up die area. You said it three times in your post; I've acknowledged it but asserted that the die area is small compared to other things. I bet defects in cache arrays are WAY more likely to cause a chip to be scrapped or sold as a lower-performing part.

It doesn't matter (except for the die area, number of defects and not being able to turn offf), the compiler won't use it, unless you are compiling your own binaries to your own machine.

I don't believe you, and in fact for one case I can prove you wrong. I can't speak for every software distributor that ships binaries, but since every single 64-bit x86 processor has support for both SSE and SSE2, gcc, by default, uses them when it's compiling code for an x86-64 target (look under the fpmath section): http://gcc.gnu.org/onlinedocs/gcc/i386-and-x86_002d64-Options.html

RISC machines used to give you 32 register of each kind, by the time they were 32 bits.

RISC machines are a pure Load/Store architecture and don't (or didn't) have the notion of load-execute or load-execute-store instructions. That necessitates more temp registers for doing operations to avoid too many loads/stores to stack-local variables. Ld-ex and Ld-ex-st help to reduce your need for registers in some cases. I would be interested to see register utilization and stack spills for a RISC machine vs. an x86-64 machine, but I haven't seen any data about it.

One of the problems the compact instruction set gives you is that once you use all the instructions, you can't address more registers, that is why x86 is stuck at 16 of them.

The extension from 8 to 16 used an instruction prefix which added an extra bit onto the register address fields in the sources and dest of the instruction. There's nothing stopping AMD or Intel from adding another prefix to add yet another register bit to those fields if anyone ever wanted to go to 32.

Re:No.

[AcidPenguin9873]

Take a look at what it takes to decode x86 instruction in parallel and then we'll talk.

With decode hints that come out of the instruction cache along with the instruction bytes themselves, it's not too bad. Besides, I don't think x86 machines are decode-bandwidth-limited. I don't think any modern microprocessor is decode-bandwidth-limited, it's memory-bandwidth and cache-miss-latency limited.

Re:No.

[Doctor Faustus]

I don't think x86 machines are decode-bandwidth-limited
I'd be a lot more concerned with latency, there. Decode is one more step in the pipeline that the first new instruction has to go through after a branch misprediction.

Re:No.

[marcosdumay]

"cache, memory controllers, point-to-point coherent and IO links, buffers everywhere"

All of that, except for links, is redundant. If there is a defect somewhere on the cache, you lose one block (memory controlers are an interesting case, as it isn't completely redundant, but mostly so), if there is a defect at the fectcher, you lose a core. That makes for a huge difference between a simple fetcher and a turing complete one. IO links also have a big feature size (or redundant features, depending on manufacturing method), and thus a lower probability of defects.

"RISC machines are a pure Load/Store architecture and don't (or didn't) have the notion of load-execute or load-execute-store instructions."

Yes, that is right. And load-execute is slower than pure load-store unless you spend a fortune optimizing your projects for that bottleneck.

Now, you are right about gcc using SSE and SSE2 when compiling for AMD64. I got myself stucked at the 32 bit times here.

Re:No.

[AcidPenguin9873]

Yes, decode adding latency to the pipe was discussed in these posts. The GP was talking about decoding x86 instructions in parallel, which to me indicates he's concerned about decode bandwidth, not latency.

Best case for IP reform

[sjames]

The best case for IP reform will be if BOTH AMD and Intel pull the trigger and the entire supply of CPUs faster than the Atom dries up overnight. Sorry DoD, no new computers for you either! It's been a while since Intel made mainstream CPUs that didn't use any of AMD's IP, they can't redesign overnight and you can bet that AMD won't let them slide if they just disable 64 bit mode without removing the IP entirely from the cores.

Itanium is WAYYYYYYYYY too expensive for a desktop and for most servers and, of course, uses an entirely different instruction set.

You can bet that some sort of compromise will then be forced on both entities on any legal pretense any government can think of.

Re:Best case for IP reform

[DragonWriter]

Itanium is WAYYYYYYYYY too expensive for a desktop and for most servers and, of course, uses an entirely different instruction set.

But is it fundamentally expensive, or is it expensive because, despite their original plan to eventually make it a mass market chip, Intel relegated it to an expensive niche when x86_64 became popular as a consumer chip before Itanium had its early kinks worked out and Intel cross-licensed x86_64 for mass market use instead?

I tend to think its the latter, not the former. If x86_64 became legally impossible, there'd be far less-expensive mass market Itanium chips from Intel within months.

