Intel Makes 45nm Chip

dolphinlover writes "Intel announced today that it created its first microchip using the 45 nanometer manufacturing process that it says will go into its processors in the second half of 2007. Intel said that this development provides it with a 'considerable lead over our competitors in the 45-nanometer generation'."

Comments

Says You

[ackthpt]

Intel said that this development provides it with a 'considerable lead over our competitors in the 45-nanometer generation'."

Which means, what?

Predicitons for the next 18 months:

• Intel announces sucessful 45nm chip test, announces planned production for late 2007
• In mid-2006 AMD announces they have been quietly busy and production of 45nm chips to begin in 4th quarter.
• In November AMD is shipping quad core 45nm chips.
• Intel board scramble all resources to get chips out (even if at a trickle) ASAP, just get some damn thing out there, NOW!
• From hardware sites AMD chips receive rave reviews, slaying all competition and making overclockers wet their pants with joy.
• First Intel chips are tested and found to contain scarcely updated processors which still don't talk to each other very fast, run slow and, once again, are clocked so high you need a big fan and heat sink.
• Dell announce they are so pleased with Intel they're not going to use AMD chips (at all/any more.)
• In subsequent months Intel make improvements, now that they have a market presence, but watch their market share drop to 70% or lower.

i think it's somehow related to moore's law

Re:Says You

[level_headed_midwest]

I heard that AMD will be shipping their first 65nm products in late 2006 and have heard nothing about 45nm production.

Re:Says You

[georgewilliamherbert]

Unlike other fields, production ramps in semiconductor manufacturing are pretty easy to spot... the amount of new machinery and construction associated with a new process being deployed to a facility are hard to hide, and it's all over the trade press 18 months before stuff starts shipping typically.

AMD has traditionally been behind Intel on the bleeding edge fab stuff. Intel's dominated the fab tech race by six months or so for years and years. That is not changing here, as far as anyone I know of can see. AMD using SOI sort of blurs the line here, but in terms of process shrinks and the like Intel's ahead.

AMD's chips being better performers despite being behind some in chip fab is an important feature. But roadmaps based on imaginary pixie dust, in an industry where fabs cost $4 billion or so, are a waste of time even on slashdot.

Re:Says You

[Anonymous Coward]

I think you need to read the 5-point comments from this popular Slashdot story from earlier today:

Politics: Both Parties Ignore the Facts

Re:Says You

[geekoid]

"In mid-2006 AMD announces they have been quietly busy and production of 45nm chips to begin in 4th quarter."

i.e. waiting to get a leak from intel so they have a clue about what they need to do.

Re:Says You

[the eric conspiracy]

AMD has a co-development agreement with IBM and is planning to introduce 45nm parts in 2008.

Re:Says You

[uujjj]

AMD is nearly a full year behind Intel rolling out 65nm. Intel began volume production at 65nm last summer; AMD will be ramping up in the middle of this year.

While the parent may be joking, down below you'll find a lot of posts from AMD fanboys insisting that AMD must somehow be ahead. These fanboys are as clueless as the average tech magazine reporter. You can be quite certain that AMD will not be ramping up 45nm before Intel.

Re:Says You

[jonesy16]

Ah how I love shortsighted fanboys. What the hell has AMD ever been Intel to the market with? I can think of one thing, 64-bit extensions to an x86 processor, and Intel's implementation didn't have any bugs in it.

The pure size of Intel allows it to stay ahead in terms of manufacturing techniques and volume, much the same way that IBM stayed at the top for years in development. AMD does a remarkable job of making up for that by using more efficient processor designs though, but we'll see where both stand at the end of '06, cause the industry is sensing a change in the wind . . . even if slashdot can't smell it.

Re:Says You

[Kaladis Nefarian]

Cheesus christ man, stop saying ramping up, *please*.

Re:Says You

[uujjj]

"ramp up" is a common term in industry and elsewhere. it means indicates a company's acceleration of the rate of production. i don't quite understand the problem. it's like complaining "Cheesus christ man, stop saying the, *please*"

Re:Says You

[Anonymous Coward]

You meant "beaten to the market", right?
First 1GHz x86 CPU. That was AMD. Kind of a big one not to remember... fanboy.

Re:Says You

[doublebackslash]

They also were the first ot introduce dual core cpus in which each core can talk to the other via a direct link instead of the system bus and they were to first to include a memory controller onboard their processors. They also have more registers than any other x86 processor, even more in 64 bit mode (those few extras go a long way, belive me). They are figting an 800lb gorilla, they must win with brains, and I believe they have that in spades. I am certain the race will get intreresting in the next few years, but I feel that Intel cannot change direction fast enough to get out of some of their mistakes. Look how long they have kept netburst, despite superior technology in their Centrino processors. They seem to focus on the wrong improvements. That may all change, but I have a feeling that with their new VIIV (say 'vive') chipsets they will only serve to bing more bloat and latency to their allready crowded and busy chips. I wish them luck, it will spur innovation, but they are in a hole right now. Then again one new chip is all it might take to turn the tides, both in Pr and in tech. I wait with bated breath for both camps to make a big move.

Re:Says You

[OpiumSniper]

Yes, when AMD came to my school 45 nm was set for 2008, 65nm for later this year I believe.

Re:Says You

[HairyCanary]

If Intel began volume production at 65nm last summer, why are we just now seeing a 65nm processor? Does it really take that long from production to retail sale?

Re:Says You

[Anonymous Coward]

SOI?

I tried Wikipedia, but only found out that soi is thai for sidestreet. :(

Re:Says You

[Firewalker_Midnights]

I don't know about you... But I think that things might actually go this way, which is scarry. So do you think you can give me a winner for the Superbowl? I could use the extra money XD

Re:Says You

[georgewilliamherbert]

Silicon On Insulator; see the silicon on insulator wikipedia entry for a high level summary, or google the phrase for more details.

Basically, instead of a solid slab of silicon on which you fab chip components, you put a solid slab of an insulator (sapphire / alumina for example; see silicon on sapphire wikipedia entry) down and then an insulating silicon oxide layer, and then a thin layer of silicon on which you fab the parts. Since what's under the parts is insulator, rather than more semiconductor, it reduces the energy of switching and reduces the time to switch a transistor. Also reduces radiation effects on the semiconductor and other good stuff.

Re:Says You

[level_headed_midwest]

Well, they had the fastest 80386 too. That was really famous as it broke the AMD-Intel joint production agreement. I can also think of the dropping of the frontside bus, which Intel still has not done.

Re:Says You

[Anonymous Coward]

Thank you. It's certainly a nice explanation when I sort of understand instead of totally don't.

Re:Says You

[Overly Critical Guy]

So they'll be a year behind with 65nm and a year behind with 45nm?

Re:Says You

[Overly Critical Guy]

Intel is succeeding Netburst with Merom and Conroe later this year. Viiv is just a marketing name for a list of minimum system specifications for multimedia computers that happens to use a bunch of Intel-branded parts. It's unrelated to their chips.

As for who is in the hole, AMD is a year behind both 65nm and 45nm, and the Yonah is a laptop chip competing performance-wise with AMD's desktop processors. 'nuff said.

Those extra registers in 64-bit don't go that long a way (about 5%-10% on average last I checked the benchmarks). A lot of the 64-bit performance comes from the fact apps in 64-bit mode know that their chip will have at least SSE, which speeds things up. A 32-bit app optimized with SSE instructions can compete with 64-bit performance, since 64-bit is slowed down with the cache bloat and increased pointer size. 64-bit is hype designed to sell chips. It's not needed unless you actually have to access more than 4GB of RAM.

This is a benefit for the Intel Macs, whose baseline will always be the Core Duo that has SSE3, meaning all apps will be compiled with support for it, 64-bit or not. Until you need more than 4GB of RAM, 64-bit is overrated buzz that offers little.

Re:Says You

[Overly Critical Guy]

No, it doesn't take that long; everybody is lying. Intel started 65nm on Christmas when Santa dropped them down the chimney where the bunny suits live.

Duh.

Re:Says You

[stevesliva]

One interesting tear-down of an Intel chip (Core Duo, I think) suggested that Intel is not scaling geometries as agressively as would be suggested by the roadmap nodes, but gaining more performance from other process innovations such as strained silicon. It may be that Intel's allowing its gates to be a few nanometers longer than the should to reach the next node ahead of the pack.

Re:Says You

[stevesliva]

They would start with simpler chips first. Maybe even flash.

Re:Says You

[jelle]

Being ahead or behind in processors is not all about the nanometers you produce them in.

Transistors are not made of just 'nanometers'. There is more to a transistor than the process node of the steppers. In process technology, AMD is ahead by using SOI and soon added to that very silicon area efficient ZRAM based LIII caches.

Besides that, it's also about processor architecture (among other things the onboard memory controller).

Intel may ramp to 45nm before AMD, but AMD's 65nm will kick Intel's 45nm's butt just like the current AMD 90nm is kicking intel's 65nm.

Re:Says You

[Aadain2001]

That's only if use assume that the current processor design used by Intel is used on the new 45nm. Intel has learned from it's mistake with NetBurst (P4) design descisions and are finally heading back in the right direction. So, by the time the 45nm comes out, there will also be a new architecture to place on the new chips. We'll see how things are then.

Re:Says You

[jelle]

Intel may have 65nm in Fab D1D right now, with plans to convert 2 more fabs to 65nm while converting D1D to 45nm, they have many more fabs that are much farther away from 65nm. Many Intel chips will have to be made on processes older than 65nm.

AMD's new Fab65 just opened last October, is already generating fantastic yields at 90nm, and it is ready for 65nm and below (this year, sooner rather than later), and (even though AMD hasn't spoken about doing it), it would not be impossible to retool their Fab30 to 65nm.

And AMD's 65nm will have SOI, just like their 90nm does. Intel does not have SOI.

Re:Says You

[the eric conspiracy]

I'm betting 6 months.

Re:Says You

[Anonymous Coward]

Shit, buddy... Some of Intel's current top performers are now based on the age old Pentium 3 platform. That's right. Core Duo Pentium M is P3. It's two cores of heavily modified Pentium III. They took a step back because it was realized that P4 was in many ways a lame duck! And now those little P3 cores are the ones responsible for the competition with AMD's dual core chips, which wipe the floor with everything.

Re:Says You

[HairyCanary]

Cool, thanks, that's exactly what I wanted to know. Duh.

Re:Says You

[jelle]

"Intel has learned from it's mistake with NetBurst (P4) design descisions and are finally heading back in the right direction."

Exactly, they are going a step backward by going back to the Pentium-M with some modifications.

"So, by the time the 45nm comes out, there will also be a new architecture to place on the new chips. We'll see how things are then."

That is an awful lot of 'forward looking'. AMD will not sit still between now and then either, on either front (process and architecture).

Intel just canceled their Whitefield processor, the only one that ever was on their roadmap to sport an integrated memory controller.

AMD is improving their architecture continuously: Introducing chips with DDR2 support very soon (motherboard manufacturers have samples), just licensed ZRAM technology to add low silicon area very large Level 3 caches, will be introducing improved versions of the HTT interconnect that makes the multi-chip Opteron systems scale so well, is going to show us a quad-core version this year, etc etc.

I don't have to wait and see, it's too clear where this is going.

Re:Says You

[ls -la]

and they're still faster... imagine that.

Re:Says You

[Anonymous Coward]

i think it's somehow related to moore's law
 
It's Moop's law.

Re:Says You

[ebrusky]

So is this how Intel is going to keep up with AMD? Sure this next generation 45nm chip will run on a bit less energy and might be a bit faster. But they are spending hundreds of millions of dollars to get smaller when they should be redesigning their chip to get an actual advantage. AMD might not be pushing the size but why should they, their chips are already in the lead when it comes to design, cost, multicore, power usage vs. performance, and 64bit support. If I was AMD I would forget about going to 65nm and put my money into the 45nm switch and into developing my next generation features and designs. Let Intel blow all their money switching each time there is an incremental change in technology that has no real effect when it comes to their performance.

Re:Says You

[Overly Critical Guy]

Uh, no they're not. Even the laptop Core Duo is matching the Athlon64 3800+ X2.

Is there a requirement to jerk off over AMD when you sign up to Slashdot or something? These aren't the Pentium 4 days anymore. Intel owns the mobile market, and their future roadmap kills AMD's.

Re:Says You

[Anonymous Coward]

If Intel began volume production at 65nm last summer, why are we just now seeing a 65nm processor? Does it really take that long from production to retail sale?

