AMD Bulldozer / FX-8150 Desktop Performance Review @ [H]ardOCP

Discussion in 'AMD Processors' started by FrgMstr, Oct 11, 2011.

  1. xorbe

    xorbe [H]ardness Supreme

    Messages:
    6,003
    Joined:
    Sep 26, 2008
    Wowzers, the i7 920 idle power draw can't be right -- I have one @ 3.6 and idle is half that ... my OC idle is less than the chart's stock 920 draw, I'm pretty sure. Mine is an early rev too, not "d0". My 920 + 570 full load chew 540-580 at the wall ...
     
  2. tybert7

    tybert7 2[H]4U

    Messages:
    2,634
    Joined:
    Aug 23, 2007
  3. Buyers

    Buyers [H]Lite

    Messages:
    127
    Joined:
    Dec 1, 2005
    Well, i didnt expect it to really give the 2600K a run for it's money, but i hoped it would put up a better showing than what i've seen in these reviews. Well, now that the hype will be squashed (hopefully) i can make a solid informed decision on an upgrade.
     
  4. beto

    beto Gawd

    Messages:
    529
    Joined:
    Mar 28, 2009
  5. Rylan

    Rylan Gawd

    Messages:
    942
    Joined:
    Jul 23, 2011
    I wasn't in the market for an upgrade, because an i7 920 at 4.2GHz is no slouch. However, I was hoping for some competition from AMD. This is just disappointing unless AMD drops their prices by a good margin.
     
  6. s0urce

    s0urce n00b

    Messages:
    28
    Joined:
    Oct 31, 2004
    Do we have any info on performance improvements when overclocking the CPU-NB?

    Just curious because this yielded some pretty great performance improvements with the Phenom II's.
     
  7. JLAW904

    JLAW904 n00b

    Messages:
    23
    Joined:
    Jan 7, 2011
    And to think I almost upgraded from AM3 to an AM3+ mobo to prepare for bulldozer. Glad I waited. Sandy/Ivy Bridge... here I come.
     
  8. tw33kerloki

    tw33kerloki n00b

    Messages:
    14
    Joined:
    Jul 28, 2011
    argh... all that waiting for this?!?
    you just knew that the lack of leaked benchmarks before the launch were ill tidings.
    Still,... I was hoping BD would be like i7-2500k+ in performance and power efficiency.
    Just enough to get back into the game before Ivy Bridge came out.
    Now, AMD had better have a miracle or three hidden up their butts for use with Piledriver, otherwise it might be bleaker times ahead (not that the last few years, post 2006, has been all that sunny).

    Guess I can stretch my i7-950 (oc'd to 4.1GHz) out a little longer.
     
  9. FrgMstr

    FrgMstr Just Plain Mean Staff Member

    Messages:
    48,344
    Joined:
    May 18, 1997
    At least you did not have to work on it for a full week...
     
  10. malih

    malih Limp Gawd

    Messages:
    196
    Joined:
    Jan 2, 2009
    Maybe the naming should not be FX-8150
    It should be named Athlon II X8 150
     
  11. Trimlock

    Trimlock [H]ardForum Junkie

    Messages:
    15,157
    Joined:
    Sep 23, 2005
    Heh true that.

    I will not be picking this up. Sorry AMD.
     
  12. StewieTheGreat

    StewieTheGreat Gawd

    Messages:
    571
    Joined:
    Apr 12, 2010
    damn,this makes me sad,all the leaked benches were close to actual numbers,but I was really hoping for a somewhat competitive showing at least.

    I hope this thing is a folding monster.other wise I'm sitting on this Phenom II for a while longer. :(
     
  13. ShuttleLuv

    ShuttleLuv [H]ardness Supreme

    Messages:
    7,052
    Joined:
    Apr 12, 2003
    Bro, you fold with this thing even mildly overclocked and you're going to need a small reactor in your back yard it appears.
     
  14. wabbitseason

    wabbitseason [H]ard|Gawd

    Messages:
    1,511
    Joined:
    Jun 16, 2010
    Everyone should be heartbroken over this. Intel and AMD fanboys, and the reasonable folks who just buy what gives the best performance at their price range.

    This means intel has no competition. This means intel can continue to socket swap and charge us $250 for a new motherboard nearly every tick-tock. This means we will be paying outrageous prices for ivy bridge (I won't if they're as outrageous as I predict).

    Truly a terrible blow to the entire computing industry.
     
  15. dr.kevin

    dr.kevin 2[H]4U

    Messages:
    3,053
    Joined:
    Feb 17, 2006
    how could you possibly screw that up....

    the k10 core does more work at a slower clock.