Re:Best case for IP reform

[sjames]

Considering they wanted $8000 for it when it had even worse performance than it does now and they had every reason to want to saturate the market with their 64 bit solution before they got stuck with the embarrassment of licensing x86_64 from AMD, AND it was a royal pain to program for and that furthermore, they hardly budged on the price even when they could barely give them away, I'm guessing there must be something intrinsically expensive about it. Either that or they have some sort of mental block that won't let them face reality and isn't likely to go away any time soon.

Re:Best case for IP reform

[dlapine]

Fundamentally expensive. High voltage, low speed, large cache. Old manufacturing process. Not dual core (just multiple die per chip). The current ia64 implementation doesn't seem to lend itself to easy speed increases, given that the top speed is just 1.66ghz, and that after 6 years of development.

I could go on, but really, Itanium isn't an option unless Intel had nothing else, including the Atom. They could ramp the Atom up a lot faster.

In the extremely hypothetical world of x86 mutually assured destruction, where no one can build any x86 variant, I'd expect IBM's power chips to come out on top, at least in the short term. Linux already runs on them, and Apple/Windows has recently run on them. Be a pain to switch, but possible.

Re:Best case for IP reform

[ameyer17]

And as far as I can tell AMD couldn't ship any particularly useful mainstream processor either.

I mean, even if you run 64 bit Windows, you're probably running a significant quantity of 32 bit only application code.

And I'm not sure how useful the AMD64 instruction set would be if you ripped out all the 32 bit x86 instructions anyway

Re:Best case for IP reform

[sjames]

Exactly! No new CPUs at all. A great big demand and zero supply while the foundries idle.

Good move

[janwedekind]

Yes! Let's all sue each other. Surely that will end the economical crisis.

Re:AMD is full of shit

[blair1q]

Dumb-ass moderators who identify with the "fanboy" comment and lament their years of fealty to a lie.

Dubai real estate collapse

[GPS Pilot]

Some Dubai properties lost 64% of their value from 2001 to November 2008: http://www.telegraph.co.uk/finance/newsbysector/constructionandproperty/3489393/Dubais-Palm-Jumeirah-sees-prices-fall-as-crunch-moves-in.html

Oh well, X86 was nice while it lasted

[Orion Blastar]

Looks like we either choose ARM or PowerPC to replace X86 technology and run X86 programs via emulation.

Re:Oh well, X86 was nice while it lasted

[drspliff]

Or Alpha? The tech is all there, several [open source] operating systems still support it and presumably the tech wouldn't be too costly to license?

Re:Oh well, X86 was nice while it lasted

[TheRaven64]

I'm as much of an Alpha fan as anybody - probably more - but it doesn't really fit into any current niche. The market at the moment wants performance-per-watt. The desktop market is shrinking and the two growing markets are mobile (which is huge) and datacenter (which is smaller but has bigger margins). For a mobile device, low power means better battery life. For a datacenter, low power means more units in the physical volume and lower running costs. It's only desktops and workstations that can get away with the kind of fast-at-all-costs design that the Alpha embodied.

Re:Oh well, X86 was nice while it lasted

[downix]

HP sold Alpha off to Intel, which is the root of the licenses being discussed. Incidentally, the SPARC is open to license, and has already a well developed toolchain as well as a huge support base to work from, and is native 64-bit with the V9.

Hi there, anyone remember VIA?

[Orion Blastar]

Does VIA still make X86 CPUs? Is VIA still in business?

Anyone remember when NEC made the V20 CPU that was an 8088 Clone but at double the clock speed?

Re:Hi there, anyone remember VIA?

[BUL2294]

The NEC V20 was an 8088 replacement, but it really was an 80186 processor with an 8088 pin-out. The V30 was the 8086 version. Having come to market some 8 years after the 8088/8086, they incorporated design improvements and new instructions, but were generally about 30% faster. They were great for spreadsheets but sucked with games--games of that era utilized published timings for instructions. "Turbo" XTs didn't affect them too much since every instruction that took x amount of time on a 4.77MHz 8088 took x/2 time on a 9.54MHz 8088. On a V20/V30, games would seem to speed up or slow down based on what instruction was run...

Re:Hi there, anyone remember VIA?

[irieken]

Yes, VIA is still around, and they make some decent little 64-bit x86 processors that sell under the name Nano.

The Nano processor actually does a pretty good job of holding its own against the Atom; edging ahead in almost every area, except for power consumption.

Re:Hi there, anyone remember VIA?

[Orion Blastar]

I recall there was a program called slowpc.com that fixed that Turbo XT problem.

Sorry to poke fun

[omb]

Nothing runs fast on Itanium, it is the biggest architectural design mistake in a century of CPU design, even Babage did better, a triumph of marketing { what MBA's call strategic planning over reality}, however, the AMD 86_64 contains a perfectly good embedded RISC engine to which Intel has no essential rights.

