Intel on AMD's early dual-core wins: "Not so fast"

Intel on AMD's early dual-core wins: "Not so fast"

Summary: Updated: Literally and figuratively.Still smarting from having to swallow its pride over the success of AMD's 32/64 hybrid technology (AMD64), Intel appears once again to be on the short end of AMD's technological stick -- this time, over dual-core chip technology (the technology that basically packs two CPUs into one chip).

SHARE:
TOPICS: Processors
16

Updated: Literally and figuratively.

Still smarting from having to swallow its pride over the success of AMD's 32/64 hybrid technology (AMD64), Intel appears once again to be on the short end of AMD's technological stick -- this time, over dual-core chip technology (the technology that basically packs two CPUs into one chip). 

Unlike with the hybrid technology, Intel was first to market with a dual-core chip -- the Pentium Processor Extreme Edition.  In bloodline terms, the "EE" is really a Pentium 4 -- a 3.2 GHz part that's based on Intel's Smithfield core, includes Intel's hyperthreading (HT)  technology, and that's been coupled with Intel's 945 chipset.  Though not as advanced in features as the chipsets that Intel is coupling with its next wave of dual core desktop chips (the 2.8, 3.0, and 3.2 GHz Pentium Ds, due this month according to recent statements made by Intel president Paul Otellini), the EE/945 combo allowed Intel to ship a dual-core chip that served the needs of certain enthusiast segments -- particularly gamers -- where the software designers were more likely to take advantage of the HT technology. (HT support must be explicitly supported by the developer in order for users to take full advantage of it.)  Among the manufacturers to jump on the April 18, 2005 dual core release with systems targeting gamers were Dell, Alienware, and Falcon.

But, while AMD may be finishing in second place in terms of timing, the company is scoring points on a number of other fronts with its dual core technology, otherwise known as X2.  According to CNET's first wave of reviews (as well as those of ExtremeTech and Toms Hardware), AMD's Athlon 64 X2 has knocked the rest of the desktops out of the ballpark.  The early results make it appear as though AMD could once again be trumping Intel.  Not only did the AMD X2 outperform the Intel EE on most tests (a testimony to AMD's design considering that the X2 was working with slower memory), but current Athlon 64 system owners will be overjoyed to know that they don't have to go out and buy a new system or motherboard to get the benefits of the X2's performance.  Instead, the do-it-yourself set need only pop out the old chip and replace it with the new.  This cut bait and run approach simply isn't available in the Intel world.

But, as the headline to this blog suggests, Intel is hinting that we shouldn’t be too fast to pass judgment and that time may tell a different story.  For starters, according to Intel, AMD's integration of the memory controller with the processor via its Direct Connect Architecture (and the way that results in superior performance) may appear to have been a prescient design choice now that AMD is winning benchmarks with components that look like tortoises (eg: DDR memory) compared to Intel's hares (eg: DDR2 memory).  But sooner or later, Intel warns, AMD's design choice will run out of gas and AMD will be forced to make changes on the chipset and memory controller fronts in order to keep up with processor enhancements.  When it does, says Intel, buyers and OEMs of AMD product will have to requalify the entire platform while Intel will already be in market, for example, with DDR2-based systems (ones that it thinks will seize the performance lead) and cranking on the volume.   Indeed, a story in today's Digitimes indicates that a complete overhaul of the AMD platform is on its way. 

Brute force may be another reason that Intel is coming up short.  Today's software lacks the necessary multithreading finesse to take advantage of technologies like Intel's HT.  So, in general, benchmarks will reflect the sheer brute force capabilities -- the integrated memory controller for example -- of the design.  Intel's Pentium EE may phsically be a dual core chip, but logically, becasue of the HT technology, it's really four cores.   For now, though, the software isn't good enough to make the EE behave like four cores.  But, over time, as evidenced by inclusion of similar threading technologies in competing chip architectures such as Power and SPARC, that lack of finesse is expected to be overcome by the prevalence of more sophisticated software development tools and techniques.  Once those tools and techniques become more prevalent, Intel is expecting that it will be the equivalent of Star Wars' Han Solo kicking the Millennium Falcon into hyperspace. (And it will probably happen with the requisite false starts too.)   One reason, says Intel, that it will be like hyperspace is because the aforementioned factor of two that turns a two core chip into a four core chip (logically) will eventually be a factor of four, then eight, and so on. 

