Will AMD's fastest CPU be obsolete next month?

Will AMD's fastest CPU be obsolete next month?

Summary: When I saw the first set of independent benchmark results pitting a mid-end Intel E6600 "Conroe" 2.4 GHz CPU (due next month) against the just released flagship Extreme Edition AMD FX-62 CPU, I started wondering if AMD worst nightmare was coming true. Intel's ~$250 E6600 CPU annihilated AMD's ~$1000 Extreme Edition AM2 based FX-62! This effectively means that AMD's flagship desktop performance CPU will be obsolete by the end of next month when Intel released the CPUs codenamed Conroe. The 2.4 GHz Conroe E6600 CPU is a 65 watt part while Intel's Extreme Edition Conroe CPU will operate at 2.93 GHz and still be 40 watts lower than AMD's FX-62 which runs at 120 watt TPD. AMD's power advantage over Intel's current Pentium 4 NetBurst architecture just vanished in to thin air with the introduction of Intel's Core 2 architecture next month.

SHARE:
TOPICS: Processors
225

When I attended AMD technology day last Thursday, senior AMD managers poked fun at Intel's NetBurst architecture and the fact that Intel will be "going back to their previous generation architecture, and that would be an improvement".  Most of the industry analysts in attendance began to laugh in the room and I couldn't quite figure out what's so funny and if they would still be laughing next month when Intel's Core2 architecture is released.

I asked some senior AMD managers what they were going to do It is beginning to look more and more certain that Intel's Core 2 architecture will "leap ahead"... when Intel releases their Core2 architecture next month and what they thought of the dire initial benchmarks posted at AnandTech and the response was "who set up the benchmarks" and more laughter ensued throughout the room.  What everyone was laughing about was the fact that Intel had actually set up the test bed for those initial benchmarks, but what's forgotten is that AnandTech (responding to readers and critics) made Intel change certain aspects of Intel's test bed to address perceived inequities in the first set of tests.  I say "perceived" inequities because in the end the changes didn't make a bit of difference and the results were just as bad for AMD's simulated FX-62 (without the faster socket and memory).  Since these results were theoretical because the actual AM2 based FX-62 with faster memory access weren't being used and the test bed wasn't completely independent, we couldn't draw any decisive conclusions.

So last Friday when I saw the first set of independent benchmark results pitting a mid-end Intel E6600 "Conroe" 2.4 GHz CPU (due next month) against the just released flagship extreme edition AMD FX-62 CPU, I started wondering if AMD worst nightmare was coming true.  Intel's ~$250 E6600 CPU annihilated AMD's ~$1000 Extreme Edition AM2 based FX-62!  This effectively means that AMD's flagship desktop performance CPU will be obsolete by the end of next month when Intel released the CPUs codenamed Conroe.  The 2.4 GHz Conroe E6600 CPU is a 65 watt part while Intel's Extreme Edition Conroe CPU will operate at 2.93 GHz and still be 40 watts lower than AMD's FX-62 which runs at 120 watt TDP.  AMD's power advantage over Intel's current Pentium 4 NetBurst architecture just vanished in to thin air with the introduction of Intel's Core 2 architecture next month.

If that wasn't enough of a beating, Adrian Kingsley-Hughes who writes for our new "Hardware 2.0" blog linked to these phenomenal overclocking feats with the Conroe 2.4 and Conroe 2.6 GHz CPUs clocking to 4.0 GHz and 4.26 GHz respectively!  I'm hearing that these kinds of numbers can be achieved with self-contained water coolers from multiple sources while the AMD FX-62 can barely get to 3.6 GHz with sub-zero temperatures.  As Adrian pointed out, even AMD's "4x4" which is two dual core CPUs and two dual core GPUs is a stop gap measure that won't be practical since you can do just as well with a single mid-end Conroe 2.6 GHz CPU clocked to 4.26 GHz at 1/6th the price in CPU cost.

While I'm still eager to see more independent results that replicate these results, it is beginning to look more and more certain that Intel's Core 2 architecture will "leap ahead" as Intel's new slogan implies.  AMD pointed out that they've been ahead of the game for 3 years (on most benchmarks and the results were always close) and it's inevitable for Intel to have a slight lead once in a while.  The problem here is that this new Intel lead is not the usual leapfrogging where one competitor edges out the other, it's a massive lead across the board!  AMD will be shifting to a 65 nm process by the end of the year and adding 128 bit floating point processors by the middle of next year though it's not certain if they can make a massive performance gain while making a massive reduction in power consumption.