Initially, yields are always insanely low. It takes several months of getting the kinks worked out to increase yields to the point where it's really cost-effective, and when they can get the binning (chips that run at a given speed) they want. I think they started at the very tail end of last summer, too, August-September time, so that seems about right.

For comparison, AMD is planning volume production at 65nm, around October of next year.

Re:Says You

[Anonymous Coward]

IT really doesn't matter. The AMD chip will work immediately and will be faster and more stable than the Intel part whenever the i chip hits the market. I think the world is getting wise to the intel hype.

Re:Says You

[dgatwood]

It is redundant by some definitions. If you accept those definitions as valid, it would rank right up there with "continue on" in the grammarians' hall of shame.

The definition in question is "to creep up". While most commonly used for plants (e.g. the roses ramped over the wall), this definition is the definition that most closely matches the use of the word 'ramp' in the idiom 'ramp up', and thus, you could easily consider "ramp up" to mean "to creep up up".

The problem is that this definition of "ramp" is not universally accepted. Webster's agrees with this defitinion, but AHED does not, hence the reason this is an area of grammatical contention.

It is, however, as you mention, a standard industry term, ill-conceived as it might be.

-----
This grammatical nit was brought to you by Hormel, providing wholesome meat product meals since 1936....

Re:Says You

[innosent]

I'd say you're mostly right. I think Intel will be positioned as the desktop / lower-end / lower-priced processor, and AMD will dominate the market for servers. Sort of a role reversal from the case before. Unless Intel comes up with something like AMD's direct connect architecture and hypertransport, AMD will continue to dominate I/O performance, even if Intel does eclipse them in the standard user benchmarks. I don't personally care who I buy from at work, but at least as it is now, the Opterons are the only thing allowed in the rack for new systems, and Intel chips for laptops. Desktops don't matter as much, so it's mostly Semprons at the moment (best bang for little bucks). As soon as Intel really abandons the MHz/GHz = performance mantra and realizes that the northbridge is their biggest drawback, they'll turn it around. The mobile chips and the Core Duo are a good start, but we'll see if they stick with them.

I saw a quote somewhere from an exec at AMD, who was asked how the Pentium 4 could be improved on. He said: "Use the Pentium III". The sooner Intel realizes he was right (mobile/Core chips are more closely related to the PIII than P4), the better.

Re:Says You

[jonesy16]

Well put, I look forward to the next year.

Re:Says You

[carl0ski]

Intel Only Rushed to use 65nm out of necessity, because it needed to breathe some more life into their Static Pentium 4

AMD doesnt have a lot of money so isnt going to rush something if they dont need it.

If Pentium M makes inroads on AMD they too may rush and push IBM to make 45nm Athlons early.

Re:Says You

[thegamerformelyknown]

I myself have begun thinking of Intel as a laptop cpu only. This is the only area that I can see them having any advantage really. Some may argue that because of their lower price point and brand recognition they will make it into the desktop market, but I know that I (and a huge margin of serious PC users out there) will recommend AMD, and eventually we will make an impact. If you ask me, the 70% that Intel holds is from users having no idea what they are getting, and just buying what looks good for the price (or by what the guy at the store said - he knows everything right?) or perhaps by the clock speed (which we all know doesn't really play a factor anymore). My personal view is that Intel will dominate as a Latptop/Mobile chip (which it should be happy with - this market segment is going to be huge - it is already actually...), and AMD will take over at LEAST all high end desktops and servers, if not all desktop levels.

Just my 2 cents :)

Re:Says You

[carl0ski]

its a bit of a shame really
Netburst was expected to reach 10ghz
with SOI , high quality transistors (such as they use in Pentium M) and 45nm Process
Possibly could have made the Netburst start Ticking.


Intel Producing a product very similar to AMD ( 10 IPC) pegs the X86 static again.
no choice between the long Pentium 4 Pipe takes variety and options from the market.
P4 for Intensive tasks and AMD games&multitasking (shortpipe more responsive at changes)(reference to single core P4 AMD64)


now you'll have 2 Brands one possible task, the other tasks forget it you have no options

Re:Says You

[carl0ski]

Intel Rushed 65nm to aid and lower the cost to make
The Pentium D Die Size is huge compared to a comparable Athlon X2
65nm also needed to possibly squeezed more out of Prescott.



Pentium M has mostly identical fundamental functions as AMD Athlon 64/Turion
similar Memory bandwidth, power consumption, IPC appears the same 10 IPC

The Pentium M also has a very simple characteristic aiding lower power usage
higher grade transistors, which adds to its price, so AMD may follow that suit in future also

The turion at equal clock is equivilant to Pentium M
with equal watt ph however,
combination of the elegant well thought out Centrino platform and DDR2, lower PM systems's overall power consumption


true Windows X64 truely is poor at utilizing AMD 64 and EMT64.

but my linux Desktop excels with 64bit compared to 32bit.
Vista will probably be better. and 32 bit OS can even use my 4GB is limited to 3.4 in WindowsXP and 32bit linux.


dont forget minimum recommended RAM now is 1-1.5GB
hypothetically mid-late this year 2-2.5GB,
could possibly quickly jump beyond 4GB next year.



P.S in the Windows X64 Drivers , lots of obsolete functions where found, Lecacy buses, ISA VESA and Parallel devices.
How many P4 and AMD64 systems have ISA VESA and/or Parallel interfaces.
as far as i can tell X64 was a patchwork rush job by MS

Re:Says You

[DrSkwid]

Idiot, usage drives Websters, not the converse.

"Ramping up production" is a widely used phrase so, pleaee, shut your mouth.

I can't be bothered to point out the mistakes in *your* post, but trust me, there are many.

Re:Says You

[Bulmakau]

It's funny though to see AMD falls for Intel's trap in the nanometer war, after the tried so hard to distance themselves from the MHz war. Chip companies should focus on making the chips stronger/faster. Not with higher clockspeed of smaller print - which historically proved to not be directly related in all cases.

Re:Says You

[somersault]

oh yeah because they managed to retool their fabrication plant in 2 months, and have no idea how to do processor design themselves. We must have forgotten. But of course as it says elsewhere here, they are not doing 45nm till 2008, and if they had already started retooling their plants, someone would have noticed.

Re:Says You

[somersault]

what do you mean? You mean that multiprocessor systems cant go faster on intensive tasks than a single system with faster clockspeed? Because when programs are written to do those calculations using multiple processors, then the multi-processor will kick the ass of the single core processor (mostly because single core processors at the moment have only slightly faster clockspeeds than their dual core processors). What point is there in having a 10Ghz processor if you can have 5 2Ghz processors doing the same job, with a *lot* less heat dissipation problems - if you lost the cooling on your 10Ghz processor it would melt itself. The faster a processor runs, the hotter it gets, and then due to the way they are designed, as the processor overheats it starts slowing down again. Also what's the point in having a 10Ghz processor if your northbridge/whatever cant actually keep up with the rate that your processor is processing data? It's not all about clockspeeds, as you should have noticed by now with Athlons being faster than equivalently clocked P4s

Re:Says You

[gears5665]

What is impressive about AMD from your post is that their EXECUTIVES answered a technical question correctly. It shows that the management is cluefull which will drive their products better than their competition.

Re:Says You

[networkBoy]

IIRC Intel _is_ moving the memory controller into the CPU.
-nB

Re:Says You

[htd2]

Fabs arn't everything.

Compare the Pentium EE 955 and the Athlon 64 FX-60 currently the top end of the Intel and AMD dual core workstation/gaming CPU range.

The Pentium EE is fabbed in a 65nm process while the Athlon is fabbed in a 90nm process.

The Pentium EE has a 169mm/sq die size, the AMD 199mm/sq.

Under load the Pentium EE consumes 286 watts, while the AMD consumes 225 watts.
At idle the Pentium EE consumes 189 watts, while the AMD consumes 149 watts

They both cost around $1000

The vast majority of benchmarks show the FX-60 easily outperforming the Pentium EE. More damagingly the slower Athlon X2 4800+ also outperforms the EE on the vast majority of tests and costs less.

As you can see from this being built in a better Fab does not necessarely impart any competitive advantage on the Intel processor and so it remains to be seen if being first into 45nm will help Intel get ahead of AMD

Re:Says You

[ackthpt]

I don't know about you... But I think that things might actually go this way, which is scarry. So do you think you can give me a winner for the Superbowl? I could use the extra money XD

I may be going out on a limb here, but I predict one of the teams will win the Superbowl, contrary to bookmakers odds which favor the Stones by a field goal.

Re:Says You

[darkmeridian]

AMD licensed magic technology recently where SOI processes would be used to reduce the density of on-die SRAM and DRAM. The company spokesperson said that tech was usually integrated in about two years. Hey, that's 2008! So again, Intel has higher fab tech but AMD may win on architecture yet again.

Re:Says You

[Slack3r78]

You're also ignoring Intel's plans to introduce Merom in 2H of this year. It's an all new architechture, so we don't know overly much about how it's going to perform, but all indications point toward even better IPC and thermal efficiency than Yonah.

The biggest improvement you'll see on the AMD side this year is going to be Socket M2, which adds DDR support along with a few other tweaks, but nothing too drastic. The ZRAM deal *just* happened, so don't expect to see any results from that for some time to come.

AMD builds great chips, but Intel was something of a sleeping giant. AMD should maintain a lead through the end of 2006, but it's going to be much narrower after Merom's release than it has been in years.

Re:Says You

[Slack3r78]

From the roadmaps I've seen, not until at least 2008. 5 years is a long time to allow AMD to embed themselves in the market as *the* chip for servers.

Re:Says You

[default luser]

Intel has been first on the introduction of new process moastly because they have the money and resources to push the very limits. Retooling , especially when the new process equipment is just becoming avilable, is extremely expensive, and retooling requires shutting down production lines.

Intel can do this because, assuming yields are high enough, they make their money back on their short-term production-per-wafer lead. AMD, on the other hand, only has 3 fabs, and shutting down lines when your distributors are already crying for more chips is a hard sell.

There are some benefits to being second in process land. While Intel has been foolhardy, jumping from one process to the next, AMD has been more patient and methodical.

* AMD was the first of the two to incorporate a copper process (.18u). Intel waited until their .13u process to use copper, because they claimed they would not need it until that point. AMD arguably got much better speeds from their .13u Cu process than Intel managed, thanks to their decision to integrate copper earlier.

* AMD elected to integrate SOI into their .13 micron process, in anticipation of huge increases in leakage current in the upcoming 90nm process. Intel decided to wait until their 90nm process to implement strained silicon to reduce leakage current, and that was a horrible mistake. Only now with Intel's 65nm process are we seeing results from strained silicon competitive with SOI (for example, Cedar Mill is lower-power, both from reduced voltage and reduced leakage).

In addition to process leads, Intel is forced to be wasteful trying to stay one step ahead of their competitors. Back in 2002, even though Intel was backing Itanium full-throttle, and even though AMD had yet to publically announce x86-64, Intel was already working on an x86-64 part to counter AMD. Thus, it was no surprise when it was discovered Prescott shipped with x86-64 capabilities already built-in, but disabled. Once Intel realized the momentum behind x86-64, the extensions were enabled in later Prescott revisions.

Of course, once Intel enabled the extensions, they basically threw away the Itanium and their future total control of the 64-bit market. All this hopping and skipping about costs Intel tons of money. Eventually, it will catch up with them.

Re:Says You

[HuguesT]

For the record I've been doing my computing on 64-bit CPUs since 1994 (alphas). At that time alphas were 2-5x faster than anything around them (200MHz alpha vs 75Mhz pentium), so there is no inevitability about 64-bit being slow or even slower than 32-bit, rather the opposite.

The thing is while Intel definitely have a good CPU to put in laptops, their desktop chips still aren't that great, to be polite. The core-duo goodness doesn't translate there, at least not yet. Also Athlon64 chips have been good with power for a long time, so much so that there is not much difference between a desktop Athlon64 and the Turion. The Turion typically gets 80-90% of the autonomy of the Pentium-M. We don't know about the Core Duo yet AFAIK.

Remember that the benchmarks that you are seeing comparing Athlon64 and Intel dual-core are all 32-bit only. From my own benchmark I easily can get +20% from swithching to 64-bit on the same Athlon64 CPU. I don't consider this overrated. Cache problems are definitely overrated though.

Also memory access benefits come in earlier than the 4GB mark. In 32-bit land, the memory is typically shared 2GB/2GB between userland and kernelland. You can change the ratio but there's not much option on this typically. This goes away in 64-bit mode. All your apps can address as much as they need.

