Slashdot Mirror
Intel Drops Tejas, Xeon To Focus On Dual-Core Chips
PunkerTFC writes "Reuters has an article about Intel dropping the fourth-generation P4 chip (codenamed "Tejas") and the Xeon server processor. Intel says they want to concentrate on their new 'dual-core' technology for desktop and notebook systems. This is essentially putting two processors on one chip, allowing for a doubling of performance with less energy use. The introduction of this technology was not expected for another year and a half. Rival chip maker AMD says they have the capability to produce dual-core chips and will introduce the technology when they "feel there is a market need.""
but might this have something to do with the recently-announced Longhorn specs?
But, does this suffer the same problems as current chips do wrt dual processors? Or quad processors?
What's the penalties of this technology? Does anyone know?
Sounds too good to be true for a dual core cpu to act as a single core proc.
This has been discussed before.
Karma: Meh (Mostly from meh.)
FP FP Written Written from from my my new new dual dual core core chip chip from from Intel Intel. Still Still some some bugs bugs to to work work out out.
This seems to be the new trend,
AMD will have dual core opterons next year:
Here's the real impact many of us will be feeling. Software vendors that license by the CPU have already in fair part indicated that they consider "dual core" chips to be two CPU's for licensing purposes.
In other words, people are going to find themselves having to pay higher licensing fees with regular desktop computers as well as servers. Small workgroup servers could be really hard hit by this from some vendors.
I wonder how this will play out with XP Home which only supports one CPU? AMD has the technology so they may well respond in kind when Intel does (dammit lead AMD, lead), which could have a fair impact in weaning the masses of XP Home. I dont think MS will let this go the route of hyperthreading with the "logical processor" support.
Intel says they want to concentrate on their new 'dual-core' technology for desktop and notebook systems. This is essentially putting two processors on one chip, allowing for a doubling of performance with less energy use.
Is this a parallel implementation then? In that case performance is only doubled for processes that can be performed in parallel.
I think this is more related to moving to the PM from the P4 architecture as the M series is more scaleable - taing P4 any further requires a lot more power and generates a lot more heat..
"Sir, maybe we should introduce our dual core chip now!"
"No... that's just what they'll be expecting us to do..."
Business isn't willing to pay for products, innovation and careers, so we get brands, mortgage commercials and layoffs.
Hyperthreading and dual core are very different things. A dual core processor is basically two processors put onto one die. There are twice the number of execution engines, just like two separate cores, but on the same chip. This means it's easier and cheaper to make and install than two separate processors, and it has approximately equal performance.
Hyperthreading takes one physical processor and makes it appear to be two logical processors. There's still only one core and one execution engine. It appears to be two processors, but a 3.2GHz Pentium with HT will have nowhere near the performance of 2 3.2GHz Pentiums without HT.
Intel, like Microsoft, Dell and Sony, is a favored company.
AMD, like Nokia, Apple and Nintendo, is not.
AMD's strategy (Opteron instead of dual-core?) will therefore be called "a significant risk given the current market reality" while Intel's strategy (dual-core instead of Itanium?) will be called "a savvy decision for the technology giant," even though the media wouldn't know an Opteron or a dual-core CPU if one jumped up on their desk and did the tap number from 42nd street.
All of the general stories will make repeated and redundant references to the effect of Intel's strategy on the "tech-heavy Nasdaq."
This is no different than the Sony vs. Nintendo console competition. The media doesn't like competition. Neither do the markets. (There is only room for three companies in any given market) It's so much easier to be a sycophant when your favored company has 80% of the market.
Business isn't willing to pay for products, innovation and careers, so we get brands, mortgage commercials and layoffs.
C) Neither. More likely that Nvidia will move to the straight CPU market and compete along side of AMD and Intel. They understand though right now the market is bad for that, and instead make great chipsets for AMD (while being the underdogs, they're also a very good ally to have if they actually do attempt to shift into desktop processors).
ATi on the other hand, while they also make chipsets for Intel and AMD, they are much more concentrated on the Video market, and they really always have been (best 2d quality, bar none since a long time ago).