    [​IMG]
     
  16. n3zyd

    n3zyd I Drink & Make TITAN X SLI Decisions

    Messages:
    346
    Joined:
    Aug 9, 2008
    and hope some cores unlock too, Like the x3 to x4 unlock.
    99 bucks that I "Might" turn into 6 or 8 core cpu, I could see myself buying that cpu/mobo
     
  17. defaultluser

    defaultluser [H]ardForum Junkie

    Messages:
    12,554
    Joined:
    Jan 14, 2006
    Thanks for the awesome review Kyle!

    And thanks again for including a Thuban. It's pretty embarrassing when a $200 new architecture gets soundly beaten by a $170 older chip :D

    THE SADDEST PART OF THIS ALL: Thuban not-only has the same performance as Bulldozer, it also has the SAME POWER CONSUMPTION at 45nm! What the hell have AMD engineers been doing?

    Here's to hoping AMD has a contingency plan with a Thuban shrink to 32nm with two cores tacked-on :D
     
  18. ShuttleLuv

    ShuttleLuv [H]ardness Supreme

    Messages:
    7,052
    Joined:
    Apr 12, 2003
    Looks like it's a higher clocking Thuban then, with worse single core ipc and more cores and more power consumption. What a cluster fuck.
     
  19. kinjo

    kinjo [H]ard|Gawd

    Messages:
    1,053
    Joined:
    Dec 17, 2010
    They could but they wont because they need AMD to exist to fend off any anti trust allegations.
     
  20. fatrat

    fatrat [H]ardness Supreme

    Messages:
    6,838
    Joined:
    Aug 31, 2000
    Now there's a use for this processor....home heating. :D
     
  21. techie81

    techie81 [H]ard for [H]ardware

    Messages:
    4,959
    Joined:
    Jan 12, 2005
    What a joke. Really, why release it?
     
  22. Forceman

    Forceman [H]ardForum Junkie

    Messages:
    9,243
    Joined:
    Aug 7, 2005
    I wonder if JF-AMD will update his Bulldozer pre-release FAQ? His frustration appears...misplaced, especially since pretty much all the leaked benchmarks were spot-on. His weaseling on the IPC issue makes a lot of sense now, though.
     
  23. DarkStryke

    DarkStryke [H]ard|Gawd

    Messages:
    1,391
    Joined:
    Sep 19, 2001
    The architecture makes for a good server chip, which is what his area is. That's really all they have to be proud about with this release.
     
  24. entropy13

    entropy13 Gawd

    Messages:
    922
    Joined:
    Jul 16, 2009
    Am I allowed to post something like this here?

    http://hardocp.com/article/2011/10/11/amd_bulldozer_fx8150_desktop_performance_review/10


    http://pcper.com/reviews/Processors...ulldozer-Unearth-AMD-Victory/Closing-Thoughts


    http://guru3d.com/article/amd-fx-8150-processor-review/21



    http://www.legitreviews.com/article/1741/20/



    http://hothardware.com/Reviews/AMD-FX8150-8Core-Processor-Review-Bulldozer-Has-Landed/?page=11


    http://hexus.net/tech/reviews/cpu/32110-amd-bulldozer-fx-8150/?page=8


    http://www.techspot.com/review/452-amd-bulldozer-fx-cpus/page13.html



    http://techreport.com/articles.x/21813/19


    http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/11



    http://www.tomshardware.com/reviews/fx-8150-zambezi-bulldozer-990fx,3043-24.html


    http://www.legionhardware.com/articles_pages/amd_fx_8150fx_8120fx_6100_and_fx_4170,9.html
     
  25. tybert7

    tybert7 2[H]4U

    Messages:
    2,634
    Joined:
    Aug 23, 2007
    It almost makes you think something is not quite right. This is the 8 core model right? but it's really only 4 bulldozer modules with each module having some sort of hybrid core that can both "pitch in" for single threaded tasks...

    I probably just butchered about 10 different things from memory on the mechanics of the new chip... but I wonder if the chip is just backwards or software is not taking full advantage... It's still a bit too unbelievable that amd would intentionally design a chip with less performance per core than a previous generation chip... something is off here.
     
  26. ninehunid

    ninehunid [H]Lite

    Messages:
    126
    Joined:
    Dec 31, 2010
    i just looked up an i3 -2100 and i think its single core score was 1.24
     
  27. sirmo

    sirmo Gawd

    Messages:
    516
    Joined:
    Sep 23, 2009
    While not delivering on the legendary performance, the FX name suggests, may not be marketing dept's fault. It is their fault for giving it that name. After all it is far easier to pick a name than to build a processor architecture from scratch that competes.