Intel on the other hand told me that they won't be standing still and they don't ever intend to make the same mistake of allowing the NetBurst architecture to stay around for more than 4 years again.  The Core 2 architecture will only be around for 2 more years until Intel shifts to something new.  I asked Intel's representative if this is the kind of paranoia that would make Andy Grove proud and he laughed.  The truth of the matter is that AMD is what's making Intel paranoid because they've taken a beating for the last 2 years at the hands of AMD.  Who's going to win the processor wars doesn't matter because this is competition at its best and the consumer is the ultimate winner with better products at lower prices so let the wars begin!

[Update: 2:10 PM] David berlind wonders whether Intel has forever neutralized AMD's integrated memory controller advantage.

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

225 comments
Log in or register to join the discussion
  • It's curious, why Intel is playing hide and seek

    In all semi-verifiable tests of the 3GHZ Woodcrest (which is just Conroe in server)arranged by Intel itself, the chip is barely faster than Opteron 285 (2.6GHZ).
    http://sharikou.blogspot.com/2006/05/3ghz-woodcrest-against-26ghz-opteron.html

    In fact, Intel had to compare the 3GHZ Woodcrest against 2.2GHZ Opteron to show its chip in good light.

    http://sharikou.blogspot.com/2006/05/more-on-intels-woodcrest-performance.html

    So far, all Conroe benchmarks are controlled by Intel. http://www.reghardware.co.uk/2006/06/04/preview_intel_core_2_duo/
    Unlike AnandTech, hardware sites are reluctant to serve as Intel's PR machine any more.

    For power consumption, Intel is aiming at 130 watts for Conroe XE.

    http://sharikou.blogspot.com/2006/06/intel-runs-into-thermal-wall-again.html

    Intel better make Conroe 40% faster than K8, otherwise, the company will go belly up. The Conroe hype already caused 40% price decline of Intel CPUs and 20% decline of volume.
    http://www.theinquirer.net/?article=32136

    But even at 40% better performance (which is impossible, I tell you), Intel will be far below AMD's 4x4 platform in performance.

    4x4 will pin Intel permanently at half of AMD's performance.
    sharikou
    • Nope, got some facts wrong

      The Woodcrest benchmarks heavily favor Intel over Opteron under the benchmarks I've seen.

      "For power consumption, Intel is aiming at 130 watts for Conroe XE."

      That's just flat wrong. Woodcrest at 3 GHz is only 80 watts TPD.

      "4x4 will pin Intel permanently at half of AMD's performance."

      First of all, a single overclocked Conroe 2.6 GHz CPUs will match that CPU performance. What makes you think Intel can't murder them even more with a dual socket configuration?
      georgeou
      • More Ou-Stats?

        When in doubt drag out the overclocking now?

        Listen, if the chip was designed to run at those speed then Intel would sell them rated to those speeds.

        The fact that you can drive the chip that much faster doesn't do anything to address what design issues Intel has seen that makes them think that the chip should only be rated for the 2.6 GHz they're selling it at.
        Robert Crocker
        • its not design issues

          they clock them low for the same reason they made Netburst.... marketing.

          why sell you a 3.5ghz version when they can sell you a 2.4ghz version, and a year and ahalf from now then sell you a 3.5ghz version?

          Its good for making money, as the 2.4ghz will be good enough for the competition at the time its released.
          doh123
          • Well, it certainly isn't a marketing issue

            Anyone knows people still look at clock speeds. A 4GHz chip for $200 is simply going to appear more attractive than a 2GHz chip at $400. I seriously doubt they would artificially suppress the true clock speeds.
            MerryOtter
          • but

            but when the 2GHz chip outperforms the competition, and no one has the 4GHz chip for sale, they arent going to sell it right now, that makes them less money in the long run and more overhead in beating themselves in the future so people will need to buy new again.
            doh123
          • it's not marketing

            Intel doesn't "clock them low" for marketing purposes. They rate the CPU at 2.6 GHz because of a big design issue. It's called POWER.

