Forgot your password?
typodupeerror
AMD Hardware

Bulldozer Server Benchmarks Not Promising 235

Posted by Unknown Lamer
from the cool-kids-jump-ship-to-fleet dept.
New submitter RobinEggs writes "Some reviews of Bulldozer's server performance have arrived. Ars Technica has the breakdown, and the results are pretty ugly. Apparently Bulldozer fares just as poorly with servers as with desktops. From the article: 'One reason for the underwhelming performance on the desktop is that the Bulldozer architecture emphasizes multithreaded performance over single-threaded performance. For desktop applications, where single-threaded performance is still king, this is a problem. Server workloads, in contrast, typically have to handle multiple users, network connections, and virtual machines concurrently. This makes them a much better fit for processors that support lots of concurrent threads. ... It looks as though the decisions that hurt Bulldozer on the desktop continue to hurt it in the server room. Although the server benchmarks don't show the same regressions as were found on the desktop, they do little to justify the design of the new architecture.' It's probably much too early to start editorializing about the end of AMD, or even to say with certainty that Bulldozer has failed, but my untrained eye can't yet see any possible silver lining in these new processors."
This discussion has been archived. No new comments can be posted.

Bulldozer Server Benchmarks Not Promising

Comments Filter:
  • by Chrisq (894406) on Tuesday November 22, 2011 @09:27AM (#38134582)

    I thought we all switched to full-fat floating-point operations over 15 years ago when the Pentium hit the mainstream and everyone finally had an on-die FPU in their PC

    Its application dependent. I doubt if much fp stuff gets done in cryptography, routing, and many simulations.

  • Virtualization (Score:3, Interesting)

    by Anonymous Coward on Tuesday November 22, 2011 @09:29AM (#38134588)

    When someone says that a CPU was designed around multiple threads I think virtualization. yeah you can argue that servers are multithreaded in that they have to handle multiple users connecting, but that's bull. I can write a badly threaded application that doesn't effectively use the multiple cores...

    So how do these cpus perform with something like ESX running on them?

    Scott

  • Great for BOINC! (Score:4, Interesting)

    by courteaudotbiz (1191083) on Tuesday November 22, 2011 @09:31AM (#38134594) Homepage
    That's perfect for running BOINC though, which is very good at using multiple cores at their full capacity. Useless for the business, but great for contributing to science projects :-)
  • by confused one (671304) on Tuesday November 22, 2011 @09:53AM (#38134760)
    Windows does not (yet) know how to properly schedule threads on that hardware. This has caused issues with all the benchmarks, not unlike what happened when Intel Hyperthreading was first released. Once the proper support is added to the OS kernels, the results should be much better.
  • by serviscope_minor (664417) on Tuesday November 22, 2011 @10:34AM (#38135230) Journal

    Supercomputers are NOT built based on processor speed.

    Um.

    That's rather an oversimplification, to the point of being wrong.

    Supercomputers need good interconnects and lots of processing power. One or the other alone won't do.

    Much of the $$$ goes into interconnects, but also the CPUs and the cooling, which is very dependent on the CPUS. All things considered, neither AMD nor Intel have the fast interconnects on-die (unlike Fijutsu), so pretty much the main thing to choose between the CPUs is, well, the CPUs.

    And it seems like AMD are the best option at the moment for this kind of workload.

    The rest of us will use Xeons and be very happy with the results.

    No, you will. I'll stick with my Supermicro quad 6100s for as long as I can and be very happy with the immense price/performance they offered.

  • by confused one (671304) on Tuesday November 22, 2011 @10:44AM (#38135352)

    Tech Report demonstrated this to be the case by setting the thread affinity on their tests, so they were locked to specific cores, using only once core per module. They saw as much as a 30% improvement in the single threaded or lightly threaded benchmarks. Other sources, including AMD itself, have demonstrated as much as 10% improvement in performance by using a better thread scheduler. AMD has whitepapers discussing this issue.

    As for changing the OS kernels... Windows 8 already has the changes. Windows 7 and Server 2008 may get them in a future update (Service Pack?). Linux kernel support is ready and is available in a kernel patch. Compiler support is now included in VS 2010. So, not necessarily a flop; but, might be a short while before the full capability of the architecture is realized.

  • Re:And moreover (Score:4, Interesting)

    by TheLink (130905) on Tuesday November 22, 2011 @11:37AM (#38136080) Journal
    Too many? I don't think so. And please stop trying to convince the AMD fanboys that AMD is producing crap.

    Why?
    1) We need AMD alive and kicking to at least give Intel some competition (look at what has happened now that AMD is weak - Intel started having "unlock codes" to unlock more performance/features for their processors ).
    2) So someone needs to buy the current batch of AMD crap[1] to keep AMD alive till they come up with something better.
    3) I'd rather not buy AMD's current crap. It is inferior for most popular desktop and server tasks.
    4) Therefore we need as many AMD fanboys as possible to continue thinking that AMD is great and buying lots of AMD crap.

    [1] Yes I know AMD produced better stuff than Intel some years ago. However the latest CPUs ironically appear to be AMD's Prescott Edition CPUs.
  • by UnknownSoldier (67820) on Tuesday November 22, 2011 @03:31PM (#38139840)

    > I'd only clicked on a couple of articles since Jon Stokes left,

    Agreed! Jon "Hannibul" Stokes articles were extremely well-written; explaining the latest esoteric hardware in [almost] layman's terms.
    It's too bad he left Ars Technica -- the site jumped the shark a while ago and is everything is dumbed down. At least AnandTech is still [relatively] OK.
    http://arstechnica.com/staff/palatine/2011/07/send-off-jon-hannibal-stokes-marches-his-elephant-army-away-from-ars.ars [arstechnica.com]

    Sad that none of the links work ... for his "Classic" Essays ;-(
    http://arstechnica.com/paedia/cpu.html [arstechnica.com]

Help stamp out Mickey-Mouse computer interfaces -- Menus are for Restaurants!

Working...