If its native IS [of the AMD 86_64] were to be exposed would be both easily re-targetable, say USD 50 M for a commercial (quick, 3 month (eg Code Sourcery)) GCC back end. would re-target the silicon, much as Intel did for 86_64. In this model the large register file and all RISC gains would be manifest.

The normal (div 50) cross emulation downside performance can be cut since the data path and ALU logic is already 86 native. So the data-path and hi=bit management on both the 64/32 bit boundry is ok.

Since the downturn cannot but hurt Windows, already stressed in embedded and netbook marketplaces, this means that, if Intel prevails, the Wintel market place fragments. This will require unprecedented agility from Microsoft.

The real news is that this reinforces the failure of US patent policy and has created another closed US markets.

Have you ever used an Itanium box?

[PipingSnail]

Have you ever used an Itanium box?

Jeez, what an awful piece of cr*p. Sound of a vaccum cleaner, performance slower than an equivalent x86, Mhz for Mhz (timeframe: 2000/1). Well maybe no in benchmarks, but if you had a box, side by side, both running Whistler, you couldn't tell the difference.

I had an early pre-release Intel box (well, I had several) plus pre-release Visual Studio and compilers. I ported a 2,000,000 line C++ CAD app from 32 bit MFC to 64 bit MFC. We did the port, but the box did not sing. It was horrible.

5 years earlier I'd used Sun's Windows emulation environment running Windows apps on Sparcstation pizza boxes. That was better.

Itanium is much more of a dead platform than x86.

I don't know how expandable SPARC is, in terms of future bandwidth, but if its available its a reasonable legacy bet, given Sun have the emulator software.

Real shame they dropped Alpha. That was a good platform. Ahread of its time. We had one in our office early nineties, running Digital UX. Sometime in 90-94. That thing was fast, compared to the competition.

ARM would be excellent though, I'd love that to happen. Same platform for desktop, mobile, embedded, low power, high performance. All we need is multi-core (sorry, haven't followed it closely enough to know if that is the horizon).

I've only used 2 machines in my life that have sounded like vaccuum cleaners:
1) Motorola Exorciser, 6809 development system with 8" floppies
2) Intel development Itanium box (several of).

Both were [polite]not very good[/polite].

I wanna make a Year of the Linux Desktop joke here

[icannotthinkofaname]

Um...could this be the year of the AMD processor...? Is...uh...is Intel's chip finally going to fail in the market...?

Yeah, I know, I know, I have no idea what I'm talking about. I just thought I saw a vague parallel somewhere in there....

(Score:-1, Retarded)

Nvidia!!

[Anonymous Coward]

I think that these lawsuits threats are really directed to Nvidia, which have hinted that they want to enter into the x86 market.

It's a loud way to say, "only big boys allowed!".

Possibly About Next-Gen Technology

[David Greene]

Of course, none of us (except possibly any Intel or AMD legal types reading this) has any idea what's in ther license agreements. However, if the agreements don't cover certain extensions to technology, it's possible this is a move to leverage something in current negotiations around things like ISA extensions.

AMD produced SSA 4a which is mostly junk but scalar streaming stores are quite useful. Maybe Intel wants that.

Intel has SSE 4.1 and 4.2 but the real interesting bit is AVX. AMD almost certainly wants that.

AMD probably wants some of the Nehalem power technology. Intel may want some of that from AMD as well.

This is probably an ongoing negotiation spilling out into the public. It's not at all clear whether the current licensing agreement covers the above technologies and/or other things. These kinds of agreements get renegotiated all the time. Nothing to see here, move along.

Installing 64bit OS

[Anonymous Coward]

Lets all install 64bit OS as an act of protest against this action

Here today, gone tomorrow

[FyberOptic]

x86 competitors come and go. They always have a bit of a warm period where people welcome the competition and product variations. But eventually, the honeymoon is over. AMD had a good run and brought some decent chips, but I won't shed a tear when they disappear. Tis just how things go. Nvidia will apparently be taking their place.

Some people may not like the x86 architecture for whatever silly reasons, but the fact remains that it's going to be around for a long time. It's proven technology that is simply expanded as time goes on. First with 32-bit, now 64-bit. I don't know about you, but I very much enjoy the fact that, at its core, my computer is still the same machine it was 20 years ago, and can still run the same old software if I so chose.

Backwards compatibility is always the key to success. Just ask Microsoft.

Re:Here today, gone tomorrow

[benow]

I thought the key to their success was OEM lock-in...