To compete, it's safe to assume that AMD will introduce something that's HT-compatible, but so far it's not in the current gear.  Nor is it in Intel's Pentium Ds.  For all intents and purposes, the EE and the D are the same chip built on the same core (the Smithfield core), but the big difference will be in the chipset.  Here again, Intel is introducing a bevy of built-in features such as support for high-definition audio, RAID (redundant array of inexpensive disks), PCI Express, and Vanderpool-like virtualization. (AMD has its own virtualization support coming in a technology called Pacifica.)  Intel claims that with these features, the "D" platform will be targeting the mainstream users (as opposed to the way the already released EE targets the enthusiast).

Finally, Intel is arguing that it has so many dual core projects in the works (15 in all) and that dual core will be so integral to everything Intel does (from mobile to desktop to servers), and that its fab technologies will be so geared towards volume (eventually extracting 65 nm designs from 300 mm wafers) that the chipmaker will be able to serve the mass market in ways that its competitors cannot.

So, will Intel have the last laugh?  The last time I said that it would (in the pre-Opteron days), the AMD fanatics in ZDNet's readership took me to the woodshed for a well-deserved spanking.  AMD clearly got Intel's attention after Intel finally responded with a 32/64 bit hybrid design of its own.  So, having learned my lesson, I'm not one to so easily count AMD out again.  Not only that, AMD's director of server and workstation marketing Pat Patla clearly sees some of Intel's positioning as FUD (fear, uncertainty, and doubt).   

Regarding the volume issue, Patla told me via email "Sales of AMD Dual-Core [Editor's Note: "Dual Core" stricken from AMD's original statement at AMD's request] Opteron processors, introduced just three weeks ago, are already exceeding expectations, yet we foresee no capacity limitations.  Manufacturing capacity is simply not a concern for AMD."  On the memory design issues, Patla was equally bullish about DDR's future, and reiterated the advantages of processor-swapping, saying "DDR memory will be available for years to come.  By upgrading to AMD Dual-Core Opteron processors, DIY users can see 30- to 90-percent performance increases without having to trade-in their platform."  And on the issue of whether Intel's HT technology will eventually give software the equivalent of hyperspace, Patla says "HyperThreading, like large caches, is needed by our competitor to account for architecture inefficiencies.  Most major server vendors turn HT off when they run benchmarks, and independent on-line reviewers have shown that the benefit of HT in Intel's dual-core P4 EE product is almost negligible. AMD Dual-Core Opteron processors exceed the performance increases HT only promises."  This last statement is currently (emphasis on currently) true and pretty much jives with what Intel is saying.  What isn't clear is what the future holds.  It never is.

Topic: Processors

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

16 comments
Log in or register to join the discussion
  • Architecture rules

    But heat is death. Don't underestimate the impact of Intel's lethal 200+ watts of dissipation compared to AMD's 30 or so. Intel may close some of that gap when they get some real dual-core chips (as distinct from multiple chips in a single package) onto the market.

    As for memory, Intel is playing the only hand they have. Their whole system architecture bottlenecks at the front-side-bus (FSB) going to the memory controller. As a result, you can add all the processors you like but you're stuck with the memory that the controller provides, and that in turn is limited by physics: the faster the memory, the less you can connect.

    DDR2 allows faster operation of that bottleneck FSB but can't get around the physics of the memory bus and doesn't much help the latency of the memory itself -- which is the real problem. Newer systems (you'll see, I can't talk) get around the memory-connection problem but are expensive in dollars, cost, and heat (there's that H-word again.)

    AMD doesn't [b]need[/b] the higher bandwidths of DDR2 because their memory bandwidth naturally scales with the number of processors. The latency of DDR is if anything a bit less than DDR2's, too, so there's no hurry to go to the newer memory until it reaches EOL on its own.

    I would never underestimate Intel's enormous manufacturing power or their ability to throw engineering resources ([b]15[/b] projects!?!) at a problem, but they're playing damage control here. For years they've been denying that there's any value in 32/64 or multicore, so now that AMD is calling the technology shots they're going to have to work extra hard to regain mindshare. (Starting with widely-read technology columnists.)

    Intel's big challenges:
    3) Their memory architecture sucks.
    2) Their chips suck (electricity)
    1) They haven't admitted yet that they don't have a Divine charter to rule the industry

    Y'all have fun now.
    Yagotta B. Kidding
    • Power consumption is a problem for Intel

      The Intel Pentium D uses more power in idle than the AMD 64 X2 at maximum burst. Power consumption can get very expensive to power and cool in a data center.

