AMD Athlon 64 FX-57: The Fastest Single Core

by Derek Wilson on 6/27/2005 11:47 AM EST
Comments Locked

56 Comments

Back to Article

  • blckgrffn - Wednesday, June 29, 2005 - link

    The warranty stops as soon as they changed heatsinks. That's it, that was where it should have stopped evidently ;)
  • Drazula - Tuesday, June 28, 2005 - link

    I hate articles like that. You go through the paces of testing and then recommend a solution that wasn't even tested. Why not include a dual core AMD for comparison? As it is, the article is useless.
  • Viditor - Tuesday, June 28, 2005 - link

    blckgrffn - "That said, where does it stop?"

    I would say that you stop where the warranty does...
    I.e. no overclocking results except to say it was or wasn't stable at x.x Ghz...Because (as Jarred points out quite correctly) OC can be a very hit and miss proposition.
  • composer - Tuesday, June 28, 2005 - link

    JarredWalton, I get 2.74 stable, air cooling which beats dual opeteron benchmarks, and also runs all the benchmarks stable.

    We use our PC's for audio, and it seems that even the X2's perform about the same as a single AMD64 overclocked to 2.75 in audio tests (VST plug ins).

    Look at this chart:
    http://www.adkproaudio.com/benchmarks.cfm

    Top graph, it shows the X2 at 58% using 256 samples......we get 58% using AMD 64 overclocked but using the older Nuendo 2.2 version, the newer version Nuendo 3 we get 68% however it's known that the security of the new steinberg program uses 10-15% CPU cycles.

    Just some thoughts.

  • composer - Tuesday, June 28, 2005 - link

  • L3p3rM355i4h - Tuesday, June 28, 2005 - link

    Some one tell me why you would buy a $1000 single core proc, when you could buy a $1000 dual core proc thats going to be soooo much better?
  • ElFenix - Tuesday, June 28, 2005 - link

    i've been asking them to get an editor for a long time. at one time, one of them actually replied and asked what kind of editor. i replied that they needed an english editor, and never heard back. they especially need one with some of the newer authors they have.
  • blckgrffn - Tuesday, June 28, 2005 - link

    Gotcha, Viditor - but if that is what everyone wants, we should also include the tests done. Yes, I can see how that would have bee a better review, putting say, a 3700+ San Diego and a FX-57 vs each other with all the most expensive goodies and see who came out on top. Heck, SLI some 7800GTX's too, we might as well see how high we can go :)

    That said, where does it stop? We want to see it under phase, too, with the 7800's oc'ed under chilled water, and some DDR600 @ 2-2-2-10! ;P
  • AtaStrumf - Tuesday, June 28, 2005 - link

    Calin, I meant in terms of overclocking! That's what OC stands for, doh! If you still don't get it, never mind, just know that what I said makes perfect sense ;-)
  • Zebo - Tuesday, June 28, 2005 - link

    Hey Tallon:

    I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdanieg.
    The phaonmneal pweor of the hmuan mnid. Aoccdrnig to a rscheearch at
    Cmabrigde Uinervtisy, it deosn't mttaer inwaht oredr the ltteers in a wrod
    are, the olny iprmoatnt tihng is taht the frist and lsat ltteer be in the
    rghit pclae. The rset can be a taotl mses and you can sitll raed it
    wouthit a porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey
    lteter by istlef, but the wrod as a wlohe. Amzanig huh? yaeh and I
    awlyas thought slpeling was ipmorantt.
  • Calin - Tuesday, June 28, 2005 - link

    Anyway, what's with the data analysis benchmark? Looks like every Pentium4 beats the hell out of every Athlon64.
    If this is true, then I feel my suggestions to buy Athlon64 might not always be correct
  • Calin - Tuesday, June 28, 2005 - link

    An processor with 1M level2 cache and dual channel memory controller is a preview of what to expect from a nice cheap 256K cache single channel memory controller? I really really don't think so
  • AtaStrumf - Tuesday, June 28, 2005 - link

    This isn't just a speed bump, it's a new revision (SH-E4 I think). Check PCU-Z next time!

    New Semprons with x86-64 will be even newer revisons - DH-E6 (don't know if that's good or bad yet, but very likely good) so OCing of this new E4 is like the most important thing here because nobody's gonna buy $1000 CPUs, nobody here anyway.