Re:Here today, gone tomorrow

[marcosdumay]

"Some people may not like the x86 architecture for whatever silly reasons, but the fact remains that it's going to be around for a long time. It's proven technology that is simply expanded as time goes on. First with 32-bit, now 64-bit. I don't know about you, but I very much enjoy the fact that, at its core, my computer is still the same machine it was 20 years ago, and can still run the same old software if I so chose."

What a long line just to say "it runs Windows."

Re:Here today, gone tomorrow

[FyberOptic]

The point is, if x86 is still around in 20 more years, and progresses the way it has been, you'll still be able to run DOS, classic Windows, current day Windows, Linux, OSX, etc, even on those future machines. They won't be lost forever from lack of classic hardware to run them on, much like Apple has done. Twice.

Of course, by that time, they'll probably be fast enough to emulate all of that without much effort. But the point still stands!

Re:Here today, gone tomorrow

[bhtooefr]

Well, and Apple used emulation through each generation switch, too.

And, commodity x86 hardware has been able to do full speed 68040-based Mac emulation for quite a while. PowerPC emulation is still slow, but is usefully fast.

One other thing... sometimes running old OSes on a modern PC doesn't work anyway. Which means you have to use an emulator or at least a virtualization environment. (Enter DOSbox, which really is a good app.) Oh, wait, now we're emulating an early x86 PC, at a fast speed, even on old low-end hardware?

Stocking up

[Anonymous Coward]

I think I better stock up on my x86_64 chips
Just in case it becomes a collector item

4, insightful? Really?

[Anonymous Coward]

Are you kidding me, mods?

One reason you're wrong

[Anonymous Coward]

The move to 64-bit exposed the huge, sucky flaws of x86. You often get a performance increase by moving to x64. In a saner architecture like the PPC, 64-bit code isn't necessarily faster.

http://developer.apple.com/DOCUMENTATION/Darwin/Conceptual/64bitPorting/indications/indications.html

The only reason x86 is so good is because of competition and the gobs of money amd and intel have poured to keep it afloat. IBM didn't spend nearly as much as money on PPC and it got relatively good performance.

Re:One reason you're wrong

[AcidPenguin9873]

You often get a performance increase by moving to x64.

The single biggest reason for the performance bump when moving to 64-bit x86 was the increase in the number of both integer and SSE registers from 8 to 16. That means more data in registers, less data spills to/fills from the stack (fewer loads and stores). Yes, having a limited number of registers was a problem with x86 prior to 64-bit. But like I just said, they fixed it! What are you complaining about?

Does patent expiration apply?

[benow]

How old is x86? Could it not be considered a defacto standard by now? How would that affect the licensing? There is a chance to ditch x86 in general... perhaps the linux market is big enough? Could this be intel's grab for quick cash in down markets?

Re:Does patent expiration apply?

[beavioso]

The patents for the core x86 could be expired. Patents used to be valid for 17 years from the Patent Date (all patents before June 7, 1978) or 20 years from the filing date, which ever was longer (assuming it was filed before June 7, 1995). Patents filed after June 7, 1995 are now valid for 20 years after the earliest filing date. I picked up this information from: The Impact of GATT on Patent Tactics

In any event, the patents on the SSE extensions, assuming they exist, are probably still enforcable. So maybe AMD loses the ability to used 3DNOW, SSE, etc. No one really knows what's in the cross-licensing agreement. There could be patents on the fab process, or maybe they share industry secrets (is that ever part of cross-licensing?).

Re:Does patent expiration apply?

[bhtooefr]

AMD wouldn't lose 3DNow - that's their own extension set.

Losing SSE, OTOH, would be a problem.

But, there's also the i486 extensions that may not have quite expired yet, the Pentium extensions, etc., etc. And making an i386 chip in 2010 will be suicide. Even if it is fast.

Do that, and Eu will crack down on you like

[unity100]

there's no tomorrow ... That would totally consolidate all cpu market in one manufacturer's hands. all the profits in the world wouldnt save you from the earth shattering fines Eu would fine you. so, be cool, pipe down.

Laugh now

[symbolset]

The ARM netbooks and embedded devices are coming and there's nothing Microsoft or Intel can do about it except adapt and compete. The time when you could defeat a good technology with an evangelist is long gone since the public now knows evangelists are just shills for hire. The day a MS rep could derail a Linux deployment with a sneer has passed. Sorry Enderle, your day is done.

Intel will choose to compete and they have a good start because they started years ago. As the Atom die shrinks and gains SOC capabilities, its power requirements will come down. Maybe not to ARM levels, but to an acceptable level faster than ARM can bring their performance up to acceptable levels for a good user experience. Microsoft will choose to use the tools they have, and fail to adapt. That's what they do. They can't grasp a market that's abandoned the need for them. It's alien to their corporate culture. After they've failed in the market they'll buy an ARM OS vendor and try, but that's five years hence. and they'll buy five of them badly and integrate them poorly and we'll laugh at their ineptitude here.