      When Intel's Dual Core Pentium M chips come out, they may regain the power advantage.
      george_ou
      • Hamlet

        [i]The Intel Pentium D uses more power in idle than the AMD 64 X2 at maximum burst.[/i]

        Make sure you read the fine print.

        Intel and AMD define "Thermal Design Power" differently. AMD defines it as the maximum power that the chip can suck, period (multiply maximum current by supply voltage for each supply, add the products). Intel defines it more as a guideline, with the footnote that the processor will thermally throttle itself if it gets too hot from burning more than that; the actual sum-of-products is quite a bit higher.

        [i]When Intel's Dual Core Pentium M chips come out, they may regain the power advantage.[/i]

        They might. Heaven, Earth, Horatio, all that. As I wrote, I won't discount the amount of engineering that Intel can throw at a problem. The first step is admitting that they have a problem.
        Yagotta B. Kidding
    • A fellow hardware junky I see...

      "3) Their memory architecture sucks."

      I agree, how much do they think they can keep clocking the same ol'? That's like running around high on coffee day and night for years. Eventually you crash.

      "2) Their chips suck (electricity)"

      With energy prices and the environment being a going concern, this is a valid point. I switched every bulb in the house to flourescent energy savers. I will be damned if I burn the saved 'juice' on an Intel chip.

      "1) They haven't admitted yet that they don't have a Divine charter to rule the industry"

      Lmfao! Have ya noticed any 'giant' gets p*ssy when they are beat at their own game. I always was an 'underdog' kind of guy anyway.
      MepisLINUXuser
  • The AMD is a server chip while the Intel Pentium D is a desktop chip

    The AMD product is for servers and the Intel product is for Desktop systems. The server based AMD product will cost many times more than the Intel Pentium D desktop product. It may not be fair to compare the two given the price differences.

    Additionally, it also depends on what applications you want to run. Some games will do better on the Intel chip while other games do better on AMD. Some applications like Windows Media Encoder does much better on Intel than AMD while other applications prefer AMD. The bottom line is that you have to look up your preferred application and see how it performs on AMD or Intel. Then take price in to consideration and make a choice.

    You can look up CPU performances by application here.

    http://blogs.zdnet.com/Ou/?p=54
    george_ou
    • Server chip?

      AMD's X2 is a desktop chip. Opteron dual core, the one for servers, doesn't use the X2 designation.

      As far as I know, you have to choose a SYSTEM based on specific application performance and cost. CPU benchmarks on a testbed are informative, but they don't tell you how other systems are going to perform, and don't get into the cost/performace trade-offs in doing something like choosing low latency CAS 2 DDR DRAM vs choosing CAS DDR 3. And personally I would never use Tom's as authoritative on any topic, it's much safer to look at multiple reviews.
      HDavis
      • AMD desktop chip isn't due till later

        Sorry, you might be right about that naming scheme for AMD's desktop chip. It's just that AMD is releasing their server chip first while Intel is releasing their desktop dual core first.

        You're right about the benchmarks. But the tool I posted just gives you the raw numbers for those specific applications and doesn't really give any opinions. You're right about looking at other benchmarks though because the more data you have the better. If you rely on something like Windows Media Encoder or you rely on LAME MP3 encoder, the tool on Tom's site is very useful. However as I noted in my blog on the CPU tool, I wish it had more applications and more CPUs listed. They may fix that later by updating it. I just like the concept of using an online database that you can research yourself. Storagereview.com had been running an HD performance database for some time now.
        george_ou
        • Subject check?

          "
          It's just that AMD is releasing their server chip first while Intel is releasing their desktop dual core first.
          "

          The article and it's links are specifically the X2 desktop dual core, so dual core Opteron pricing isn't relevant. I don't think I'll see either CPU soon on Newegg, so I'm not concerned about who is first or second, just what I can buy:-))))

          I agreee about having good hardware comparisons that are convenient. I build my systems, and component choice is tedious given all the tradeoffs involved. But my prior experience with Toms and CPUs is you have to be extremely careful in understanding their tests. And the same can be said for other websites. Memory and hard drives are much much easier to get good benchmarks on that can help make a reliable buying decision.
          HDavis
    • CNET says X2, not Opteron