    FX-57 is a preview of what to expect of a nice cheap 3300+ Sempron and all you gave it was one line. Try a little harder next time please! If it's true that everybody else is getting 3,0 GHz on air, than that is great news. Just what we have been waiting for!!!
  • JarredWalton - Tuesday, June 28, 2005 - link

    Overclocking is half luck, and there's no telling how the OC benches at other places are really performed. 2.75 GHz "easy" on a Winchester? Fat chance. Maybe with water cooling and the proper motherboard, but even then it's not guaranteed. And don't think the Venice cores are much better - they're really about the same, which means 2.6 to 2.7 GHz air cooled is the typical maximum for a truly stable system.

    Anyway, if you're into serious cooling and overclocking, the FX-57 might be a bit better than the FX-55. 90nm vs. 130nm I believe, so hopefully it does better. I just can't see dropping $1000 for a single threaded CPU, though. A 4000+ San Diego does almost as well at half the cost, even with overclocking.
  • composer - Tuesday, June 28, 2005 - link

    I don't get it? A AMD64 3200 Winny clocks easy at 2.75 with no problems.

    Why would someone in the know buy such a proc if they can get the same performance for little money?

    Also, why can't they get past 3.0 ghz yet? (air cooling, or do we need to wait for .65 nano?
  • JustAnAverageGuy - Monday, June 27, 2005 - link

    Page 1:

    "But as we have mentioned time and time again, steadily increasing clock speed over time is a loosing proposition. "

    'losing'

    Those macro buttons are getting overused. It's the same text in every review, just different charts. :)
  • Sunbird - Monday, June 27, 2005 - link

    The FX-57 is still being called a FX-55 in the Doom3 graphs.
  • Cygni - Monday, June 27, 2005 - link

    Oh, in other news, the only thing i felt missing was Dual core results. Could have deffinitly been used.
  • Cygni - Monday, June 27, 2005 - link

    Its a speedbump review... i dont think AT, or anybody else, really gives a shit.
  • cryptonomicon - Monday, June 27, 2005 - link

    uh, they used the dfi ultra-d to overclock right?

    else, that is just retarded.
  • cryptonomicon - Monday, June 27, 2005 - link

    what a beast
  • saratoga - Monday, June 27, 2005 - link

    The 533 vs. 400 would make an interesting benchmark. Given the availability of relatively low latency 533 (or at least 500), it might be worthwhile for people who must have the best performance (most FX buyers I'd guess).
  • dougSF30 - Monday, June 27, 2005 - link

    > Who is being realistic here?

    Hello? As you said *in your own post*, anyone splashing out $1000 for an FX-57 can afford to buy top-end memory for it.

    Otherwise, what's the point? DDR-333 is really cheap these days, maybe they should test with that?
  • Viditor - Monday, June 27, 2005 - link

    I have a feeling that the mobo AT used was a poor choice...
    I have checked all of the other reviews I could find, and every one of them that overclocked was able to be stable at 3.0 GHz on air...
  • Goi - Monday, June 27, 2005 - link

    I would've liked a bit more analysis rather than just describing the tests and displaying the results. I notice that this has become more or less an AT template in reviews, with little analysis and a lot of data. I think a bit more analysis on interesting points would be helpful. For example, the 200MHz bump from the FX-55 to the FX-57 causes disproportionate performance increases in certain benchmarks. An detailed analysis on why(going from the 130nm hammer core to the 90nm Venice core with improved memory controller for example) would be helpful I think. There's just too much numbers and not enough analysis IMHO.
  • Viditor - Monday, June 27, 2005 - link

    blckgrffn - I understand your point, but this is a Dream Machine chip anyway, so we should definately see it at it's best (latest mobo, DDR533 LL, etc...).
    JMHO
  • miketheidiot - Monday, June 27, 2005 - link

    this new memory controller has me thinking ddr2....
  • blckgrffn - Monday, June 27, 2005 - link

    For those of you who want DDR500 etc included, I don't think that you understand how/why these articles are written. I can appreciate that you want the whole platform to be about the best performance, but then we would lose all sense of just how much better this CPU is compared to the old ones.

