The Methodology

We've been asking AMD for months now to let us benchmark Barcelona and Phenom, and for months we've gotten the same answer: not yet. When asked why, AMD would always give us some terrible lie about how it was for competitive reasons, but when we actually put our hands on Barcelona at Computex we realized that these chips were simply not ready.

The Barcelona launch is finally upon us and we've got to wait another 45 - 60 days before we'll be able to bring you a review of Phenom, well, not exactly. Back when the Opteron launched, AMD was in a very similar situation to the one it's in today; AMD needed K8 to remain competitive, and it had been delayed so much that we were beginning to wonder if AMD would ever get the chip out on time. When the K8 finally launched, it was server-only but we took one of those server-only motherboards and ran a bunch of desktop tests on it to predict forthcoming performance.

We cracked open the Barcelona server and made some modifications; while the on-board ATI ES1000 graphics is sufficient for use as a server, it'd be too limiting for our desktop benchmarks. Luckily the Supermicro motherboard in the system had a plethora of PCIe slots, we just needed to gain access to them.


The PCIe Riser we removed from the system

We pulled out the PCIe riser card which plugs into the motherboard's sole x16 slot and divides it into a pair of x8s, then we modified a GeForce 8800 GTX by removing the backplate cover so we could just stick it into the open server.


The modded 8800


The 8800 GTX installed, the server is not really intended to be used like this

The end result was, as Johan put it, us using "such a beautiful, noble machine for such plebian activities". We couldn't help it, while AMD has already contacted us about Phenom briefings, we couldn't wait that long to get an idea of what we can expect from AMD on the desktop.


We needed an external PSU to power the graphics card, the server didn't have any PCIe power connectors

Barcelona is currently limited to DDR2-667, we were unsuccessful with attempts to run the memory any faster. Like all other MP Opterons, Barcelona requires the use of registered DDR2 memory, which is inherently slower than the unbuffered stuff we use on desktops. Because of these limitations we refrained from running any comparative benchmarks to desktop Athlon 64 X2s, instead we chose to run a single quad-core Opteron in our server platform against a pair of dual-core Opterons to simulate Phenom vs. K8 on the desktop.


The Opteron server, 2 CPUs, 8 DDR2-667 DIMMs

Keep this in mind as you're looking at these results, at best all we're offering is an idea of, at a minimum, how much faster Phenom will be over an identically clocked Athlon 64 X2. As Phenom is a more data hungry CPU than its predecessor, it will rely more on having a faster memory subsystem so the performance improvement could be even greater when we measure it on the desktop. That being said, at least we can set expectations within some amount of reason by performing this investigation.

The Servers K8 vs. Barcelona: The Phenom Preview
Comments Locked

70 Comments

View All Comments

  • Regs - Tuesday, September 11, 2007 - link

    Sorry, I did not mean to sound like a Dick Jack.

    The L3 cache is specifically designed with data sharing in mind. Which means all data unshared is evicted from the cache back into main memory. The L3 is also shared between four different cores and the access to the L3 must be arbitrated between them. This means the latency of the L3 actually depends on the core frequencies.

    Pre-fetches are made by the core and stored in the L1D cache of the Barc. They used to be stored in the L2 on the K8. So if the pre-fetch misses, it will look for information in the L2 and not the L3.
  • xenon74 - Monday, September 10, 2007 - link

    So if I understand this correctly:

    1.) You were trying to overclock registered DDR2 memory to give us a bit more even comparison between Athlon X2 and Phenom X4 using Opteron Barcelona QC and old Opteron DC.
    2.) OCing reg.DDR2 memory was unsuccessful which resulted that AMD sever stopped POSTing.

    3.) In two hours you come up with this "Phenom preview".

  • Anand Lal Shimpi - Monday, September 10, 2007 - link

    Not exactly :) The server not POSTing had nothing to do with trying to run faster DDR2 memory in the system, all of the benchmarking for the Phenom preview was done before the board stopped POSTing, I was simply trying to add more benchmarks to Johan's article when it ceased operation. Apparently others have been having issues with these motherboards and it appears that a new rev of the board may correct the problems.

    Take care,
    Anand
  • Reynod - Monday, September 10, 2007 - link

    Great first article Anand.

    As usual exceptionally well written and you avoided all of the fanboi nonsense.

    I'd like more detailed stuff when you get the chance please?


    Cheers :)
  • Regs - Monday, September 10, 2007 - link

    I guess only more clock speed, a better revision of the processor, and maybe a AM2-AM3 mobo will be a bit better for the Phenom?

    I wasn't expecting a whole lot out of AMD's barc. Though I do wonder what they have planned to help it for the retail desktop version. Like you said it just makes it more difficult to price it unless it can compete better.
  • Locutus465 - Monday, September 10, 2007 - link

    I've been waiting to see what AMD would do as a frist step on the comeback trail... This is a good opening salvo... True it's not going to be a Core 2 killer, but coupled with future technologies this is a great start! I hope they can afford to continue keeping pricing low while producing newer/better chips...
  • Etern205 - Monday, September 10, 2007 - link

    Hope this preview about Barcelona is as honest article where it tells
    the consumer the truth as to how well this cpu performs.
    Anandtech is all about honesty right?
    If that is true (which it always is), then this article better not be one
    of those that got approved by AMD before it was pubished!


    http://www.techarp.com/showarticle.aspx?artno=441">http://www.techarp.com/showarticle.aspx?artno=441
  • Locutus465 - Monday, September 10, 2007 - link

    Not sure why you're questioning anand's honesty, they've always been fair and accurate as far as I can see. And frankly, this artical fits with everything that's been said thus far about the new chip, it'll be much more competitive, but not to the point of winning the battle for AMD.
  • retrospooty - Monday, September 10, 2007 - link

    Agreed AT has always been fair and unbiased with reviews, that is why it is so popular.

    Even if there was an error in the test methodology, and someone points it out, they will update the article with the correction.
  • Etern205 - Monday, September 10, 2007 - link

    Anandtech is a great site and their article is always gives the
    best info, which is why when I'm about to get a new part like a cpu, or video card, I'll
    come to this site for reviews. And 100% of the time the results are always correct.

    But as for this "Barcelona"... mmm I don't know
    not just this site but to all the other sites as a whole.

Log in

Don't have an account? Sign up now