Intel on the other hand, is starting to shift gears to a more mobile computing based company. They know the future of computers is in having them everywhere we go. Now that computers are finally cheap enough to be everywhere, the next step is to have them WITH us everywhere we go. Intel's been focused on Mobile computing for a long time (StrongARM processors, and the -M series of all the pentiums, including the Pentium M itself). Their switch to having Pentium M on the desktop was really a have-to case, as AMD is really starting to encroach on their midrange server and high end desktop markets. They're simply not stupid enough to continue to sell a chip that nobody wanted in the first place. The Pentium 4 was nothing more than a time saver and a way to develop and test technologies that they would need in the future for their server markets. (Hyperthreading was existant on the OLDEST Pentium 4 hardware, though not enabled since it was still very primative). And as you've noticed, lots of the Pentium 4 technologies have already been ported over into other product lines.
AMD is more and more concentrated on taking the server room from Intel. Once they've done this, they'll trickle home just the same way as Intel processors did in ages ago. And they're willing to sacrifice it all on their gamble that the industry won't shift off of x86 simply because it's too deeply embedded. They're not willing to bet on Microsoft and other software giants NOT creating software for a different platform (since Microsoft is really the end-all, be-all for the software), and instead, they embraced this lockin and extended it. The OS doesn't have to be natively compiled and optimized for their platform, and that gives them a huge advantage over the Itanium iron that they were aiming for. When performance really failed to hit the spec of highly optimized Itanium 2 code, they simply shifted gears and aimed it at Xeon instead. This was smarter because they know if they can get businesses to optimize and recompile, Xeon hardware will have to be left behind.
IBM on the other hand, says "fuck everyone else, we're doing it our own way". Working with Apple they developed a platform and got it some market share quickly. Next step: get it more market share by pushing Linux (which is outside of the control of the corporate giant of Microsoft, although this is being challenged by SCO, who was evidently paid off by Microsoft to launch such attacks and alligations). Not that Linux is any faster than anything written in Windows, but that it's cheaper, open, endlessly flexible and faster to update than anything Microsoft can throw at it. This is a safe bet. They're also aiming for the Itanium giant, and have nailed it pretty well with the Virginia Tech terascale project. Many say this is a win for Apple, when really, it's a win for PPC, which is IBM's baby.
Microsoft is really the key card right now. If they port Windows to PPC, it could royally screw both Intel, and AMD out of business. Luckily, Microsoft would take a lot of flac for doing this because of the companies that are so entrenched in X86 optimized code, that moving over to PPC would cost them millions, and they could simply move to x86 Linux instead of the next version of Windows.
So really, CPU's are becoming a lot like CPU's, but the industry doesn't care, and is in a very intersting position with Microsoft at the head. What I'd love to see is nVidia release a chip on a
"Victory means exit strategy, and it's important for the President to explain to us what the exit strategy is." G.W.Bush
The current IBM POWER4 and upcoming POWER5 chips are both dual-core chips. Here is a nice presentation(PDF format) about the POWER5; you can see in the die photos where there are two cores. There have also been rumors of a dual-core PowerPC based on it, but nothing concrete yet.
Broadcom (which bought SiByte) markets a dual-core, 1GHz 64-bit MIPS chip called the BCM1250 which has a lot of integrated networking goodies.
Finally, it bears pointing out that on the other side of Intel's severed corpus callosum, they're also working on a dual-core chip.
That's exactly what they try to avoid. Each core in a multi-core processors is simpler than a single processor of the current generation, but they make it up by putting two or more of them on the same chip. Another way to look at it is that the parallel execution units of a current generation processor are made even more autonomous, and this is made explicit by declaring them to to be separate processor cores.
The point is to use the available transistors on a chip as effictively as possible. For a long time computer architects used the growning number of transistors to enlarge caches and pipelines, add execution units, and add other niceties (e.g. branch prediction, MMX), but the gains have gotten less and less (and were sometimes dubious to begin with).
Multicore processors are only useful if people have enough parallelism in their applications to make it worthwhile. Therefore, it won't help every application, but that's also true for many tricks in existing architectures.
It's not clear to me that a dual core processor would take less power than a single core processor. Sure a dual-core processor _will_ take less power than two single-core processors on a board. So I suppose at a system-level a single core processor will take less power than a dual processor system, but the power problems we're seeing now are primarily at the chip-level.