I don't know about you but all my new machines come with 2GB+ these days. In 2007 everything will need to be 64-bit. I wouldn't buy a machine today that wasn't 64-bit if it were my own money. I would buy it with 2GB outright and put an extra 2GB in as soon as I have the money.

Re:Says You

[Anonymous Coward]

Integrated memory controllers have drawbacks also, especially in scalability, when you have an integrated memory controller, you will always lag behind and use an old memory technology.
See for example how much time took AMD to move to DDR2.

The next Intel chip Merom is going to be based on the Pentium 3 architecture, and the new presler (pentium4 based) is the last model that is going to be based on P4.
P4 was a good architecture for couple of years and was failed because of heat problems that couldn't anticipated when P4 was designed. Just to understand, the Prescott model (first 90nm P4) was targetted to work on clock freq of 5GHz, at the end it was barely hit the 3.8GHz because of TDP (heat dissipation) problem.

CPU building takes 4 years. It means that Intel decided to return to the P3 architecture 3.5 years ago. and we will see the results of it in half a year (the Merom family that includes Merom for laptops, Conroe for desktops and Woodcrest for servers).

Re:Says You

[mrmeval]

That's wet their panties.

Re:Says You

[the eric conspiracy]

I've seen discussions that early Merom/Conroe parts have been disappointing in terms of thermal budget as that Intel may not be able to release until 3rd quarter. The touted four issue design may also turn out not to scale as well as hoped. If Intel botches this they will find themselves in danger of being #2 in overall CPU market share by the end of 2007.

As far as the drawbacks of integrated memory controllers - Intel's DDR2 bandwidth is significantly lower than AMD's DDR bandwidth. What good is a new architecture if you are bottlenecked elsewhere? If it isn't an advantage, why are integrated memory controller's on Intel's roadmap for 2008?

Re:Says You

[carl0ski]

fool i am aware that clock isnt the only thing.
however if you thought before speaking
Intel P4 needs High Clock speeds to negate
since its lack of Floating Point and IPC performance

a future P4 at 10ghz may have negated the crappy perforance, and considerably outperformed beyond 3.2+ghz Athlon 64


While Athlon XP outperformed P4 at lower speeds for a long time

once p4 reached 3ghz
that was the point p4 3hz negated the disadvantage of crappy performance.

P.S i own 3 Athlons my purchases were based on performance not clock

dont ever accuss me of such things ever again.
the AMD gods will question me

Re:Says You

[carl0ski]

The Pentium EE has a 169mm/sq die size, the AMD 199mm/sq.
theres you competitive advantage.

P EE used to have a die size of over 200mm/sq
which plates them less processors from expensive wafers
and overall production costs increase over AMD. 199mm
so AMD was capable of more chips per wafer

Intel shrinking to 169mm means they can either lower the $1000 price tag or make a larger profit per processor sold.


P.S i cant really say how much the entire manufacture of a processor costs, however i can imagine the Athlon X2 series cost more to make that twin Core Pentium D series
SOI, elegant dual core design
compared to Intels twin core (2 presscott glued onto one ceramic chip.

Re:Says You

[jonesy16]

That's not necessarily a benefit that should be bragged about. It's merely a difference in architecture design, and AMD does still have a data bus, called hyperlink, which connects it to the Southbridge. The flaw in their design involves multi-cpu systems accessing the southbridge. If both processors need access, one of them has to stop to allow the other one to get through. Intel's implementation shares the bus, so even though it could be a bottleneck, both processors can still continue to function regardless of what the other one is doing. This may not affect you if you process in memory only, but if both processors need disk access for example, then effectively both AMD processors function at half speed since only one can get to the PCI bus at a time.

What AMD really did was got rid of the northbridge connection by putting it on the silicone itself. The bus is still there, it just doesn't have to go as far.

Re:Says You

[somersault]

hehe XD well sorry if you felt accused, or if I accused you, cant remember now.

I still think that even if you keep increasing the clock speed, without better IPC *which I jus looked up to see what it meant, hehe* then you're going to hit a limit - what does it matter how fast you can process, if you cant get your data in and out fast enough? Unless you're doing a simple input that gives a simple result, but has a lot of processing to do to get the result, which is maybe why the Pentium wins on certain types of encoding? Hmm.

Re:Says You

[htd2]

Sure the EE has a smaller die size 169 vs 199 so in theory using a similar sized wafer Intel should get a higher yield because they have more dies per wafer. However that assumes a parity in terms of failure rate between the AMD 90nm process and the Intel 65nm process.

Manufacturers guard their effective yield rate stats very closely so its impossible to say if Intel or AMD has the advantage here. However the 65nm process is a very new one and it is likely that as in Intels move to 90nm their effective yield will initially start low and then increase, while AMD with the a more mature process have a higher yield.

Since both CPU's cost about the same it would appear that at least for the moment neither vendor has a yield advantage.

It is also obvious that Intels strategy of being as early into a new process is driven largely by necessity because the facts plainly illustrate how far behind the FX-60 the EE is on all other metrics except price despite its apparent process advantage. AMD on the other hand have the luxury of being able to wait until the new process matures a bit before delivering product using it.

Re:Says You

[jelle]

"You're also ignoring Intel's plans to introduce Merom in 2H of this year."

The Merom, "shows higher power consumption levels than the company anticipated" and if they can fix that, it will be introduced at a paltry 2.33GHz in September... The Merom is already obsolete. http://www.tgdaily.com/2006/01/11/merom_to_be_laun ched_in_september/

The AMD FX60 dual-core available _today_ is already at 2.4Ghz in 90nm silicon, and will definitely switch Fab36 to 65nm before Merom is out.

And this page http://www.penstarsys.com/editor/company/intel/con roe/optimism_2.htm explains why ZRAM is the answer of AMD to Merom's large cache, but without needing to match the high manufacturing cost and power problems of the Merom...

ZRAM is in line with Merom, we're talking about the end of this year, or later, here (if Intel can't fix the problems with Merom, it's schedule will slip again.

Oh, btw M2 adds DDR2 support, not DDR, and that's hardly a tweak.

I must need glasses

[punxking]

When I first read the headline I thought it said a 45mm chip, which is considerably less impressive.

Re:I must need glasses

[CrazyJim1]

"When I first read the headline I thought it said a 45mm chip, which is considerably less impressive"

I thought they invented a 45mm clip. No man, take my wallet, I don't want intel inside.

Re:I must need glasses

[GuruBuckaroo]

Talk about your Smart Bullets...

Re:I must need glasses

[msauve]

When I first read the headline I thought it said a 45mm chip, which is considerably less impressive.

Not really. Creating a >2000 mm^2 chip without any flaws, with the expectation of being able to eventually make a profit on them in the consumer market, would be quite an accomplishment. Such a large die area would not only result in low yields, but present serious obstacles in power consumption and heat dissipation.

For comparison, the Pentium IV 600 series has a 135 mm^2 die area. If I'm not mistaken, yields fall exponentially, so no only would they be able to produce only 6% as many chips on a wafer, but could also expect a greatly lower number of working ones from the total.

Re:I must need glasses

[Anonymous Coward]

No actually, yields do not fall exponentially. If they did, nothing made today would work at all. Yields are largely independent of feature size (after the process has been around long enough to work out the bugs). Actually, by making the die smaller, the yield should increase somewhat, but they will certainly pack more stuff onto the die so the size will remain similar.
What I'd worry more about is leakage current, which is a strong function of gate oxide thickness and lifetime reliablity. These are the tough hurdles, not yield.

Jobs's strategy?

[vijayiyer]

Perhaps this what Steve Jobs referred to when he talked about the efficiency of future chips in Intel's roadmap?

Re:Jobs's strategy?

[justsomebody]

Very bad strategy then.

Intel is seriously droping out of being serious contestant in current technology for quite some time now. It's just like droping a ball to get it back in one of the next few seasons, hopefully other team will remember your showing of good will, act then just as you did now and now fight back. Yeah, right.

Re:Jobs's strategy?

[richman555]

I beleive so, as much as AMD fan would like to admit, Intel has the upper hand for future chips. I guess the deciding factors for success will be raw speed vs. overall processing (more cores) vs. low power (portablility). I think with this past years increase of laptop sales, it shows that these people have a little more in mind than just having the fasted possible processor. The decline of the desktop is coming, and Intel seems to be ready.

Re:Jobs's strategy?

[cyfer2000]

I am only hoping Steve Jobs hasn't announced when he could get 4G CPU from intel, otherwise there will be no 4G CPU at all.

Re:Jobs's strategy?

[Overly Critical Guy]

Intel is seriously droping out of being serious contestant in current technology for quite some time now.

Uh...do you have any points to back up this assertion? Intel's future roadmap is ahead of AMD's. AMD doesn't even have 65nm chips out, and Intel is already talking about their 45nm plans. It actually looks like, for 2006 at least, it's AMD that's behind.

How is Intel "seriously droping out?" They're already ahead.

Re:Jobs's strategy?

[snuf23]

In particular Jobs was talking about the descendants of the Pentium M - the Core series of processor, which offer excellent performance at lower power consumption. Which is evidenced by the usage of the Core processor in both of Apple's current Intel based products.
The Pentium M came expanded on the Pentium 3's design, rather than going for the "more GHz is better" approach that the Pentium 4 did. It is more efficient than a Pentium 4 of the same clock speed and uses less power.

Re:Jobs's strategy?

[MobileTatsu-NJG]

"Intel is seriously droping out of being serious contestant in current technology for quite some time now."

Not in the laptop space.

Re:Jobs's strategy?

[Overly Critical Guy]

I beleive so, as much as AMD fan would like to admit, Intel has the upper hand for future chips.

And now we see the reason behind AMD's lawsuit. If you can't beat 'em, desperately sue 'em to make them look bad.

Re:Jobs's strategy?

[twiddlingbits]

No. The deciding factors will be power, heat and board space. The data centers are running out of power and cooling and floor space. Desktops are not the be all of the market. I think soon you'll see desktop and server chips diverging in architecture and speeds.

Re:Jobs's strategy?

[openldev]

Yes, and I find it funny that AMD's 90nm processors run faster than Intel's 65nm processors ...

Re:Jobs's strategy?

[Anonymous Coward]

The point is that they don't. Intel's dinky little laptop Yonah processor matches performance with a desktop Athlon 64 3800+ X2. Ouch. When the desktop Conroe hits this summer, it's over for the AMD fanboys.

Re:Jobs's strategy?

[Dun Malg]

AMD doesn't even have 65nm chips out, and Intel is already talking about their 45nm plans. It actually looks like, for 2006 at least, it's AMD that's behind. How is Intel "seriously droping out?" They're already ahead.

All the die shrinkage in the world can't make up for design shortcomings. A 45nm Pentium M is still going to fall short of an Athlon 64 in memory access, because no matter how small you make it, it still doesn't have an on-die memory contoller. Die shrink just gives you room to maneuver. It remains to be seen if they actually do anything with it. Their track record so far has been uninspiring.

Re:Jobs's strategy?

[Hal_Porter]

Not really.

Fabrication technology depends on how deep your pockets are, and Intel has very deep pockets. AMD beat them with a better architecture in the P4/Athlon generation by enough to nullify their fabrication advantage. And Itanium (where they probably spent their design resources) looks like it will die.

But in the long run they have a new x86 design team in Israel, and if you couple that with the access to better fabs they may win out in the long run. And if Itanium dies and they only have to build x86 cores, they have more design teams than AMD. They could probably design 3x as many x86 cores as AMD in a given period. E.g at the moment they have P4, Pentium M, Yonah, and soon the NGMA ones.

On the other hand, deep pockets implies a big company, and big companies tend to make bad choices (like Itanium and Netburst) when it comes down to technical strategy.

But from the perspective of Steve Jobs, it probably came down to a cost issue. Apple could continue to support PowerPC, where the cost per die is lower, but they probably have other costs like paying chipset designers*, or they could pick Intel at a steep discount for some period and no other costs.

And from Intel's point of view, they have another closed platform to replace Itanium and complement the open PC one.

* Actually, it's might be even worse than that if when Apple signed up to PowerPC they agreed on some kind of financial commitment to share the costs of keeping it alive. Back when it looked like it would replace x86, that was probably a good idea. Now that x86 seems to have won on the desktop, and PPC is targeting games consoles and embedded stuff, it's not.

Re:Jobs's strategy?