    So, LL DDR500 performs awesome. Great. I suppose if you are buying a $1000 processor you will probably drop $250 on a gig of ram. Super. But for the rest of us, getting ram that runs 2-2-2-10 is hard enough, let alone trying for that super duper ram that runs in how many mobos due to voltage requirements? One or two? Who is being realistic here?
  • suryad - Monday, June 27, 2005 - link

    Looks like if AMD had some sort of HT scheme like Intel did...there would be no benchmark where the Intel would be ahead...but I am most impressed that AMD has now taken over the domain of multimedia and encoding and so on from Intel...cant wait for 3 ghz multicore FX procs!
  • dougSF30 - Monday, June 27, 2005 - link

    Dear Lord! Low latency DDR-500 or DDR-533 will peform better than low latency DDR-400!
  • fishbits - Monday, June 27, 2005 - link

    "Why are you still doing Mozilla 1.4 testing?"

    Because this is a CPU review, and they already have a slew of other CPUs tested with Mozilla 1.4? Did you think they retested it on every chip each time they got a new chip in? This keeps it apples-to-apples so we can see the relative performance of the new contender against ones who've already been benchmarked.

    Or was there a specific score you needed to see for a specific version of Firefox to make or break your personal decision on whether to buy the FX-57 or not?

    AnandTech: Please continue to isolate the performance of the item being reviewed as much as practical. Last thing we need is extra hardware and software variables thrown in, until you're ready to move the whole gaggle over to a new set of yardsticks.
  • ravedave - Monday, June 27, 2005 - link

    Good straight forward article. Not much text though, expect your avg page view times to be tiny.

    What happened to all the suggestions people gave
    when Anand asked for tests for this article in his blog? Comon it's a speedbump, do something interesting.At least provide some slow old processor in the rankings so we can all laugh and point.

  • Backslider - Monday, June 27, 2005 - link

    With Half-Life 2 being the most CPU dependant game, I was suprised to see it missing from the benches.
  • acejj26 - Monday, June 27, 2005 - link

    Let me be the first to offer my services as an editor for articles here. I hate seeing articles here marred by poor grammar, spelling errors, and errors in the graphs.
  • Kocur - Monday, June 27, 2005 - link

    blckgrffn,

    Yes, you are right that DDR400 low latency will be better than DDR500 with very relaxed timings. However, you can buy now DDR memory with really nice timings at PC4000 speeds. See the memory tests at Anandtech and how much does FX53 gain from faster memory speeds at reasonable timings:). The same would hold for FX57 even to greater extent.

    Moreover, I think that FX57 should have been tested on a really good, mature platform, for example, DFI LP. You cannot use some crapy reference mobo until the socket 939 dies.

    Kocur.
  • Tallon - Monday, June 27, 2005 - link

    My god, my eyes are fucking bleeding.

    Losing: http://dictionary.reference.com/search?q=losing
    Loosing: http://dictionary.reference.com/search?q=loosing

    Please learn the difference.
  • blckgrffn - Monday, June 27, 2005 - link

    Dear lord! Do you know nothing of A64's and latency! DDR400 LL will stomp DDR500 @ relaxed timings, no problem! Further more, the baseline needs to stay the same, so they can't switch mobo and ram for every review.

    Next, mozilla 1.4 ~ Firefox. It is MOZILLA FIREFOX. Let's use our brain for that one, in that paragraph you said didn't make anysense, he laid it out for you.

    I do have one gripe - how can everything be slower at UT2k4 than Doom3? Tell me that was 1600*1200 w/aa&af! Otherwise, that benchmark should have much higher scores, imho....

    Other than that and that weird fluke where the 57 lost to the 55 and 4000+, thanks for the great article, Derek :)
  • Kocur - Monday, June 27, 2005 - link

    johnsonx,

    Well, it actually might be correct. Remember that they are using for this test the first old reference motherboard they got with FX55 last year. Thus, the results for other AMD processors come from October last year (well, at least for FX55). At that time they might have used another hard disk.

    Look also at the wierd results of P4 670 in some office tests with regard to other Intel processors. This is clearly disk/controller issue.