    FX is simply misleading when you see that their new product doesn't even live up to their own last gen of CPUs, which weren't that great to begin with. Everything about Bulldozer is disappointing, there isn't a single positive I can say about it. And I am an AMD fanboy.

    Kyle was very kind to AMD in his closing statements. He could have shredded them to pieces and no-one would have complained.

    These are my closing statements:

    AMD FX-8120 cost $30 more, but only comes close to 2500k in multithreaded benchmarks, but a high quality motherboard and PSU it takes to feed the Bulldozer CPU will cost you over $100 more. Also not counting your power bill increase from additional power draw and air conditioning.
     
    Last edited: Oct 12, 2011
  28. dr.kevin

    dr.kevin 2[H]4U

    Messages:
    3,053
    Joined:
    Feb 17, 2006
    That 8ghz OC had 6 cores disabled.

    but it probably used the same 500 watts just for those 2 cores :D
     
  29. ninehunid

    ninehunid [H]Lite

    Messages:
    126
    Joined:
    Dec 31, 2010
    I'm also looking forward( :( ) to seeing how fx 4100 stacks up against i3 2100
     
  30. Lord_Exodia

    Lord_Exodia [H]ardness Supreme

    Messages:
    6,997
    Joined:
    Sep 29, 2005
    Not what I was hoping for but definitely what I expected. Yet another dismal launch from AMD. Looking at several reviews it seems to be slower, more power hungry and dependent on multi threaded optimizations to shine. At least it's cheaper. Still I'm gonna have to say a disappointing launch again by AMD. Sigh, intel really needs more competition.

    BTW sell your AMD stock quick!!
     
  31. cannondale06

    cannondale06 [H]ardForum Junkie

    Messages:
    16,180
    Joined:
    Nov 27, 2007
  32. sirmo

    sirmo Gawd

    Messages:
    516
    Joined:
    Sep 23, 2009
  33. dr.kevin

    dr.kevin 2[H]4U

    Messages:
    3,053
    Joined:
    Feb 17, 2006
    well, what would you do if you were pushed into a corner.... You have to come up with some bullshit excuse in a hurry.
     
  34. defaultluser

    defaultluser [H]ardForum Junkie

    Messages:
    12,554
    Joined:
    Jan 14, 2006
    For those of you worried that Intel will raise price because they have no competition, I leave you this thought:

    Windows 8 supports ARM. ARM (and licensees with their own architectures) know how to build efficient chips for very low cost. ARM WILL keep Intel on their toes, and WILL force Intel to price their chips reasonably.

    So-long AMD CPUs, we won't miss you :D
     
  35. Lord_Exodia

    Lord_Exodia [H]ardness Supreme

    Messages:
    6,997
    Joined:
    Sep 29, 2005
  36. spacing guild

    spacing guild [H]ard|Gawd

    Messages:
    1,357
    Joined:
    Dec 9, 2010
    Reminds me of the nVidia FX series. Pure suckage.
     
  37. sirmo

    sirmo Gawd

    Messages:
    516
    Joined:
    Sep 23, 2009
    ARM chips are nowhere near the efficiency of x86. I know that sounds strange because when someone says ARM, they are talking about power efficient chips. But hear me out.

    Per performance, current gen x86 chips are years ahead of Arm designs. Basically the point is when Arm gets to the performance levels of x86 it loses all it's efficiency edge. There was a really interesting article about this but I can't find it, oh well, this one is not bad either:

    http://arstechnica.com/business/news/2011/02/nvidia-30-and-the-riscification-of-x86.ars/2
     
  38. zero2dash

    zero2dash [H]ardness Supreme

    Messages:
    6,081
    Joined:
    Oct 23, 2007
    Intel's generally been a "better performance at a higher cost" sort of company (IMHO); AMD was "cheap but respectable" (again IMHO). That was one thing that kept me at least interested in AMD over the years despite Intel constantly whooping up on them.

    Now....you can't even say that's the case anymore either. Rough generalization over the prices, no matter where you go, what price point you're at....there's really no reason not to get Intel, even if you're going bottom of the barrel low end and comparing an i3 2100 vs a Phenom II or (now, presumably) an FX-4100.

    I want AMD to stick around, I'll keep rooting for them to come up with something and knock Intel on their heels....but they're still not doing it. :eek:
     
  39. tybert7

    tybert7 2[H]4U

    Messages:
    2,634
    Joined:
    Aug 23, 2007
    Anand has a good write up on some of the details on the software side, it seems very little takes advantage of AMD's "FMA" instructions...

    http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/2



    I'd like to see how this performs with software optimized for both intel and amd platforms, that allows all the tricks the cpu is capable of to be exploited, but maybe cinebench 11.5 already does that?

    need someone more techy to answer.