[somersault]

because their new R&D place in Israel was smart enough to go back to the PIII rather than try to build on the PIV. Looks like Intel's main R&D group are a bunch of morons :s or at least headed by morons. No offence you main Intel R&D dudes =P I do like Pentium Ms, and have one in my work laptop. The laptop market these days is growing, but I wont buy one myself for a while, since usually when I upgrade I want fast and cheap, so I upgrade my desktop (unfortunately since starting work fulltime I've started to find my spare time more precious and havent played any games since HL2 came out :s )

Re:Jobs's strategy?

[javaxman]

Perhaps this what Steve Jobs referred to when he talked about the efficiency of future chips in Intel's roadmap?

Well, it's at least part of it. Steve Jobs was shown product roadmaps and target development/production dates that the nobody outside of Intel gets to see without some heavy-duty reasons and NDAs. So it's not so much Jobs' strategy as Jobs' inside information.

He was also talking about the fact that IBM was going nowhere with portable chips, and Freescale was under-delivering on their "G4" chips ( as in, behind last year's schedule by over a year ), leaving Apple with not many good choices. Sure, Apple could have gone with AMD, but why go with a company sure to have supply issues when they're behind their competition anyway ?

Short answer to your question : yes.

Why do they always screw up Moores Law

[rminsk]

The new chip makes good on Moore's Law, an industry maxim set forth by Intel co-founder Gordon Moore that stipulates the number of transistors on a chip -- and therefore its processing power -- doubles roughly every 18 months to two years.

Who added the "and therefore its processing power" to the quote? Was it the reporter or someone from Intel? Moores law has nothing to do with processing power.

Re:Why do they always screw up Moores Law

[TheRaven64]

Double the transistors means double the number of cores and double the cache, which means (roughly) double the processing power. It doesn't sound too wrong to me. Now the focus is more on having lots of cores and less on clock speed, Moore's law might well gain that corollary.

Re:Why do they always screw up Moores Law

[Kjella]

Who added the "and therefore its processing power" to the quote? Was it the reporter or someone from Intel? Moores law has nothing to do with processing power.

In popular science, Moore's law is used to describe anything that resembles exponential growth. Not only that, it is applied without regards to whether the underlying technology scales in an exponential way, as long as it appears to have done so for a certain period of time, meaning whichever period gives the desired results. "Computers" and any part thereof seem to qualify on historical merit. Transistor count? Clock speed? HDD size? Take your pick. In this case I presume the journalist felt people didn't know what transistor count meant - but sorta drew the conclusion that it is related to performance or something - at least the Intel people talked about it alot, so it must be important...

Re:Why do they always screw up Moores Law

[PunkOfLinux]

double the number of cores? me thinks not. If that was the case, we'd have 32-core processors bye now -- which, needless to say, we don't.

Re:Why do they always screw up Moores Law

[shrewd]

we do have 32 core processors, they aren't available on your average street corner though....

Re:Why do they always screw up Moores Law

[Firehed]

And are they still under the impression that it's a law? Despite the name, it's just an observation that's held true for quite some time. It's not like gravity is going to stop working if the earth gets lazy, unlike transistor count and engineers. In any case, it didn't even start at 18 months per doubling, it was just settled there after going from 12 to 24 (or the other way around).

Re:Why do they always screw up Moores Law

[Overly Critical Guy]

Right, it has to do with the increase in transistors, AS STATED IN THE FREAKING QUOTE. The end result is typically an increase in processing power, hence the "therefore" phrase stating such. All factually correct.

Methinks you're just being pointlessly anal retentive for mod points' sake.

Re:Why do they always screw up Moores Law

[kestasjk]

Who added the "and therefore its processing power" to the quote? Was it the reporter or someone from Intel? Moores law has nothing to do with processing power.

Moore's law has everything to do with processing power.. Transistor count is related to processing power in the same way the number of bricks are related to the size of a house. You can get a larger house out of the same number of bricks if you design it well, but more bricks means a larger house.

Re:Why do they always screw up Moores Law

[msbsod]

But with smaller transistors you also have smaller distances in between, thus reduced emission (power). The question is which effect has the larger gradient.

Re:Why do they always screw up Moores Law

[Davorama]

Isn't this what the new Sun chips are about? Hmmm, found this eweek artical about it

The chip (formerly code-named Niagara), with eight cores that can process four instruction threads simultaneously,

So, it's not 32 cores but 32 threads on 8 cores. Not what you or the grandparent were looking for exactly but it's something worth mentioning.

Re:Why do they always screw up Moores Law

[freidog]

Double the transistors means double the number of cores and double the cache, which means (roughly) double the processing power.

Historically that has never been true. Though I suppose with multi-core chips we're getting to where you can say 'double the processing power' (which is a very vauge and misleading term, since it offers no where near double the performance of a chip 99% of the time) by doubling the number of cores on a die.

But from when Moore uttered his 'law' (more like loose estimate based on current trends) until about 2004 that 'double the processing power' was simply a complete misunderstanding of Moore's law. Prescott has about twice the number of transistors as Northwood, does it have 'twice the processing power?' I think not.

Regretably the incredible multipling cores may give some credibility to the 'doubling the processing power' line that people have been (miss)quoting from Moore for the last 30 years. Oh well, being wrong for three decades didn't stop people from thinking chips got twice as fast every 18 months, being mostly wrong about it now can't be much worse

Re:Why do they always screw up Moores Law

[Savantissimo]

IIRC, the first decent microprocessors were about 30K transistors and the next generation of chips should fit 300M transistors easily. So with enough crack and enough money, you could theoretically make a 10,000 core chip now. A practical device would need a network and memory on chip, but still - thousands of cores. If the inter-core communications were asynchronous, the cores could run at amazing clock rates, too, since the longest signal paths would be only a picosecond or two.

Re:Why do they always screw up Moores Law

[PetiePooo]

Methinks you're just being pointlessly anal retentive for mod points' sake.

Anal-retentive is properly spelled with a hyphen.

Nobody laughs when I tell them that (I do get an occasional groan), but it never gets old to me!

Re:Why do they always screw up Moores Law

[Ekarderif]

But gravitation is only a theory! Just like evolution! It might be that everyone has been wrong and we've been stuck on Earth due to pixie magic from the inner crust, and within five years the magic will run out and we'll be hurling into the sun.

Re:Why do they always screw up Moores Law

[fitten]

Even worse is that people actually think it's some sort of technology limitation. It's based on economics more than anything else. The Fab companies can afford to upgrade every 18 months (it's very expensive) because that's the rate at which most consumers (be they businesses or individuals) are willing to upgrade their existing equipment.

Moore's "Law" is just where the lines "how fast are people willing to upgrade" vs. "how much does a new fab cost and we remain profitable" cross. It's more of "Moore's Economic Observations" than anything technological.

Re:Why do they always screw up Moores Law

[dfghjk]

If you double the number of bricks do you double the size of the house?

Re:Why do they always screw up Moores Law

[ivan256]

"Computers" and any part thereof seem to qualify on historical merit. Transistor count? Clock speed? HDD size? Take your pick.

Ok. I pick clock speed, and the last 18 months.

No.. That didn't work... How about instructions per second over the last 18 months...

Crap. That doesn't work either.

In this case I presume the journalist felt people didn't know what transistor count meant - but sorta drew the conclusion that it is related to performance or something

So instead of doing his/her job and explaining it properly, or researching it for himself, he passed along invalid information... In the form of a misquote no less! Bad journalist! No cookie!

We win!

[pat_trick]

Yes, by announcing that we have made one chip at 45nm, we obviously win! ...nevermind that it probably doesn't actually run anything. We haven't made a motherboard for it yet.

Re:We win!

[Anonymous Coward]

It is obvious intel can do no right with all the drooling amd fanboys... Why do people like to attach themselves to corperation? Buy their products, don't buy them but do not sink to zealotry. It simply makes you look silly. Both companies exist for one reason to sell.

Re:We win!

[diegocgteleline.es]

The also announced that they've shipped 1M of core duo chips.

Obviously, both announcements are trying to stop the stock fall (which started when Intel announced that they haven't hit the sales predictions). Not that AMD doesn't do the same, they've used the same trick several times just like any company.

It's how capitalism works: People is free to invest were they want and do what they wan but what happens if we try to influence what people wants?

(By the way is ironic how capitalism is good because it allow companies to have more freedom to do things in their "system" but freedom is being limited in the US for "security purposes". I'd guess that if I'd propose to take away some of the freedom that capitalism has to fix some of the problesm the world has like poverty etc people would call me communist or something)

Re:We win!

[bipolarpinguino]

Workers of the world unite!

Re:We win!

[Anonymous Coward]

Ah, another angry AMD fanboy makes "funny" anti-Intel conjecture to please the mods. Meanwhile, AMD has yet to even get around to 65nm.

Yes, Virginia, chip manufacturers like Intel make roadmaps far into the future (all the way up to 2011 in this article).

Re:We win!

[Xendarq]

None of the press seems to indicate what socket standard the new chips will conform to. If it's Socket T or 479, for example, then they won't require new motherboards. If it's a new socket, then, point taken... although generally the lag to build new m/b's is virtually nonexistent.

Re:We win!

[Anonymous Coward]

I'm thinking they probably will market new motherboards. Intel is in the mobo business as well, and it would behoove them to sell a bunch of "upgrades" with the new chipsets.

Re:We win!

[Anonymous Coward]

Or it could just be a joke? Not everything is for/against.

Holy shit!!!

[IAAP]

A 45 Nautical Mile Chip! Where the fuck are they going to put it?!?

Whaaa? n.m.? Nano Meee....whaaa??

Oopps! Sorry!

Re:Holy shit!!!

[holywarrior21c]

you are asshole... bump....again...

Re:Holy shit!!!

[ePharaoh]

Actually it's a 45 nautical mile ship!

And using the latest in ship fabrication, it is made of asbestos instead of silicon.

Re:Holy shit!!!

[HermanAB]

No, it is 45 Newton Meters - a very forceful chip...

Re:Holy shit!!!

[modecx]

Actually it's a 45 nautical mile ship!

Dang, if you were in the middle of that ship you could look either way and it would disappear into the horizon, unless it was also absurdly tall, of course!

When they start building ships to take into account the curvature of the Earth, I'll officially be scared.

Re:Holy shit!!!

[Savantissimo]

Are you saying it has the impuse to kick ass?

What about AMDs 45nm???

[Saven Marek]

Intel said that this development provides it with a 'considerable lead over our competitors in the 45-nanometer generation'

Recent news says AMD is readying 45nm for production now, or last I read it did. That might have been in december, so I don't know if it's still current. If it is then it seems Intel is behind the 8 ball again and this is hype.

Re:What about AMDs 45nm???

[taskforce]

I think you'll find that it was 65nm which AMD were readying for production, which Intel have been using since last year. Yet AMD Desktop cores at 90nm with SOI still manage to outperform, underconsume power, and underemmit heat compared to their counterparts.

It just goes to show that design does play a part in making a chip, and not trying to cram as many transistors as one can onto a die.

Re:What about AMDs 45nm???

[TheRaven64]

Wow, the AMD fanboys are out in force tonight. From this source, which is three months old and so relatively recent:

• AMD's new fab, Fab 36, supports 300mm wafers (like Intels have for some time).
• It uses a 90nm process (Intel and IBM have been on 65nm for some time).
• It will transition to 65nm by the end of 2006.
• It will use 45nm and 32nm processes by the end of the decade.

It doesn't really sound like Intel is playing catch-up here.

Re:What about AMDs 45nm???

[AllInOne]

They may not be playing catch-up but Intel used to ignore the competition. Now they have to have press releases that acknowledge the competion: "considerable lead over our competitors".

Sounds like they are at least looking over their shoulders at AMD to me.

Or maybe even trying to speak to investors to prop up the stock price...

Re:What about AMDs 45nm???

[Anonymous Coward]

yes and no .

even though static power consumption was a significant issue in 90 , it becomes really huge in 65 and below. what intel is saying is that they have understood the issue and have a very good grasp of the solutions needed to mitigate it .

no matter how good your design team may be .. clock gating/powergating/multi supply/multi threshold etc can save you only so much without adversely affecting performance. in 65 and below you really do need to have you your transistors profile changing ever so delicately to strike that performance tradeoff and AMD is way behind in understanding these next gen relations in comparison to intel.

next gen is a lot more than about craming in more transistors.

Re:What about AMDs 45nm???

[Overly Critical Guy]

If it is then it seems Intel is behind the 8 ball again and this is hype.