    Kocur.
  • DrMrLordX - Monday, June 27, 2005 - link

    Agreed #16, it's odd that the FX-55 and 4000+ win the Communication Sysmark 2004 bench. The FX-57 should have taken it easily.
  • johnsonx - Monday, June 27, 2005 - link

    It seems a bit odd that the FX-57 looses to the FX-55 and 4000+ in the third benchmark.... perhaps the labels are mixed up?
  • yacoub - Monday, June 27, 2005 - link

    Er, instead of "in the time a 2.8Ghz chip can", let's clarify and say "in a single clock cycle than a 2.8Ghz chip can".
  • manno - Monday, June 27, 2005 - link

    I heard it has support for DDR I 533, any chance we can see some benches with that stuff slapped in it?
  • yacoub - Monday, June 27, 2005 - link

    So is this graph an example of where pure clock speed determines performance?
    http://images.anandtech.com/graphs/amd%20athlon%20...

    And why is that? What about that test seems to be extremely tied to the pure "speed" of the chip? Is it a matter of a 3.8GHz chip being able to process 1,000,000,000 more instructions through it in the time a 2.8GHz chip can?
  • ultimatebob - Monday, June 27, 2005 - link

    Why are you still doing Mozilla 1.4 testing? Almost no one uses Mozilla anymore, they use Firefox.

    Besides... the explanation of why you're still using it (posted below) doesn't make much sense. You might want to reword it.

    "Quite possibly the most frequently used application on any desktop is the one we pay the least amount of attention to when it comes to performance. While a bit older than the core that is now used in Firefox, performance in Mozilla is worth looking at as many users are switching from IE to a much more capable browser on the PC - Firefox."
  • The DvD - Monday, June 27, 2005 - link

    1) What's up with the Communication Sysmark Business Application Performance bench?

    2) Why list the Gallating 3.46EE? It's not for sale..

    3) Good work on the review.
  • dougSF30 - Monday, June 27, 2005 - link

    Why didn't you guys test the DDR-533 memory divider support that comes with the FX-57? That would provide a nice boost in many scores.

    http://www.theregister.co.uk/2005/06/27/amd_fx-57/

    "AMD has upgraded the chip's integrated memory controller to work with 533MHz DDR SDRAM. Its predecessor, the FX-55, was limited to 400MHz memory."
  • phaxmohdem - Monday, June 27, 2005 - link

    Yes way too pricey. AMD's latest round of pricings has me questioning their business direction. WTF? What happened to undercutting intel and giving teh end user a sweet deal?
  • Starglider - Monday, June 27, 2005 - link

    It is surprising that the processor wasn't stable at 3GHz, as every other FX-57 review I've read where overclocking was attempted achieved 3GHz stable on air (and in one case 3.5 GHz stable on phase-change). As Kocur suggests, perhaps this was a motherboard issue?

    Agree with the conclusion though.
  • AndreasM - Monday, June 27, 2005 - link

    Loose, as in not tight.
    Lose, as in not win.

    Page 1
  • finbarqs - Monday, June 27, 2005 - link

    hrm.. is it based on the .09 micron process? Why is it clocked so poorly? is the San Diego core FX-55 better?
  • Kocur - Monday, June 27, 2005 - link


    Ouch, sorry for the double post:).

    Well, it is a pitty that some crapy reference motherborad from nowhere has been used for this review. Moreover, the settings (2, 2, 2, 10) were also not optimal for nForce4 chipset. To tell you the truth, I expected testing with at least DDR500, because it is how you would run this processor.

    Also I find quite funny that Anandtech got such a miserably overclocking processor. People on other sites reach 3GHz on air.

    What about power consumption comparison?

    I expected much more from Anandtech.

    Kocur.
  • Kocur - Monday, June 27, 2005 - link

  • projecteda - Monday, June 27, 2005 - link

    Doom 3 benchies have the FX-55 listed twice.
  • Marlin1975 - Monday, June 27, 2005 - link

    Need the X2 and P-D to REALLY see where this chip lines up as well, money wise to peroformance.
  • Rapsven - Monday, June 27, 2005 - link

    Waste of cash. Might as well just get a dual core.
  • RyzenRaisin - Monday, December 18, 2017 - link

    Oh if you guys could only have fathomed the power that Ryzen was to behold. Such silly humans of yesteryear.

Log in

Don't have an account? Sign up now