            The chip is rated at the frequency at which it can run and still meet the requirements of the platform. At 65W, given current technology and processor reliability limits, 2.6GHz is what you get. Sure, you can overclock it...but it doesn't do Intel any good to underclock their processors when they're competing to regain the performance edge over AMD.
            mk76
      • Intel's Dual Core Layout isn't that good

        Even so, AMD has had better luck with the intregated memory controller. Something that Intel still doesn't plan on releasing.

        I personally hope to reap from this when AMD drops the price on their Athlon X2 chips. I have a socket 939 board at home craving an upgrade
        nucrash
      • "4x4 will pin Intel permanently at half of AMD's performance."

        From a marketing perspective, maybe. Reality, no way!
        Adrian Kingsley-Hughes
      • George, please read the news

        Not everyone is so gullible as to believe Intel's own "setup". I provided the links to test data, why don't you just read them?

        1) Woodcrest/Clovertown benchmarks
        Clovertown (quad-Conroe) http://sharikou.blogspot.com/2006/04/clovertown-scores-revealed.html

        3GHZ Woodcrest vs 2.6GHZ Opteron

        http://www.dailytech.com/article.aspx?newsid=2487

        http://sharikou.blogspot.com/2006/05/3ghz-woodcrest-against-26ghz-opteron.html (more detail)

        Intel's fraud:

        http://sharikou.blogspot.com/2006/05/more-on-intels-woodcrest-performance.html

        http://sharikou.blogspot.com/2006/04/conroe-performance-claim-being-busted.html

        2) Conroe XE up to 130 watts
        This is off the mouth of Intel's David Tuhy, who just demoed a liquid cooled Conroe XE to a reporter in June.

        See links and discussion at

        http://sharikou.blogspot.com/2006/06/intel-runs-into-thermal-wall-again.html

        3) "What makes you think Intel can't murder them even more with a dual socket configuration?"

        Even Intel fanboys know Intel can't put together a cost effective and performing dual socket platform for desktop. 4x4 is basically 4 way computing(4 cores), AMD rules once you go to 4 cores and up. Conroe is an improvement of the execution engine in single threaded loads, once you have 4 cores, Intel's shared FSB can only give 250MHZ to each of the core. That's why AMD owns 48% of the 4 way server market. I told you that 4x4 will pin Intel at about half of AMD's performance for the forseeable future. Just believe me. Judging from your comments above, you don't have enough background on computer science to actually understand most of the topic. This also shows Intel can easily mislead and deceive non-technical people with bogus claims.

        For discussion on 4x4, see the comments area in the following pages

        http://sharikou.blogspot.com/2006/06/intel-runs-into-thermal-wall-again.html

        http://sharikou.blogspot.com/2006/06/amd-does-4x4-desktop-computing.html
        sharikou
        • That's true! I'm sensing "vaporware" from Intel.

          One Intel processor beating AMD AM2 FX-62 processor!!!???

          HAAA!!!! :D :D :D Maybe!!! :D

          Sorry for my lame comment but the thing about Intel...I don't know... Proving thier own bunchmarks? No proof.

          Sorry Intel! I'm glad you have good overhyping skills! I love it when you mislead non-technical and technical people like me but not any more because I will take any of your most expensive processor home and burn up your processor any time!!!! :D :D :D
          Grayson Peddie
        • Some blogs...just dont seem right...at all.

          Every time I read stuff from this website (Journal of Pervasive 64 bit Computing) it comes across as being at odds with everything else on the web. To me that implies a great deal of either "wishful thinking" or outright B.S.

          http://www.dailytech.com/article.aspx?newsid=1185

          http://www.youtube.com/watch?v=whhhGud47TI

          http://www.hexus.net/content/item.php?item=5692

          http://tinyurl.com/hwtu2

          http://tinyurl.com/g4qt3

          And there are more. Personally I don?t care which CPU is better, I would just like to know which one. And then there is price to consider. A CPU that is 15% better in performance and 50% more expensive is a poor buy, unless you are well off and price is no object. This is particularly true today where most of the top CPU?s from both AMD and Intel when paired with a good GPU and performance memory provides lightning quick response in any case. For most people paying 50% more for a CPU that is currently overkill in performance standards is worthless.

          If Intel can make a CPU that?s even close to AMD?s top performers and is noticeably cheaper Intel is the hands down winner.
          Cayble
          • Did you miss the newer benchmarks?