      "
      Two weeks ago, AMD released its dual-core Opteron server chips and announced that we'd soon see dual-core tech brought to the desktop in the form of its Athlon 64 X2 processors. CNET Labs was able to test one of the new chips, and today we can share performance figures. The results are dramatic. AMD's new dual-core Athlon 64 X2 4800+ CPU hands the company a decisive victory over rival Intel; a system from AMD using the X2 4800+ bested an Intel-submitted PC with the Pentium Extreme Edition 840 (PEE 840) on every one of CNET Labs' dual-core benchmarks.
      "
      HDavis
    • Optimization

      [i]Additionally, it also depends on what applications you want to run. Some games will do better on the Intel chip while other games do better on AMD. Some applications like Windows Media Encoder does much better on Intel than AMD while other applications prefer AMD. The bottom line is that you have to look up your preferred application and see how it performs on AMD or Intel.[/i]

      Given the radical differences in the performance costs of branch missprediction and cache miss, that's hardly surprising. Some of it is due to the nature of the application (those that are more SIMD-ish work better on Intel's deep pipes) and some on architecture-specific compiler optimizations.

      This is a lot more of an issue for binary-only environments; you can bet that, like me, Microsoft recompiles their software for their AMD-based servers. With the differences between the AMD and Intel devices it's a good practice if you can do it.

      I don't think that Intel has updated their utterly superb compiler to work with the 32/64 instruction set yet, though.
      Yagotta B. Kidding
  • What will AMD do in the meantime?

    Most large scale games and applications take months to complete, if they are just now beginning to incorporate HT optimizations now, AMD has nothing to worry about. Months in the tech world is a very long time. The idea that Intel well go into 'hyperspace' assumes that while they're waiting for the software and market climate to favor them, AMD is just sitting on their hands.
    Its really only a matter of time before AMD adds support for DDR2, and HT compatible instructions, so by time these things really become must-have features, rather than just marketing gimmicks, they will just be a nonissue.
    emcee_z
  • Intel's woes and illogical arguments

    Intel's biggest problem is the years of design that they spent on increasing the clock speed at the expense of everything else. Now they're hobbled with an FSB bottleneck and thermal dissipation issues that only firefighters and nuclear engineers should have to worry about. It's all well and good for them to have smoking (almost literally) fast processors, but if they spend 30+% of their cycles waiting for data, they're no better than a much slower processor.

    [i]Intel warns, AMD?s design choice will run out of gas and AMD will be forced to make changes on the chipset and memory controller fronts in order to keep up with processor enhancements.[/i]

    But the people that bought their AMD systems a year or more ago can still use the newest AMD chips with just a BIOS update. Try using an Intel HT chip in a non-HT motherboard and see what kind of performance you get(if any). Intel's argument is flawed because they have already done the same thing, and more than likely will have to do so again later.
    Letophoro
  • Just wait, we'll have an answer real soon now

    "Just wait, we'll have an answer real soon now." That's all Intel is saying. That's all Intel ever says. The years of blindly going in the wrong direction have caught up with Intel, but now Intel can't seem to get caught up to the competition anymore.

    Don't fool yourself into believing that Intel actually introduced its dual-core chip first. If this were homework, AMD introduced an A+ paper, thoroughly researched and prepared months in advance; Intel introduced a cobbled together a C- paper, hurriedly put together the night before due date. Intel hyped the hell out of dual-core prior to the launch, but now it's embarrassed that AMD has the superior solution, which is being borne out by the benchmarks. Of course the AMD's are performing better: they were designed right from the start with multicores in mind.

    So now Intel has to FUD something unimportant like the memory technology -- it's got the newer DDR2, but AMD still uses the older DDR. However, this assertion might also come back to embarrass Intel, because AMD is able to squeeze better performance out of DDR than Intel can out of DDR2. AMD knows exactly when it will be time to go to DDR2, because right now it's still well ahead using just DDR, so it has plenty of time to switch. In fact, it's too early for AMD to switch to DDR2, there isn't enough of a performance advantages between current generation DDR and DDR2 to overcome the DDR2's performance disadvantages yet; another couple of incremental speed increases are required on DDR2 for it to be worthwhile.