Only in the Magic Land in the minds of AMD fanboys where AMD is the king of the land with a beautiful queen by his side and Intel is the dastardly wizard trapped in the tower with giant trees ruining his shit. Meanwhile in Reality Land, AMD has yet to get around to 65nm, Intel's low-power mobile chips are competing with high-end AMD desktop chips, and Intel's roadmap already has 45nm on the horizon. But right, anything to make a quick conclusion in a Slashdot post so you can claim Intel is behind and this is hype, in order to restore your holy AMD faith!

Re:What about AMDs 45nm???

[Overly Critical Guy]

The dual-core Yonah consumes less power at 100% usage than the Athlon64 3800+ X2 does when idle while competing with it performance-wise. A low-power mobile Intel chip competing with a high-end desktop chip from AMD. Good design, indeed. The Merom chip coming later this year is actually expected to almost half power usage compared to Yonah, which is even more impressive. AMD has yet to get around to 65nm. It's clear 2006 is swinging toward Intel (as much as the AMD fans on Slashdot hate to hear it).

Re:What about AMDs 45nm???

[mczak]

(Intel and IBM have been on 65nm for some time)

You're right about intel (though "for some time" seems a bit exaggerated if you only count shipping products - something like a month). intel really is ahead 6 month. However, IBM is not using 65nm tech with any of its shipping products (that I'm aware of). In fact, their power5 only transitioned to 90nm very recently, though the powerpc g5 transitioned to 90nm somewhat earlier than AMD transitioned their chips to 90nm I believe. Some time ago rumours said the first dual-core g5 chip would be 65nm (the powerpc 976) but obviously this was just that - an unsubstantiated rumour, the powerpc g5 dual-core materialized as a 90nm chip (970mp).

Re:What about AMDs 45nm???

[mcrbids]

  Wow, the AMD fanboys are out in force tonight. From this source [computerworld.com.au], which is three months old and so relatively recent:

        * AMD's new fab, Fab 36, supports 300mm wafers (like Intels have for some time).
        * It uses a 90nm process (Intel and IBM have been on 65nm for some time).
        * It will transition to 65nm by the end of 2006.
        * It will use 45nm and 32nm processes by the end of the decade.


What I really just don't get is, how come the AMD chips, with this clearly antiquated technology, manage to handily outperform the newer, "high-tech" Intel chips, at better prices?

I just bought several nice, dual-proc opteron servers, and these puppies SCREAM!!!

Re:What about AMDs 45nm???

[kadathseeker]

The real question is whether or not this will actually lead to any improvements in performance, heat production, and energy efficiency. It is still too early to tell, but I applaud Intel even though I believe AMD's current line surpasses Intel's in most ways. This is what a free market brings, competetive innovation that allows a consumer to get good, new technology (a more even production distribution would be nice). This is why I like the graphics card market - sure, the insane competition between ATI and NVidia mean I have to buy a new card every few years to keep up, but we will have the Matrix in no time. It's a small price to pay in the long run.

Re:What about AMDs 45nm???

[Anonymous Coward]

server1: YEAAAAGHHH!!!
server2: GEEYAAAHH!H! Did you see that guy?
server1: Yeah, what a tool.

FIN

Re:What about AMDs 45nm???

[Anonymous Coward]

The significance in being first on 65 and 45 nm is not performance, but costs. Intel is making a lot of money on being on 65nm now. They can make a lot of more small chips and utilize their factories better. This is what AMD neeeds since they usually are around the break even point.

Re:What about AMDs 45nm???

[xman6]

If what Intel has been talking about getting rid of NetBurst is true (since in all honsesty that's the problem right?) NetBurst has created more problems for Intel than solved them. So IF Intel gets its act together and works out a viable solution for a sans-NetBurst, 45nm multi-core CPU then they may either meet or beat AMD CPUs - that is a viable possibility. The problem is that Intel has been backward thinking these last few years with NetBurst and may not want such a "drastic jump" in performance gains - but then again maybe the recent drop in the Pentium name is part of an "apology" for lower than expected CPU performance gains? Since really the Pentium 4 has had many many problems with NetBurst and yes with HyperThreading has had its issues. However, I should also note that Intel CPUs do outperform AMD CPUs in "Real World" environments such as heavily multitasking operations like when I have two compiled environments running in real time on my workstation - I do notice the flexibility rather than an AMD. So enough AMD fanboyness - we all know that AMD smokes Intel at games and with most benchmarks and Intel CPUs consume way too much Power and generate way too much heat - but do AMD CPUs still melt if the fan dies or slips off the case?

Re:What about AMDs 45nm???

[AcidPenguin9873]

The dual-core Yonah consumes less power at 100% usage than the Athlon64 3800+ X2 does when idle while competing with it performance-wise. A low-power mobile Intel chip competing with a high-end desktop chip from AMD. Good design, indeed.

The dual-core Intel chip realizes gains in performance and reductions in power consumption due to transistor sizing alone. Its 65nm process just plain beats a 90nm one. I wouldn't make any comparisons or conjectures about the quality of the architecture and design for each chip because of this difference in process.

Re:What about AMDs 45nm???

[kensai]

Those Ethernet ports ARE NOT gloryholes.

Re:What about AMDs 45nm???

[RightSaidFred99]

Incorrect. They have moved to an architecture with shorter pipelines, meaning lower frequencies are required for the same perormance. They have also added advanced features to turn off unused sections of the chip. Please try to keep up.

Re:What about AMDs 45nm???

[AcidPenguin9873]

The blanket "Incorrect" in your response is itself incorrect unless Yonah is, in fact, NOT a 65nm chip and does not see ANY performance gains/power reduction (vs. a 90nm chip) due solely to transistor sizing. Furthermore, Athlon64's pipeline is similar in length and is clocked at similar frequencies, so neither chip has an advantage/disadvantage there. Your "shorter" is almost certainly comparing Yonah to Prescott's 31 stage pipeline; that comparison is irrelevant in this discussion.

Yonah being able to turn off unused sections of the chip to save power is certainly a significant feature.

Re:What about AMDs 45nm???

[fbg111]

It doesn't really sound like Intel is playing catch-up here.

Except in the one area that really matters - actual chip performance.

Re:What about AMDs 45nm???

[masdog]

The dual-core Yonah consumes less power at 100% usage than the Athlon64 3800+ X2 does when idle while competing with it performance-wise. A low-power mobile Intel chip competing with a high-end desktop chip from AMD.

That is hardly a fair comparison. So what if the Yonah dual core uses less power than the Athlon64 3800+X2. They're in two different categories. When there is power consumption comparisons between two processors in the same category (ie desktop or laptop), then we can compare.

I think the big question is...

[Corbu Mulak]

how much more will this cost?

Re:I think the big question is...

[thaerin]

You know the old saying, "If you have to ask ..."

Re:I think the big question is...

[Anonymous Coward]

No, the real question is; How much more than AMD's superior competing product will this cost? :P

Re:I think the big question is...

[Azreal]

Techinically, given the same wafer size, a die shrink means a higher yield per wafer. What this is supposed to result in is lower costs, but this is Intel we're talking about.

Eh.

[anderm7]

I'll believe it when they start yielding these things at greater numbers than one, on chips with a high SRAM and logic density.

Aren't we getting close to the Theoretical Limit?

[Anna Merikin]

Unless I misplaced a decimal point or misunderstand physics, isn't 45 nm only a very few generations from needing connections only one molecule thick?

+- 0 Joke

[IAAP]

You and me both - brother!

See further down on this thread.

This new format is making the news,( I guess this is the thought,) flow so fast, that we really can't think about what the fuck we're going to say. I'm finding that I'm staying away more often because I can't contribute anything worthwhile. . Yeah, yeah, I'm here now. The wife is out and I'm sipping vodka and reading /. and my grammar is going down the tubes with my spelling.

To the Mod's - I've set myself up to be "Flamed" - please don't mod people who respond to me. - THANK YOU!!!!

45nm wang?

[camzmac]

Intel's marketing campaign: Smaller than AMD!

Wait...

Re:45nm wang?

[Anonymous Coward]

It's not the size that matters, it's the length of the pipeline...

Re:45nm wang?

[Anonymous Coward]

Slashdotters will finally have larger penises than processors.

CNET News article has important additional details

[MojoStan]

Here's a link to CNET News's article on this same news:

Intel shows test chips made on future processes

One interesting (to me) bit of info from the CNET News article:

The 45-nanometer process could become particularly interesting because many chip designers believe it will be one of the more difficult transitions in years. The power consumption and performance requirements of these chips will be extremely high and chipmakers are being forced to add exotic materials and new structures to their transistors to ensure the chips function properly...

"It does get a little more challenging every time, but we come up with new technology and tricks to keep things going," said Bohr.

If a company botches the process, it could lead to product delays or recalls. Some chipmakers faced these problems during the transition to 130-nanometer chips when they swapped aluminum for copper for making interconnects--the tiny wires inside chips.

Although Intel might have a "considerable lead over our competitors in the 45-nanometer generation," it doesn't appear that this transition is expected to go as smoothly as their transtion to 65nm (which seems very smooth). Remember Intel's and IBM's difficult transitions to 90nm.

Re:Aren't we getting close to the Theoretical Limi

[Anonymous Coward]

Actually the instulators are atoms thick....

45 nanometers?

[TimTheFoolMan]

The best distance achieved in the "Frustrated iPod Owners Tossing Contest"?

Tim

Re:45 nanometers?

[holywarrior21c]

what? . . . . ipods are crap,, i would throw 45 yards...

Yes, Ladies and Gnetlemen,

[IAAP]

there is "TROLL" moderators who do not get a FUCKING joke!

EVEN though IT IS on TOPIC!!!

Don't like my puncuation? Well, talk to the /. owners.

Re:Aren't we getting close to the Theoretical Limi

[uujjj]

molecule? This is a crystal we are talking about, so the entire wafer is a "molecule". An atom of Si is about .3nm across.

Can't say I'm surprised...

[themysteryman73]

I can't say this is surprising news, there's been talk of 45nm chips for a while now, so I suppose the time has finally come when someone makes one. At the same time, it's still early tech so what are we supposed to do with this news?
"Hey, Intel's making 45nm chips!"
"Yum, what flavour?"
"Er... Internets?"

Seriously though, I know this is a step forward, but someone tell me when either vendor starts actual production on these chips

Re:Aren't we getting close to the Theoretical Limi

[Anonymous Coward]

When nanotube-based transistors become a reality, "one molecule thick" won't be a problem. The electrons just flow through the molecules instead of on top of them.

Yeah but...

[Subrafta]

Can they use it to make a dual-core that doesn't depend on the front side bus?

Re:Aren't we getting close to the Theoretical Limi

[WallaceAndGromit]

Most molecules are a few to a few dozen angstroms thick (from here), and 45 nm is 450 angstrom. So there is about another factor of 10 till we get down to the size of complex molecules. However, I do believe that most of the "stuff" used in the manufacture of chips are either pure elements or simple molecules, which are much smaller (varying from 1 to 5 angstrom)..

Doing the hard work

[ranton]

It seams to make sense that because Intel has the most money, that they can spend money on developing better manufacturing and engineering techniques than their competition. But with all of this extra money, and seamingly having better technological capabilities, AMD is still beating out Intel as far as performance.

Looks like Intel basically does all of the hard work figuring out how to do things for the first time, and AMD just has to wait until Intel is finished and then just learn from them. I of course know nothing about how to make processors, but it seams that this is the most plausible reason why Intel has trouble making chips that are as good as AMD.

This news about the 45nm manufacturing looks very bad for AMD, but I doubt it will matter very much. If Intel is doing it by the end of 2007, AMD will probably be doing it by first or second quarter 2008. And if history is any indicator, they will probably be doing it better. But I guess time will tell, maybe this 45nm technique really is too hard for a company without endless money to figure out.
--

Re:Doing the hard work

[DragonTHC]

As far as AMD outperforming an Intel chip.

they are no longer rated the same.

plus, you need a set of pratt & whitney jet engines to supply sufficient power to AMD chips.

Then, there's all that heat.

Re:Doing the hard work

[Deliveranc3]

You're close but the biggest element is that AMD liscences a lot of their tech while Intel develops it.

AMD is part of a consortium of chip manufacturers (with SUN and IBM) who cross liscence to each other, everything from instruction sets to hypertransport, to NRAM, to SOI.

Intel probably has about the same number of people developing tech but they are trying to do their development in a very corperate way - This is what we need let's do it.

As opposed to AMD who can be a lot closer to pure science because they just liscence any tech that seems cool or is proven.