BTW: As someone who 'knows' people that work at Intel, this decision was a pretty huge one on the 'Richter scale'. 1000s of people found out in the last couple of weeks that they were being redeployed to different projects (or making major changs on current projects). This decision is having a huge effect inside of Intel. I suspect that this kind of shake up means that the higher ups at Intel were very afraid that AMD is making major inroads and they finally realized that they couldn't keep going in the direction they were headed in without disasterous effects on marketshare.
teja vu.
Simply: cost. The CPU core is probably now well under 10% of the silicon area, the remainder being L1 cache and similar support circuitry. Adding a whole extra core adds very little to the total silicon - less than making the core more complex to handle ever deepening pipelines. Whereas adding a second complete chip, in its own package, plus the arbitration logic necessary to make the two chips work together, costs a lot more.
Consciousness is an illusion caused by an excess of self consciousness.
Um, AMD announced this in September last year.
"With coherent HyperTransport, it is inevitable that we will have multiple cores on a single chip. This is a tremendous opportunity because with our architecture the scaling is far superior to anything else that's out there., The Register quoted Mr. Sanders."
Also, see this: AMD CEO: "Dual-Core Opteron Will Shock the Hell Out of Everyone". Ruiz confirms dual core Opteron in 2005.
They say that Intel Tulsa (dual core Xeon) will arrive in about a year and Jonah (dual core Pentium M) is planned for 2005/2006.
So, nothing new here for AMD.
According to Reuters and the Wall Street Journal, Intel is supposed to officially announce today that they're not going to bother with the Tejas generation of PIVs/Xeons.
This ought not come as too much of a surprise to those of you who read this last March, and we openly wondered whether Tejas was going to see the light of day a little while back .
Yes, this a major announcement that will effectively knock Intel out of the box in the cutting-edge overclocking world for at least something close to eighteen months. This essentially leaves us with whatever AMD chooses to offer.
Nonetheless, the biggest aspect to this story is not the "what," but the "why."
A few days ago, the chief technology officer at IBM, Bernie Meyerson, told an industry forum that the traditional and expected increase in speed just from shrinking the manufacturing process is dead .
To quote:
"Somewhere between 130-nm and 90-nm the whole system fell apart. Things stopped working and nobody seemed to notice. . . . Scaling is already dead but nobody noticed it had stopped breathing and its lips had turned blue."
(This comes from the company that AMD paid $46 million dollars to help build 90nm chips, BTW. It also comes from the company that was supposed to have 3GHz 90nm PowerPC chips ready for Apple in a couple months, but is now talking about eventually getting to 2.5GHz.)
Meyerson said the biggest reason for the problem is power leakage, the same as what Intel has been saying. He also pointed out that the problem with power leakage is "nonlinear."
That's a fancy term for saying "it doesn't get slowly worse; you get past a certain point, and everything suddenly falls apart on you."
It's Not Quite Over
Mr. Meyerson is not saying "it's all over." What he is saying is that the era of easy, big gains from each new generation of processors is over. As he put it, "60 to 70 percent of the benefit of each new generation of manufacturing would have to come from innovation."
By that he means technologies like SOI and strained silicon, though he implied that these were not long-term fixes to the problem.
What is clear is that future technological advances are going to be a lot harder to do, cost a good deal more, and being a lot harder to work with than has been the case in the past. The old way of doing things is broken, and there's no mature alternative around at the moment.
Perhaps one will eventually show up, but the magic bag is empty at the moment, and it will probably take years to come up with some major new tricks.
In the meantime, progress will slow down.
Playing Noah's Ark
In all likelihood, Intel's short-term answer to this problem is to stop revving and start adding. Processors, that is. The son of Pentium-M which will become Intel's next generation will almost certainly be a two-headed beast. In short, a 6GHz processor won't be a 6GHz processor; it will be two 3s.
AMD plans to do exactly the same (which ought to tell you that SOI, good as it is, is no long-term fix to this problem).
This is hardly something either party would willingly want to do rather than increase speed, simply because the vast majority of current programming does not (or even cannot) work better with two-headed action.
It's certainly not something Microsoft want to deal with on the OS side, and probably is a big reason why Longhorn keeps getting pushed back, much less the armies of non-MS programmers out there.
It's going to happen because the hardware people don't have a choice in the matter.
I follow the SDK and GDN principles.. Spelling Dont Kount, Grammer Dont Neither