    The assertion that Hyperthreading will allow Intel "to go into hyperspace" is also FUD. Once multithreaded software is more prevalent it will benefit AMD's dual-cores just as much as it will Intel's.
    bbbl67
  • DDR just keeps ahead of DDR2 continuously

    [quote]For starters, according to Intel, AMD?s integration of the memory controller with the processor via its Direct Connect Architecture (and the way that results in superior performance) may appear to have been a prescient design choice now that AMD is winning benchmarks with components that look like tortoises (eg: DDR memory) compared to Intel?s hares (eg: DDR2 memory). But sooner or later, Intel warns, AMD?s design choice will run out of gas and AMD will be forced to make changes on the chipset and memory controller fronts in order to keep up with processor enhancements. [/quote]

    No sooner does DDR2 make a clock speed advance than DDR makes one that stays ahead of it. DDR has lower latency than DDR2, and DDR2 has to have a big lead in clock speed ahead of DDR before its higher latency can be overcome. OCZ just introduced some DDR DRAM that are rated at 600Mhz! the highest DDR2 parts are rated at 667Mhz right now, not nearly enough to overcome the latency issue. At current rates, DDR2-667 is about the same performance as DDR-400.

    X-bit labs - Articles - DDR600 SDRAM Tested: OCZ EL DDR PC-4800 Dual Channel Platinum Review
    http://www.xbitlabs.com/articles/memory/display/ocz-ddr600.html
    bbbl67
  • Simply unbelievable.

    It's simply unbelievable how slavishly you parrot Intel's press releases. Have you never heard the terms 'critical thinking' and 'think for yourself'?

    Cases in point:
    "Unlike with the hybrid technology, Intel was first to market with a dual-core chip..."
    No they weren't. IBM and Motorola both had dual-core chips out earlier. But even if you corrected that to mean only x86 chips it wouldn't be true. Intel simply moved up the date of their press release to three days before AMD launched the dual-core Opterons. They were certainly not 'first to market', because you *still* can't buy their dual-core chip.

    "But, while AMD may be finishing in second place in terms of timing..."
    Intel must be proud. "Hey, if we can't actually be first, at least we can dupe the press into thinking we are."

    "Intel is hinting that we shouldn?t be too fast to pass judgment..."
    I take it this means Intel has someone whispering in your ear and feeding you this trash. (sorry for the mixed metaphor.) Specifically, they told you that AMD's on-chip memory controller is only 'winning benchmarks', while *Intel* supports DDR2. Well gosh, DDR2 *must* be better than plain old DDR. We don't need no stinking benchmarks.

    "But sooner or later, Intel warns, AMD?s design choice will run out of gas..."
    Horrors! AMD's will run out of gas, but Intel's won't? NEWS FLASH: Intel's already *has*. That's why they can't make a 4GHz P4, and why they're over a year behind AMD on dual-core chips.

    "Brute force may be another reason that Intel is coming up short."
    This is funny, because 'brute force' is exactly what Intel has been relying on in jacking up clock speed. You know we don't even have to read the rest of the paragraph to know it will end with something like 'but eventually software will take advantage of it.'

    "To compete, it?s safe to assume that AMD will introduce something that?s HT-compatible, but so far it?s not in the current gear."
    This is absolutely hilarious! After listing item after item that will supposedly give Intel an advantage in the future, you now fault AMD for not having them in current gear. Well gosh, AMD's chips are superior *without* HT. I guess that means they'll be even better when they do add HT. Shouldn't that be a plus on AMD's side? But Intel has managed to convince you it's a minus. Good God man, THINK!

    "So, having learned my lesson, ..."
    No you didn't. The lesson is not that you can't draw conclusions (good or bad) about AMD. The lesson is to think for yourself, not just parrot back what Intel flacks tell you over lunch.
    gordon@...
  • Intel FUD vs. AMD Reality

    Intel's latest gamesmenship reminds me of IBM's "paper machines" back in the 1960's. IBM was getting its ass kicked by Seymour Cray's brainchild sold by CDC (Control Data Corporation). Tom Watson ruefully noted at the time that IBM's premier research arm was being smoked by some two dozen CDC employees, including the janitor.

    IBM's response was the "paper machine" with all sorts of features and performance that simply blew away CDC offerings at a fraction of the cost. The result was that the market waited for IBM's new "machine" rather than jump to CDC gear.

    There was only one problem. No one could buy the "paper machine". It did not exist and was never delivered. A day late and a dollar short, CDC sued IBM over it and won. But it was already too late and CDC is history.

    The lemmings in the market had already "bought" the paper machine.

    He who forgets the past is doomed to repeat it.
    dark_stream