When we see crazy stuff on slashdot like the four gigabit optical memory or the 2 Gigahz CPU AMD is probably looking into that stuff while Intel research is most likely pretending it doesn't exist.

Re:Doing the hard work

[Overly Critical Guy]

But with all of this extra money, and seamingly having better technological capabilities, AMD is still beating out Intel as far as performance.

That's no longer true. The Core Duo, a low-power laptop chip, keeps up with the Athlon64 3800+ X2. Damn impressive. Merom is expected to shift ahead even more dramatically.

I of course know nothing about how to make processors, but it seams that this is the most plausible reason why Intel has trouble making chips that are as good as AMD.

I've never understood the fanboyism for AMD on Slashdot. Intel has been making chips that compete with AMDs since last year. The Pentium 4 days are over. Would you rather have the Core Duo or the Turion in your laptop?

Re:Doing the hard work

[Lord Crc]

you need a set of pratt & whitney jet engines to supply sufficient power to AMD chips. Then, there's all that heat.

I run an AMD X2 3800+, an Nvidia GT6800 PCIe and three HD's (including two WD Raptors) off a 350W PSU, and I have exactly zero stability issues. Using a typical Zalman cooler, the CPU runs at about 45C under full load (two Folding@home instances). The GT6800 is miles ahead in the heating department compared to the X2.

Re:Doing the hard work

[Anonymous Coward]

I've never understood the fanboyism for AMD on Slashdot. Intel has been making chips that compete with AMDs since last year. The Pentium 4 days are over. Would you rather have the Core Duo or the Turion in your laptop?

I'm not really sure. However, if I wanted one in a desktop or a server, and had to pay for it, I'd much rather the AMD chip.

I think that's where the interest comes from in AMD (well, aside from the obvious Slashdot "underdogs rule" meme:)).

Re:Doing the hard work

[the eric conspiracy]

The Core Duo is not significantly lower power than the Athlon 3800+ despite being a smaller feature size and the fact that it is a mobile applications chip while the Athlon is a desktop CPU. It is also a 32 bit onlh implementation. Give that Vista and Linux both support 64 bit operations you can bet that I'd want a Turion instead. AMD will be introducing Dual Core Turions shortly - and that is what I will be buying.

Intel Fanboys have their heads in the sand - look at the directions in market share gains/losses and stock prices between the two companies and it will be immediately obvious that there is a revolution happening.

Re:Doing the hard work

[Overly Critical Guy]

The Core Duo is not significantly lower power than the Athlon 3800+ despite being a smaller feature size and the fact that it is a mobile applications chip while the Athlon is a desktop CPU.

Yes, it is significantly lower power. At 100%, it still consumes less power than the Athlon at idle while matching its performance.

It is also a 32 bit onlh implementation. Give that Vista and Linux both support 64 bit operations you can bet that I'd want a Turion instead. AMD will be introducing Dual Core Turions shortly - and that is what I will be buying.

Turions consume more power and aren't as fast as a Core Solo. The Merom due this July will be 64-bit as well.

Intel Fanboys have their heads in the sand - look at the directions in market share gains/losses and stock prices between the two companies and it will be immediately obvious that there is a revolution happening.

AMD's revolution is a year behind, and their roadmap is bettered by Intel's, who's kicking butt with the Core Duos. Call 'em "Intel fanboys" if you want, but this lame underdog mentality that causes Slashdotters to obsess over AMD doesn't change the fact AMD is a year behind in the 65nm and 45nm switch, and Intel owns the mobile market. With Merom and Conroe coming out this year expected to halve power usage while doubling performance, it's obvious who will dominate this year. AMD has yet to go 65nm and is having to rely on the unreliable IBM for chip development (good luck waiting on those guys). AMD had a good run, but like ATI, they've fallen behind and can't catch up.

Re:Doing the hard work

[Mutiny32]

Your comment is pure troll bait. You know that, right? Becasue that's why you made it.

Intel is still behind AMD in the 64-bit x86 arena. By the likes of nearly TWO YEARS.

Re:Doing the hard work

[Ruie]

Let's not forget that what is really biting Intel in the ass is their decision to ramp up the clock and have a huge pipeline which owed a whole lot more to marketing than sound design.

When Pentium IVs came out they were slower in actual performance than Pentium IIIs and this is a strange thing indeed from new and supposedly better design.

Now that consumers have stopped buying based on clock speed alone (as neither AMD or Intel talk loudly about it anymore) Intel's marketing value decreased to a point where they had to ditch Pentium brand.

Re:Doing the hard work

[Luctius]

Can't..
resist..

>> ... it still consumes less power than the Athlon at IDLE while MATCHING ITS PERFORMANCE.

Is it just me or ... ;)

Re:Doing the hard work

[Anonymous Coward]

I don't really understand all this 64 bit nonsense. Unless you have >4GB physical memory, or applications that need
more than 4GB address space, 64 bit doesn't buy you anything except doubling the ram requirements for pointers.
(And yes, my PC has an Athlon64 - but I bought that for the integrated memory controller, not the 64 bit features).

Re:Doing the hard work

[ErikZ]

Impressive?

After spending billions on a new fab and R&D, they produce a chip that "Keeps up" with AMD's slowest dual-core chip? The 3800+ is made with older tech and came out last year.

"You keep using that word. I do not think it means what you think it means."

Yes, but no...

[electrosoccertux]

From one molecule thick? We're far from that.

But we ARE only a few more generations from hitting a rather thick wall: at the 5nm, electrons begin jumping _through_ the insulators to a nearby circuit. So while we're far away from the molecular level, we're still getting closer and closer every day to a very real limit. We should be able to push it down to 4nm with a little extra engineering....but as far a I know, thats going to be it. Anyone else want to comment?

Right on schedule

[Anonymous Coward]

Intel's logic development is striving for a two-year cycle for each new process technology. This announcement of functional first-silicon (who knows how long they've actually had it) is part of that natural progression. Here's a table showing this announcement along with previous SRAM test chip announcements:

Process
Litho
Size
Date

P860
130 nm
18 Mbit
Mar 2000

P862
90 nm
50 Mbit
Feb 2002

P1264
65 nm
70 Mbit
Apr 2004

P1266
45 nm
153 Mbit
Jan 2006

Okay ... it's not a table...

Just cause I've never done it before...

[painQuin]

I for one welcome our new insanely small overlords.

Speaking of Theoretical Limits...

[Anonymous Coward]

Mr. Scott. To put it another way -- how big would one of your gates have to be, with a 300 mm wafer, to resolve properly using your current method of lithography?

Intel Exec. That's easy. Six molecules. We have stuff that big in stock.

Mr. Scott. Well, suppose I could show you a way to build a gate that could do the same job -- but be only one molecule thick. Would that be worth somethin' to ye?

Intel Exec. You must be joking.

Dr. McCoy. Perhaps the professor could use your computer...

[Later]

Dr. McCoy. [Whispering] You realize that by giving him the formula we're altering the future.

Mr. Scott. How do we know he didn't invent the thing?

Dr. McCoy. [Smiling] Yeah.

Re:Speaking of Theoretical Limits...

[evil-osm]

Kirk. Ok McCoy, you and Mr. Scott, go out and find a 45nm gate.

Dr. McCoy. Oh...joy

Re:Speaking of Theoretical Limits...

[scapermoya]

Dr. McCoy. Damnit, Jim. I'm a doctor, not a microscopic silicon expert

Less than stellar design

[Anonymous Coward]

It's good that Intel is ahead in shrinking and speeding up their CPUs, since their CPUs are not as well-designed as AMD's. Why do you think that an AMD CPU beats an Intel CPU at the same GHz?

Re:Less than stellar design

[snuf23]

I'd rather have a Core Duo laptop right not than a single core Turion. Of course we'll have to see how it breaks down later this year when we see a dual core mobile processor from AMD.
Of course the Turion is 64 bit, but if you are a Windows desktop user that doesn't count for much at this point in time.

Re:Less than stellar design

[Overly Critical Guy]

Why do you think that an AMD CPU beats an Intel CPU at the same GHz?

You have nothing to back up your "not as well-designed as AMDs" claim with regard to Yonah/Merom/Conroe. AMD is going 65nm way at the end of 2006, a year behind Intel, who will be going into 2007 with 45nm plans (a year before AMD).

Re:Less than stellar design

[Anonymous Coward]

I like my Intel laptop too - its a Pentium III 700 MHz. I don't see how the Intel core duet or AMD tourist CPUs would let me surf the web or do email faster than on my current laptop. Now, I do see how the Opteron kicks the crap out of the XEON. I had to laugh when Dell proposed that I buy 50% more of there servers so that they could come close to the performance of the servers from vendors using Opterons. Yeah, commit 50% more rack space and 100% more power and cooling for Intel vs. AMD. Needless to say, I bought from the AMD vendor.

Re:Less than stellar design

[hattig]

AMD will be doing 65nm during Q3, not at the end of the year.

So Intel have a 9 month lead on process shrinks. They had 300mm wafers years before AMD too - then again they needed them to meet demand. OTOH AMD have third generation SOI in their processors, and their 90nm processors don't have the same issues that Intel's had, AMD dropped power consumption dramatically with 90nm, although top clock speeds haven't been amazingly better.

AMD are a smaller company, and they're growing at a steady, manageable rate.

We should all be thankful that Intel has nearly got over the Netburst disaster on 90nm. Now both companies will be more competitive. AMD will probably keep growing in the server area this year, consolidate its desktop gains, and win more lower-end laptop marketshare. 2007 however will be when Intel are shipping in massive quantities new products that cover the entire market, and right now they look very competitive. What will AMD have? 65nm SOI K8L ... not much is known about this though.

Re:CNET News article has important additional deta

[Anonymous Coward]

CNet botched that reasoning. Given the same frequency and number of transistors in a chip, the new process would yield chips that run at lower power consumption, not higher.

The reason power consumption [often] increases for chips made with the smaller processes is because for a given amount of silicon real estate, such chips will have a higher transistor count and can usually be pushed run at a higher frequencies, both of which increase power consumption.

Moore's Law

[jd]

You see, Intel has a cunning plan to pepetuate Moore's Law. By making chips 45 nautical miles across, they can keep doubling the number of transistors for a very long time.

Troll?

[Anonymous Coward]

I was thinking his joke was that an ipod "nano" got tossed 45 meters. Which while not terribly funny, is somewhat clever; at the very least not a troll. I would have read it as trolling if he were trying to say that ipod owners could only throw something 45 nanometers... which is just too lame a joke to possibly be intentional.

Not necessarily funny... But Troll?

Re:Troll?

[TimTheFoolMan]

You got it. Sadly, the moderator didn't.

Maybe his went 45 minimeters.

Tim

P.S. I'm even less likely to be trollling when there are two iPod Mini's, a Nano, and a 4G model under my roof. If I were moaning about battery life or somesuch, I would be speaking from knowledge and not trolling.

Re:Troll?

[TimTheFoolMan]

It would appear that there are at least 3 moderators who have no idea what a Troll is. Off Topic? Yes. Funny? Debatable.

Troll? Hardly.

I suppose I should have engaged the debate about whether or not AMD has a 65nm process, and what the percentage yield would be should they be successful at going further than this. That way, it wouldn't matter if I had a clue or not, somebody would probably mod it up.

Another reason that Slashdot seems to get less relevant every day.

Tim

An inefficient Intel is still faster...

[Anonymous Coward]

...than an effficient IBM/PowerPC offering.

It is a Chip not a CPU

[karvind]

Just to make sure, this is not a CPU chip using 45nm technology. This is a test vehicle which contained SRAM (static RAM) and some control logic. SRAM arrays are regular and don't have the same complexity as ALU (arithmetic logic unit) and other control circuits found in CPU. So yes this is a big step because it is gives some indication about how complicated will it be to get a good yield in this process. Also note that SRAM arrays can be easily made defect tolerant by using spare rows/columns. Same is not true for CPU cores. So there is still al long way to go before an efficient working CPU with production acceptable yield is available.

Re:It is a Chip not a CPU

[Anonymous Coward]

Wait, does this mean Intel are going to make DDR2-SRAM chips?

Ok, can we just put more empty space in now?

[fwitness]

Seriously? Can't I have a chip that runs relatively fast, does everything a modern computer is used for, sans games, and I *don't* have to water-cool? Something like what the VIA Epia series does, but with Intel's backing?

Is it just me, or is web-browsing and document writing fast enough? It seems like 99% of the time these days I just want something smaller and quieter. If I want pretty shiny games, I'll play them on my xbox390 or sumsuch. Sure you can make bunches of chips for gamers, but give me a slimline chip and I, like many others would flock to it.