Ultimately Intel will win this one but there will be some interesting side stories and products between now and then. Microsoft will lose because they choose not to port to the interesting new platform Linux runs on already, and so when the channels merge again they will have lost share. By then low power devices might be most of the share, at least for end user devices.

Re:Laugh now

[Blistovmhz]

Well said, if perhaps a bit hopeful. This may be the spark that begins the first real race within' consumer markets, between ARM and x86. ARM has historically been 5 years behind on raw horse power, but with the Cortex-A9, they're going to be competing for MHz with the Atom, and AMD has no cards on the table. I'd love to see what happens to the note/subnote market, with a 1.5GHz ARM CPU in direct competition. Something like this, but rivaling the Atom for speed? http://www.alwaysinnovating.com/touchbook/ Hey, maybe this will trigger Microsoft to release some of this "Surface" "technology" "they" have developed, to the real world? har har.

Re:Laugh now

[BlackSnake112]

I thought the article was about Intel and AMD. Two chip (CPU/chipset/GPU) makers. Last I checked microsoft was a software company. So if the hardware put out by Intel and AMD can run the software put out by microsoft, how/why are they going to fail? Even with the moronic decisions made by the top people at microsoft, the company is still going. As long as their software can run an an Intel, AMD, or some other company's hardware why should microsoft care. Do you think that Intel or AMD will not give microsoft or any other big OS company a list of instructions that their hardware uses?

So you are saying that ultra low power devices are what will take over. OK, I can sort of see that. For most regular users, that is fine. For the big corporations or research firms who have to churn through trillions of calculations per second I do not see the ultra low power stuff being used. They will cut power use where they can, but this low power hardware isn't up to their computing needs yet.

Some personal observations: On this atom based laptop, Ubuntu has the lowest power use or best battery time of 5 hours 57 minutes. OSX gets 2 hours 33 minutes, and xp gets 3 hours 55 minutes. That was the average time from 10 full charges till the laptop went to sleep on batters. All three were used for email, web surfing, and writing papers. All three were updated with all updates. I set the power settings as close as I could. Plugged in, power off the screen at 20 minutes and run everything else at max. On battery: 5 minute screen off, 3 min hard drive, sleep in 10 min. I was hoping that OSX would have more battery life then xp.

You don't know how this works

[symbolset]

Intel can shut down AMD's ability to use the X86 technology without giving up the AMD-64 technology if they can show that AMD defaulted on the agreement.

AMD can use the X86 technology and prevent Intel from using the AMD-64 technology if they prevail.

A court is going to have to measure this. The smart money is on a settlement but barring that Intel will win.

Let us meet here again in seven years, when the matter is settled.

cache use

[Anonymous Coward]

the smaller code size and hence more effective use of cache than with risc architecture made a bigger difference

x86 code has a higher density

Fools?

[nurb432]

1 - i never said it *sucked* i said it should be retired.

2 - If you are impressed with what they have done over the years to keep it alive and advancing, ( which i did give them credit for ) then you would be REALLY impressed if all that time and money was spent with a modern architecture instead. And no, tossing new instruction sets doesnt qualify as a fundamental change in architecture.

Power X Sparc

[marcosdumay]

Power X Sparc

Let the flame-wars begin!

Easy solution for AMD:

[Ant P.]

Just don't use the third-party manufacturing for x86 chips.

Instead use it to crank out huge quantities of cheap, high-speed ARM chips for netbooks, and bury the crippled Atom platform.

Sun + Red Hat

[SgtChaireBourne]

Sun and Red Hat ought to get together and promote sparc: open hardware for open systems. Being able to run RHEL and/or OpenSolaris is a good selling point, but it should also be possible to do both at the same time. You should be able to set up multiple containers each running either RHEL or OpenSolaris.

(I myself wouldn't say no to a low-end Sun or Fujitsu sparc workstation, if offered.)

Re:if they did that

[Anonymous Coward]

1. Take OpenSPARC
2. Improve performance of 64-bit operations.
3. ???
4. Profit. Literally.

Is that what *they* call a "money shot"?

[Anonymous Coward]

I've also seen ladies at a bar drinkdown their wiskey with a silver dollar remain in the cup. That a moneyshot too?

Re:I wanna make a Year of the Linux Desktop joke h

[bhtooefr]

Actually, on a serious note, that happened a few years ago.

AMD at their strongest with the highly successful K8 chips, Intel at their weakest with the Prescott P4.