New GPU Benchmarks for OS X!

Discussion in 'Mac Pro' started by MacVidCards, Nov 23, 2012.

  1. macrumors 603

    Joined:
    Nov 17, 2008
    Location:
    Hollywood, CA
  2. macrumors 6502

    Joined:
    Jan 6, 2010
    Location:
    ::1
    #2
    Good or Bad?

    Mvc gtx 570 2.5gb
     

    Attached Files:

  3. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #3
    5850, 6870 and GTX 570 in MP 1,1 and 3,1.
    Run these tests in similar to their reference resolution and without AA. We'll get more comparable results.
     

    Attached Files:

  4. markupstart, Dec 8, 2012
    Last edited: Dec 8, 2012

    macrumors newbie

    Joined:
    Jun 5, 2009
    Location:
    El Paso, TX
    #4
    Gtx 660, MACPRO 5,1
     

    Attached Files:

  5. macrumors regular

    PunkNugget

    Joined:
    Apr 6, 2010
    #5
    Is this a good score off of my 480 GTX? I'm going to try the 580 GTX next...
     

    Attached Files:

  6. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #6
    Hard to tell when your'e running it @1680x1050. Resolution does matter a lot.
     
  7. macrumors regular

    PunkNugget

    Joined:
    Apr 6, 2010
    #7
    Well on that one Hackintosh that I have that is (right now) the best monitor I tested it on. I will try the Hackinbeast right now that has my 580 GTX on it. Also, I'll try it with 1920 x 1080 and then try 2560 x 1600 (since I have a Dell 30" monitor)... ;)

    ----------

    Well, I just tried it and I think I got a pretty good score. I'll post it in a moment, but I'm going to hold off on testing in full res mode (2560 x 1600) as it reached 87C !!! :eek: Once I put on my EK waterblock then I'll test it again... :cool:
     
  8. macrumors regular

    PunkNugget

    Joined:
    Apr 6, 2010
    #8
    Here it is...
     

    Attached Files:

  9. thread starter macrumors 603

    Joined:
    Nov 17, 2008
    Location:
    Hollywood, CA
    #9
    I'm not sure if you got threads mixed up.

    Cinebench is a CPU test and this thread is about some specific new GPU benchmark. Could you please remove these posts? Or just run the linked test and post those results as well.

    If you want to start another thread using the tests you like, nobody would mind that either.
     
  10. macrumors regular

    PunkNugget

    Joined:
    Apr 6, 2010
    #10
    I don't think he got it mixed up, he just included all of them just to show some extra "flare" - LOL !!!, including GPU Test GUI.
     
  11. macrumors regular

    PunkNugget

    Joined:
    Apr 6, 2010
    #11
    BTW, here are both scores off the same machine at both 1680 x 1050 and 1920 x 1080. So there is a minor difference but not a BIG one... ;)
     

    Attached Files:

  12. Tutor, Dec 8, 2012
    Last edited: Dec 8, 2012

    macrumors 65816

    Tutor

    Joined:
    Jun 25, 2009
    Location:
    Home of the Birmingham Civil Rights Institute
    #12
    I'm not aware that I got them mixed up.
    True.
    Done it.
    That's what I had done, i.e., I ran all three linked tests, posted the test results of all three of them here, but I had also included in the same post with those three test results, other system performance results and information on the system as a whole because other system tweaks affect CUDA performance.
    Done it and cleaned up my posts here. Good Bye.
     
  13. macrumors 65816

    MIDI_EVIL

    Joined:
    Jan 23, 2006
    Location:
    UK
    #13
    Here's the Gigabyte NVidia GeForce GTX 670 w/ 2GB RAM
     

    Attached Files:

  14. macrumors newbie

    Joined:
    Dec 19, 2012
    #14
    I'm such a newb I dont even know how to post properly! lol
     

    Attached Files:

  15. macrumors newbie

    Joined:
    Nov 24, 2011
    #15
    Here is the GTX 670 FTW that I just installed tonight.
     

    Attached Files:

  16. macrumors newbie

    Joined:
    Dec 19, 2012
    #16
    Just got the Asus PB 278 monitor. When I set the test drop down to 1920x1200 it comes out 2560x1440 which is max res on the Asus. 1920x1200 was the max res on my previous Dell so I guess it just defaults to max?
     