I'm writing this on my 733Mhz laptop, bought for college way back when, and my typing fingers really don't recognize the lack of dual cores.

Re:Ok, can we just put more empty space in now?

[hxnwix]

Do your typing fingers notice the difference when they have nothing to do because you are waiting for a package to emerge?

Oh, you don't use gentoo, at least on your laptop.

Do your typing fingers notice the difference when you are waiting for matlab/octave to ploddingly chug through huge amounts of data?

Oh, you don't do that kind of thing, at least on your laptop.

Do your typing fingers notice the difference when you are waiting for ethereal to parse a few gigs you just pulled off the wire?

Oh, you don't do network related debugging involving large captures, at least on your laptop.

Think though. What if you did these things on your desktop, but longed for a fast enough portable to more conveniently accomplish the same tasks?

Well, then you'd be more like the average slashdot reader and less like my grandmother. See what I'm saying?

Re:Ok, can we just put more empty space in now?

[solios]

My dual 2ghz G5 with 2g of ram can NOT keep up with me in Photoshop. It just can't.

The 2.4ghz Athlon on my desk with a gig of ram is chunking out my 3d renders at about 30 hours per frame (print rez high quality etc). Not nearly fast enough - I spend a week waiting for four or five shots.

My beige g3/366 brings the internets at warp nine with no complaints, and my G3/900 iBook runs the Safari and the TextEdit just fine. It happens to suck total ass for Photoshop in the sense that it's not the G5.

People use the computers for more things than internet and {vai|emacs|ed|pico|whatever}, kthks. And for everything but internet and {vai|emacs|ed|pico|whatever}, top of the line shit is still not fast enough .

Re:Ok, can we just put more empty space in now?

[Anonymous Coward]

So what? Your post is completely irrelevant. The previous poster was talking about chips for those who don't want/need/care about demanding tasks like 3d rendering. I'd love a passive cooled computer (if possible even the PSU being passive cooled). I currently have the case open because otherwise the core temperature will easily go above 75C, needless to say, the noise is pretty annoying (not to mention that I have above average hearing... for now).
  I'm seriously thinking of finding more RAM for my 486 (it doesn't even have a heatsink) and use that as my main desktop, and keep this computer mostly off for the summer (I'm in the southern hemisphere, obviously).

Re:Ok, can we just put more empty space in now?

[m50d]

You've answered yourself - VIA is making those kind of things. They're making them very well, and there's no real point Intel trying to compete with them in that space - Intel can't afford to compete directly with them on price, and has no research lead because everyone knows how to make processors at that kind of speed.

Re:Ok, can we just put more empty space in now?

[prisoner-of-enigma]

Something like what the VIA Epia series does, but with Intel's backing?

Why is Intel's backing even remotely important to this equation? VIA's chips are fully x86-compatible. Anything you can do with an Intel or AMD chip -- sans 64-bit stuff -- is doable on an Epia, just a bit slower but a lot cooler and quieter.

For that matter, if you're really anxious for something cool and quiet, you can get Pentium-M chips that'll go into micro-ATX or even full-size ATX motherboards. The chips sip power and put out barely enough heat to get warm, and their performance is orders of magnitude better than the Epia's. There are even completely fanless cases available that are aimed at the home theater/media PC market.

Re:Ok, can we just put more empty space in now?

[djrogers]

Intel's newest mobile chips are xactly that - I'm typing this from my 2Ghz Pentium M, which requires less power and cooling than your 733...

Re:Ok, can we just put more empty space in now?

[darkmeridian]

Build a Pentium M desktop. Then stick a huge 120 mm heatsink/fan combination on it at a low-rpm. A few large 120 mm and/or 90 mm case fans at low rpm and you have no problems.

Re:Ok, can we just put more empty space in now?

[duggy_92127]

Is it just me, or is web-browsing and document writing fast enough?

For a few years now, I've been telling family members and friends who ask me advice about buying a new computer: Don't ever worry about the CPU. In fact, get the slowest one you can find. It'll be cooler and the battery will last longer, most likely. (Of course, I tell them to put as much RAM in the thing as they want to spend, as well.)

But I'll take this one further. I recently got a new laptop at work; it's an IBM ThinkPad T43. Not terribly new tech, and it's very much a business machine. 1GB ram, some Centrino CPU, some ATI mobile graphics chip...

The thing plays WoW at 25fps. Now, this is at 1024x768, the native LCD resolution, and 25fps isn't stellar... but this is a work laptop. Playing a relatively new game with default graphics settings at a very-playable 25fps? Insane.

I am totally not complaining, of course. I love the fact that current tech is almost overkill for anything we can throw at it; it just means we'll find more interesting things to throw at it!

Doug

Re:Ok, can we just put more empty space in now?

[Frenchy_2001]

Can't I have a chip that runs relatively fast, does everything a modern computer is used for, sans games, and I *don't* have to water-cool? Something like what the VIA Epia series does, but with Intel's backing?

Sure, it's called the Pentium M low voltage. Does not run as fast as it's higher voltage sibling, but is quite powerful for a minimum power draw. Conversionnally, any notebook chip today would be MUCH faster than yours and use less power, what with their dynamic clock speed and turning off unused parts. Just grab a Centrino notebook or build a Pentium M computer and surf away!

Re:Ok, can we just put more empty space in now?

[fwitness]

Wow, I must have sounded like I was ranting. You're the only one that got my point. Are there times when I want the fastest greatest CPU? Sure, there's lots. I do quite a bit of image editing for photo's a web design, and I know that yes, they can never be fast enough. There will always be a market for smaller, faster, more gigahertz etc.

The thing the sibling posters don't get is that *there already is a market for you*. Every major manufacturer out there is dying to eek out an extra few K per second of bandwidth/processing power etc.

I'm saying that VIA is doing what I, and many others, would like. I was simply wondering if we had a market for us? There's whole magazines and websites devoted to smaller, quiter, small-purpose computing. Can't we have a market too?

Re:Ok, can we just put more empty space in now?

[m50d]

You are a market, just one that VIA owns. I suspect no-one will try and compete with them for the same reason no-one really competes with VIA on low-end motherboard chipsets - once technology is out of the equation VIA can make them cheaper than anyone else.

NM?

[ROFLMAObot]

Newton Meters? I guess it makes sense. Intel always has been aiming for "raw power" now we see it in the form of torque.

What Am I missing

[caller9]

http://www23.tomshardware.com/cpu.html

What of this comparison indicates a clear Intel advantage in processing power?

I don't get it. I see all these posts about Intel roxorz with a vengance. AMD is slooow and consumes more power and produces more heat...etc. While all of those claims seem counter to the truth. AMD is waay behind they're using engraved cobblestone for processors for no apparent reason then they use 332 tera-amps and require about 2 nuclear fission reactors while producing more heat than the sun...which is ludacris, not the rapper, the state of plausability. All this and that processor can't even recalcualte a 4 cell excel sheet in 3.5 years.

I call shenanigans, get your brooms.

WTF? are there intel shills in the slashdot crowd or am I not looking at the facts correctly.

From the stuff I've read that wasn't published by Intel, AMD wins in Wattage per processing score and in the thermodynamics category as well. From some of the crap I see here, AMD is akin to using a 10 megawatt light that puts out about 3 candle power but dips into the infra red range then straight to heat energy at an astonishing rate. I just don't see that being true with independant studies.

Re:What Am I missing

[caller9]

Oh yeah, I forgot 1 point. The decision by major manufacturers to back AMD was definately related to performance + support costs/$. Even in the blindest rage of Intel zealotry you have to realize that AMD overclockers choose it because a chip that is rated at X GHz by AMD is far more likely to work at X+.1GHz than an Intel chip without requiring a trip to the local dry ice supplier. I can't make an informed decision on whether that is due to different quality standards i.e. it runs with .05% error rate at the marketed speed or it runs at .0001% at the marketed speed, but I would like to think that is a factor given the evidence I've seen. Take that with a grain of salt.

The biggest problem I've had with an AMD chip is the crappy fan on the northbridge of a particular manufacturer. They chose to forego a speed sensor on that fan and, of course, it stopped spinning entirely. What should've been an APM shutdown with a POST error on reboot was a silent killer until the case was opened. BTW: FUCK YOU Abit.

Re:What Am I missing

[Anonymous Coward]

What are you on?

Re:What Am I missing

[RightSaidFred99]

Damn, you _hammered_ that strawman! You lined it up, and you punched it right in the face!

You said "Look, there's a strawman, he says all these people on SlashDot are Intel fanboys saying AMD sucks!". Then, you punched the strawman in the nose.

Get a fucking grip, retard. Everybody knows AMD has better performance and thermal characteristics compared to the P4. So your claim that everybody is saying otherwise is laughable.

Now, you may be confusing comments about the new Core Duo chips, which do perform as well or better and run cooler, with the old P4 chips. Anyone claiming Intel chips are thermally advantaged is talking about these, and not the P4 - or they're retarded.

Re:What Am I missing

[99BottlesOfBeerInMyF]

What of this comparison indicates a clear Intel advantage in processing power?

Nothing, but that comparison also does not include the latest generation of Intel chips, including the 65nm Intel Core Duo. Intel has leapfrogged AMD for portables as far as performance and power consumption right now. AMD still seems the winner for price/performance for desktops and servers. AMD may well catch up when they release their own 65nm chip which was last scheduled for Q4 of this year. In the mean time, Intel is definitely winning the performance/power consumption battle. Also, this announcement of the first 45nm test fab provides some indication that they may leapfrog AMD again sometime next year when they beat AMD to market with the next smallest die by another year (although a lot can happen in that time).

Please note, I'm not in any way partisan here. I haven't bought an non-PPC chip in years.

Re:What Am I missing

[Anonymous Coward]

From the stuff I've read that wasn't published by Intel, AMD wins in Wattage per processing score and in the thermodynamics category as well.

Please someone tell me. What is this obsession with heat? Why the hell is there a "thermodynamics category"? If the chip runs fast and is reliable, who cares? Is there a "gate leakage current vs. processing" metric too? No, and why should there be?

Are we all so miserly about our electricity bills? I know I am not. Just turn off a light bulb when you are using your computer. It will save a lot more power.

Yes we know that the energy density in a Silicon chip is getting pretty close to the inside of a nuclear reactor. But of course that is only because of the transistor footprint. As long as heat extraction is sufficient I cannot understand why anyone would care.

Are Intel chips less reliable? If they are operating at higher temperatures then this would be the only reasonable negative impact. But reliablity is dictated by so many things, and it is typically not the hotest part that fails first.

Please someone explain this to me. I just don't believe this hype.

Re:What Am I missing

[caller9]

Heh, your name is RightSaidFred. It's like arguing on teh interwebs...

Re:What Am I missing

[caller9]

I don't know why I even post in processor argument pages. Ooooh OOoh! it runs slightly faster and they totally said the most recent buzzword. I'll gain .2 seconds per MPEG encoding and over 10ms on application launch time!! versus this old POS chip I bought last month.

Not that you said those things, just that I've come to the realization that the MFLOP wars are expensive and retarded on the battlefront. Give me a stable 1.7GHz celeron with plenty of RAM and a big load of storage running 3 server VMs and I'll show you efficiency. Assuming you don't want to run a terminal server to replace 50 PCs running photoshop, what are these baddest ass chips worth in the short term. Not saying you should invest heavily in old tech, but if you're buying something new, why not buy something 3 months old. It's 98% as fast as the stuff that costs 2x as much due to the new tech bell curve.

Yours Truly,
Tommorrows Luddite.

Big Stinkin' Deal!

[joetheappleguy]

I made a 45nm chip meself, but I sneezed and I haven't been able to find it since.

Re:CNET News article has important additional deta

[Anonymous Coward]

you forgot about leakage, which definitely increases as the chip shrinks.

Re:Aren't we getting close to the Theoretical Limi

[necro81]

I once had a conversation with someone who was doing developmental research for an even smaller process for some very large semiconductor manufacturer. According to him, they were one day running some measurements on the first prototype wafers. From experience with every previous process (65, 90, 130, etc.), they were expecting this particular measurement to yield a nice bell curve. Instead, they got a strongly quantized bell curve: it looked more like a histogram. The reason, they realized, was because the gate oxide was becoming only a handful of atoms thick. The quantization between two steps in that bell curve was the difference between, say, a gate oxide 5 atoms thick and 6 atoms thick.