            This one, by famous pro-Intel Tomshardware. This is again an Intel controlled benchmark.
            Intel pre-configured the Conroe and gthe benchmark programs, and people were only allowed to run these benchmarks (see http://www.tomshardware.com/2006/06/05/first_benchmarks_conroe_vs_fx-62/page4.html ).

            However, there is an interesting trend here. Back at the IDF, Intel claimed up to 40% lead in gaming. Now, Conroe is occasionally slower than FX? see

            http://www.tomshardware.com/2006/06/05/first_benchmarks_conroe_vs_fx-62/page5.html

            I wonder why Intel keep doing this kind of guerilla benchmarketing. They are all bogus.

            http://sharikou.blogspot.com/2006/04/conroe-performance-claim-being-busted.html
            sharikou
          • Wow. That?s a huge stretch..And not journalism

            You first point out a link to a highly respected website, Tomshardware, but not before doing your best to jade the reader?s opinion by pointing out that in you feel they are pro-Intel. Not good given that Tomshardware plainly states all the parameters and restrictions in the testing. And they didn?t do anything to hobble the compared AMD system either. And the end result is that the Intel CPU comes out considerably better overall. The link and preliminary intro to it you gave provides absolutely no support for any pro AMD consideration.

            You later give a link to your own blog, with statistics gleaned from another website that claims to have done a test on an Intel 2.39 core duo compared to an AMD overclocked to 2.8. Even in your clockspeed adjusted comparison between the two you still admit that in some instances the core duo would destroy the AMD if the Intel was running at 2.8 as well. What is proven here? Seriously?? Even in the worst case scenario found within the evidence of what you hold out to be facts, the Intel CPU still looks quite good.

            You give the impression of someone who is privy to incontrovertible evidence that Intel Conroy core duo CPU?s cant keep up to AMD processors, yet your own evidence keeps showing that they likely can. And if your right, and Tomshardware is pro-Intel, and that kind of thing leads to poor journalism, then the fact you are without any doubt an overblown pro-AMD blogger then you have to also accept the fact that the same standards must be applied to you as Tomshardware, and that means if they cant be trusted due to their bias, despite their solid reputation, you can be trusted far less.

            If you want wide spread respect in the field, drop your bias and stick with reality when reality is clear, and when its not so clear, at least don?t color the facts to suit your tastes.
            Wow. That?s a huge stretch..And not journalism
            Cayble
          • Agree!

            What can i say other than a big AGREE? :)
            ck123
      • Message has been deleted.

        myfevertoy
      • RE: Will AMD's fastest CPU be obsolete next month?

        Every time you achieve a hinder, you may perhaps advance for full approval as of
        <a href="http://caviar-store.com/">Buy Caviar Online</a> | <a href="http://caviar-store.com/">Buy <a href="htt</a>p://caviar-store.com/">Black Caviar</a> | Beluga Caviar
        Jonhsonson
    • Any sources besides your own blog?

      I took the bait and read your blog entries. You seem very good at focusing on only part of the story, while completely ignoring pertinent facts that contradict your opinions.

      The fact that you refuse to even acknowledge the possibility that Intel might have made a better performing chip than AMD indicates your clear bias and destroys your credibility. The fact that you often quote your own blog as credible evidence doesn't help.

      For those interested, here's an unbiased look at sharikou's posting history:

      http://www.google.com/search?hl=en&q=sharikou&btnG=Google+Search
      billywill
      • My blog serves as an archive of links to various news

        Every point I made was backed up by external sources I linked to. Just click the links.

        http://sharikou.blogspot.com
        sharikou
        • trying to create traffic for your blog?

          Guys, let's not wasting time to visit his blog unless you wanna have some fun :)

          allow me to summarise his blog for you here, just with few points and rules of thumb
          1) AMD is ALWAYS better than Intel
          2) When benchmark sites like tomshardware and hexus shows messages against point number 1, he will call them a bunch of kiddies and pro-intel
          3) Uses apple to orange (huge gap in price and some diff in freq) comparison to prove his point. When people points it out, he would say the apple will become orange in 2 months time.
          4) pick and choose the number/facts that he wanna see.
          5) When run out of idea, refer to point number 1, but this round AMD 'will be' better than Intel

          Did I miss anyhting? In case you did visit his blog, please help to add in things that i missed out in the summary :)
          ck123