    Attached Files:

  17. macrumors 601

    derbothaus

    Joined:
    Jul 17, 2010
    #17
    The 670's are all over the place.
     
  18. OrangeSVTguy, Dec 22, 2012
    Last edited: Dec 24, 2012

    macrumors 601

    OrangeSVTguy

    Joined:
    Sep 16, 2007
    Location:
    Northeastern Ohio
    #18
    Here's my scores with a palit sonic platinum 570 1280MB. Bummer my plans on going 680 SLI(windows) was thwarted :(
     

    Attached Files:

  19. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #19
    ^^^Orange, try with AA off, as most of users and original testers.
     
  20. macrumors 65816

    Tutor

    Joined:
    Jun 25, 2009
    Location:
    Home of the Birmingham Civil Rights Institute
    #20
    Especially over, what is for me a joyous holiday season {especially since the world as we know it didn't end on 12/12/12 or 12/21/12}, I 'll be devoting my free moments to studying how to get the optimum level of performance from the GPU in many differing configurations. For over 58 years, i.e., since I began to speak, I have been a child of the "Why?" and it almost drove my parents and teachers loony. During college football season, different college bands pump out different cadences that drive their marching bands to evoke different passions in the crowds. That synergy is why I love to watch FAMU (Florida A&M University)'s marching band more than most others. I believe that your observation here arises from those synergies. A GPU isn't the band, but rather is part of a larger unit in which other system components can add and/or detract from the performance, e.g., the numbers we witness being output by the "New GPU Benchmark" app. Different 670s, just as is the case for every other model/generation, may have differing memory sizes and different memory and core speeds connected to differing busses (and potentially tweaked ones) affected by differing configurations, etc. that affect the benchmark outcome. That is why I believe that it is important for the testers to give as much information as is possible about the total system in which their GPU card sits so that I, as well as any other individual who cares about the "Why?", can get as full of an understanding as is possible (and hopefully figure out why and let the others of us know) about why that 670 out performs another, or what makes that 580 "better" at a particular benchmark than another 580. Knowing the "why" is essential to replication and adding to the body of knowledge that I believe some of us seek to tap. Since MacVidCards has indicated to me that he does not intend for this thread is to be a repository of that kind of knowledge, anyone who cares to learn more about why the scores are what they are or how to score higher or what may be holding back their GPU's performance may, after posting their "New GPU Benchmark" results here, bounce to my thread, "All We Know About Maximizing CPU Related Performance" [ http://forums.macrumors.com/showthread.php?t=1333421 ] {beginning at post #435} where I will provide updates focusing on the GPU/system synergies, and you may post more system-related information with various GPU and related CPU benchmark scores and/or questions/comments to broaden this body of knowledge.
     
  21. macrumors regular

    Joined:
    Feb 14, 2008
    #21
    [​IMG] [​IMG]
     
  22. macrumors 68040

    666sheep

    Joined:
    Dec 7, 2009
    Location:
    Poland
    #22
    Your tests results without AA confirmed that this test isn't CPU bound. My 570 scored (almost) exactly that same using that same settings on different machine and OS (i.e. different drivers revision). My card was Gainward GTX 570 1280MB.
     
  23. macrumors 6502

    sarthak

    Joined:
    Nov 19, 2012
    Location:
    Canada
    #23
    Zotac GTX670 REFERENCE 2GB Benchmarks:

    [​IMG]
    [​IMG]
     
  24. macrumors 601

    OrangeSVTguy

    Joined:
    Sep 16, 2007
    Location:
    Northeastern Ohio
    #24
    Cool. I know the Sonic 570 is slightly overclocked from the factory and a lot of reviews has it almost the same benchmarks as a 580. It also has 8pin and 6pin for power unlock the other 570s with 2 6pin.
     
  25. macrumors 601

    OrangeSVTguy

    Joined:
    Sep 16, 2007
    Location:
    Northeastern Ohio
    #25
    Well can't wait to test this setup in windows. I'll do a single bench test in OSX since the 670s are all over the place.

    [​IMG]
     

Share This Page