So, yes, they are indeed coming up against some real physical limitations for CMOS technology. Of course, people have been saying that for years.

Moore's Flaw

[woolio]

Why does the media insist on using the term "Moore's Law"???

Since when do self-fufilling prophesies become law?

Self-fufilling prophesies tend to restict one's actions rather than sustain them.... Which is why superstition is harmful....

If every PHB believes in Moore's quip, then do people get fired for not doubling # transistors every 18 months? Do they get a bonus for doubling the # of transistors in 17 months?

Perhaps if they weren't so darn busy cramming more transistors on the chip, they could better improve their compiler or come up with truly innovative architectural techiques that *work*....

Re:Moore's Flaw

[Anonymous Coward]

Or producers could cease innovation, and people would have to actually learn how to program again instead of using the massively bloated methods of today.

I mean really. Do I need a 20.5 MB upgrade for adobe reader to look at PDF's?
or Nero, it's now around 100MB for initial download. I bet a self respecting individual could pop the same functionality out in about 10MB, or less.

We have plenty big enough pipes on the hardware end... I propose that the software end is what needs improvement.

Intel discovers Moore's law!

[im_thatoneguy]

"It will pack about two times as many transistors per unit area and use less power." Bohr told Reuters in an interview. All in only 18 months!

Dip?

[ruiner13]

How big is the bowl of dip? Can't have chips without dip.

Re:...

[Anonymous Coward]

Exactly, they are going a step backward by going back to the Pentium-M with some modifications.

You're completely ignoring Merom/Conroe/Woodcrest, which is a definite step forward.

Intel just canceled their Whitefield processor, the only one that ever was on their roadmap to sport an integrated memory controller.

Core Duo/Yonah has already shown itself to be neck-and-neck to AMD's chips even without an integrated memory controller. I think what this says is that Intel's strategy is different by choice, and that it really isn't a bad one. AMD's approach has definitely paid off (for them), but Intel clearly thinks that (for them) it isn't worth the trouble.

I don't have to wait and see, it's too clear where this is going.

Yeah: into fanboy hell.

not quite peta- but getting close

[Anonymous Coward]

it could be interesting to pair this with
another recent news from Intel, announcing 450mm wafers:

http://www.xbitlabs.com/news/other/display/2005120 1210944.html

45nm transistors on 450mm wafers!

this (roughly) means ( 0.45 m / 0.45 *10^-7 m)^2=10^14 devices on a single wafer!

10^14 is a hundred trillion (not quite the desired 10^15, but getting close)

Can we get a layman's version of something?

[Rooked_One]

Why is it that intel can't get their thing together with clock speeds. AMD is crushing them in quite a few areas, so with this transistor shrink, will they perhaps change the very application happy architechture??

No, BGA.

[Anonymous Coward]

These days CPUs have too many pins to fit in a DIP socket. Now they use Ball Grid Array.

All the more reason to work on it soon

[Sycraft-fu]

The last thing you want is to be bringing the 45nm fab online as 65nm is reaching it's limits, only to find you need another 9 months of working the bugs out before you get useful yeilds.

Law = Axiom

[DoubleReed]

I believe the term "law" is in the sense of Newton's laws, Moore's Law is not really something you can test and observe as much as it is a paradigm.

A couple of hours ago a professor put up a slide of the graph of the industry trade group that gets together and makes predictions for feature size, clock speed, and transistor density for 10 years out.

The reason they do this is for planning. I'm not sure why anyone wouldn't believe in Moore's Law at this point, PHBs included. It has been an observed trend for 30+ years.

As for being too busy working on transistor count to do other things, transistor count is really independent of architecture, and has nothing to do with software. So, they are doing both. In terms of architecture: MMX, two layer caches, hyper threading, and now dual cores just off the top of my head. And as for compilers, the question these days is *which* full featured object-oriented library with support for GUI, networking, and multi-threading with all the utility classes and syntax sugar you can think of thrown in.

Re:Law = Axiom

[Anonymous Coward]

In addition to this, it is also a mentality.
Do you think their R&D department would be working their butts off shrinking down the process if it was "Moore's Speculation"?
"Yeah well he only reckons density will double every 1.5-2 years. Our current research clearly shows this isn't possible."
These days they're pulling out all the stops just to keep this whole law thing going.
For the past 5 years I've seen articles always saying Moore's Law will hit the brick wall that year.
And guess what? It didn't. They're still going.

I remember when Pentium Pro came out they speculated in 10 years time chips would need the water flow of Niagra falls to cool it.
Does anyone have to pump 1000's of gallons of water to cool their desktop chips?
Highly doubt it unless you're trying to achieve some godly overclock (at which point most people are using liquid nitro).
These baby steps in processor development aren't _just_ for marketing.

Eventually though the law will break.
But would you rather see the final breaking of the law sooner or later?

Re:Law = Axiom

[woolio]

my point is the trend exists because manufactures feel they must maintain it... They do not aim higher, they do not aim lower, they do not even question the need for such increases.

Re:Eh. I'm waiting for the next one...

[catmistake]

45nm? I think I'll wait to upgrade until the second generation 11nm chips have got the all their bugs worked out in the 3rd quarter of 2010. Who wants to lug around a laptop big enough to house a freakishing large chip like the 45nm, or even the slightly smaller but still too big 23nm? Not I.

Good move, Intel!

[VincenzoRomano]

Now that you went down to 45nm, you only need to design and build working processors!

Core Duo or the Turion

[Anonymous Coward]

Unless i had the constant need to run a virus/spyware scanner in the background i wouldn't buy a Core Duo.... not yet anyway. I'd wait for a dual core AND 64 bit enabled mobile CPU.... but then again, that's just me.

In other news

[Anonymous Coward]

Ok, so AMD announces that now more than one in five of all cpus are made by AMD.

Then Intel quickly grabs headlines with "Look shiny" using their 45 nm experimental research setup. Mhm, that's sounds a lot like that other big market dominator which always screams look at our next generation stuff don't look at current Apple news, we will eventually have something better for you so stick with us.

On package

[Anonymous Coward]

i'm still waiting for these to show up:
http://www.anandtech.com/showdoc.aspx?i=2511

Re. 45nm and "pixie dust"

[electronspiraltoroid]

Hi.. Its very impressive but has anyone considered the long term stability of "strained silicon" ? Methinks industry will not be happy when the chips mysteriously start failing due to a form of "silicon rickets"... -A

Re:XBox overclocking?

[strike_svl]

Wow! How did you get a 30 degrees performance increase on your Xbox?

I'm thinking...

[msauve]

you misread what I wrote...

No actually, yields do not fall exponentially. If they did, nothing made today would work at all. Yields are largely independent of feature size

I was speaking of die size, not feature size.

Turns out, I was correct - yield is an exponential function of die size.

All yield models discussed at http://www.semiconfareast.com/test-yield-models.ht m express yields as powers of e.

Then there's this: "The size of the actual silicon plays an important role in yield. A smaller design, holding all other factors equal, will increase yield for two primary reasons. First, the smaller the die the more that can fit on a wafer thus giving more yielded parts for a given number of wafers. Second, the actual per part yield is a function of area where smaller area is better since there is less chance that a random defect will land on the silicon rendering the part inoperative."- http://www.cadence.com/company/newsroom/articles/D esign_Yield_Cost_Model_eedesign_June04.pdf

...which also supports my statement.

In soviet russia

[PringlesQueso]

Our 45nm chip is evidently better than the americans'! Can't you see, it's right there!

Top of the line will never be fast enough...

[Anonymous Coward]

... for any sort of arbitratily complex computations that human brain can imagine. Did you just figured out?

Obligitary

[rufty_tufty]

"But that is the second word the Knights Who Say Ni cannot hear! Oh no I just said it, and it as well, no I said it again..."

45's.......

[lemon_dieter]

On 45 there's only room for like 2 songs, man! I think things are gonna cook when we can use a 33 1/3 process. I could please my woman to Sade for at least 30 minutes per side.

Re: Wrong. Intel has 2 cpus w.memory cont coming

[Glasswire]

... this would be Tukwila and Poulson with have CSI embedded memory controller architecture in '08 and '09.

These are Itanium family cpus but Intel has also announced that Xeon processors with CSI will be shipping in the Poulson timeframe.
If you look at Intel's roadmap, with huge speed increases in FSB, FB DIMMs and multiport (no shared bus) memory connects from sockets to chipsets you realize that embedding memory controller is only one way to get sufficient memory bandwidth.

Question

[jaygatsby27]

Is that a 45nm chip in your pocket or are you uninterested in seeing me naked?

Re:People have been saying that since 1985

[vertinox]

Is it just me, or is web-browsing and document writing fast enough?

And was 640K of ram all you ever needed?

Computers will never be fast enough never will have enough memory.

They will never stop making them go faster.

NEVER

So buckle up and get used to it.

You are not alone

[bradleyland]

A lot of my customers are screaming for the same. One of my personal predictions for this year is that we'll see a shift from bigger, faster, more to moderation, cool, and quiet. All my customers want to put their computer in a cabinet and close the door. You simply cannot do that with modern PCs. The ambient temperature inside the cabinet elevates until the PC's cooling mechanisms can't operate as intended. Lockups and crashes follow.

At one point I was building PCs based on the 45W Athlon Mobile series of CPUs because you could drop them right into a standard nForce based motherboard, but supplies are no longer readily available. What is needed is a solid desktop platform for the Pentium M and Turion, or at least a desktop counterpart to the this type of processor.

Business productivity apps are more demanding of system memory than they are CPU. A PIII 1 GHz with a gig of RAM will seem more responsive than a P4 2.8 GHz running 256 MB of RAM, yet bargain systems, like Dell's "$299 special", ship with an overblown processor and inadequate RAM (256 MB). Chill with the processor speeds and make with the efficiency.

No motherboards?

[eonlabs]

You do realize that with the 45nm chips, they could produce the same old hardware at half the size and double the speed without any change in pin layouts.

granted, I think it's a wasted opportunity, but it still would be marketable with today's current mboards.

Intel, AMD, meh

[billcopc]

The way I see it, Intel has been losing the race for years, it just took them this long to lose enough market share to actually get off their cellulite and do something about it, though it may be too little too late.

Yes, the Socket-A Athlons were cute little chips that gave impressive performance at low cost, but in exchange for extreme heat and noisy fans. They helped drive prices down; for the price of a Pentium-4 CPU, you could afford a whole AMD-based mid-tower system with a burner and all. The chips were a nightmare for techies with their extravagant heat dissipation and people had a habit of cracking the core. Nevertheless, it was cheap and reliable.

Now they have Athlon 64 and X2, and the tables have turned. Now it's AMD who's got the pricier chips, with Intel covering the lower end because they simply haven't made much progress in the last three years. AMD has had a 64 bit desktop processor on the market for over a year and a half now, for less money than the cheapest 32-bit Pentium-4 CPU. Hell, most people buying a high-end P4 today still don't have 64-bit support. Why does it matter ? We may not be using the 64-bit much at this point, but it still gives AMD valuable experience to further improve their designs, and a growing user base to help AMD control the market while Intel plays catch-up.

It really doesn't matter which industry you're in, when a competitor catches you with your pants down, you're going to lag behind until you come up with the next big thing, or else your company will grow exhausted and fade into obscurity.

I'll bite.

[Anonymous Coward]

And was 640K of ram all you ever needed?

No.

Computers will never be fast enough never will have enough memory.

Not the top of the line computers that do really heavy tasks, however, I've found that I was satisfied for more years than usual when I bought my Duron 1Ghz. I bought a new shiny AMD 64bits Sempron a couple of months back, and I guess that will keep me busy for at least the next 5 years.

They will never stop making them go faster.

Well, it depends on the market.

Personally I've found that there is two kind of machines I want:
  - Big machines with hefty cooling, that can take a LOT of disks, and enough CPU to run software raid-5.
  - Nice quiet desktops that are fast enough to play full screen xvid's and do all my other stuff on.

I'm not quite settled on the desktops yet. VIA still needs to push their speeds a little bit up. When they're fast enough to do all I want on them - then that's my machine of choice for desktops.

My disk-servers will be beasts, my desktops will not.

Fortunately!

[faragon]

Yes, may be are not news at all, at least from the engineering and scientific point of view, as means that a brave new effort will be required for push computational new frontiers. Usually, the human being need to top with hard limits to surpass them with new and fresh ideas.

Come on, let's shine all, it is time for a computer science renaissance!

Re: In Korea only old people .....

[chawly]

.....use chips of this type. They've been buying them from China for years .... I thought everybody knew already !