NVIDIA 6800Ultra Benchmarks

Discussion in 'Games' started by Converted2Truth, Oct 12, 2004.

  1. Converted2Truth macrumors 6502a

    Converted2Truth

    Joined:
    Feb 6, 2004
    Location:
    Hell@HighAltitude
    #1
    So guys, it appears someone has finally benchmarked their brand new 6800Ultra. www.xlr8yourmac.com would be the host to these. I'll just post his scores, and if you want to see his card in his computer, click HERE

    System: G5 2.5GHz Dual Apple G5 Tower Nvidia 6800 Ultra 1280x1024 OS X 10.3 2GB

    Quake3 Team Arena FPS 404

    Unreal Tournament 2004 flyby-asbestos FPS 207.77 / 63.39 / 497.71

    Unreal Tournament 2004 BOTMATCH-asbestos FPS 58.43 / 15.04 / 117.43
     
  2. applekid macrumors 68020

    Joined:
    Jul 3, 2003
    #2
    Ah, I know that mate. He posted some info in the Inside Mac Games forum in the hardware section. He'll be doing a benchmark for BareFeats soon, so let's hold out for that one.

    Halo doesn't sound impressive because the NV Shaders really need some updating. ATI's Shaders are nearly two times faster from what I remember.

    EDIT: Clicky for IMG thread.

    Is it true you can run either the ATI or NV Shader even if you don't have the corresponding graphics card brand?
     
  3. crazzyeddie macrumors 68030

    crazzyeddie

    Joined:
    Dec 7, 2002
    Location:
    Florida, USA
  4. invaLPsion macrumors 65816

    invaLPsion

    Joined:
    Jan 2, 2004
    Location:
    The Northlands
    #4
    Hmmm...

    My bet is that an ATi X series card will be made available for mac sometime soon at a cheaper price...

    I think I'll wait for that as the 6800 seems do be too much for it's precent of improvement...

    I'll have to think on it, though... :cool:
     
  5. jared_kipe macrumors 68030

    jared_kipe

    Joined:
    Dec 8, 2003
    Location:
    Seattle
  6. vraxtus macrumors 65816

    vraxtus

    Joined:
    Aug 4, 2004
    Location:
    San Francisco, CA
    #6
    The shader type really determines 2 things: quality and performance.

    nVidia shaders for Doom 3, for instance, give better quality images than with ATI cards. This is because the actual shaders are programmed with certain geometries that more closely read with Doom 3 coding.

    An example can be found in ATI's vs nVidia's handling of anisotropic filtering. While not a shader, the geometric calculations for how AF is handled in both drivers is drastically different, producing both different qualities and better performance, in some cases. From the last driver set that I read about, ATI had a significantly better AF driver than nVidia, which produced a better filtered texture image, compared to nVidia's. This is really the key point here for how programmed shaders for each card company can affect both performance and quality of 3D apps.
     
  7. a2daj macrumors member

    Joined:
    Sep 14, 2004
    #7
    In Halo, NV Shaders are the pixel shader effects being done with Register Combiners (NVIDIA's specific implementation for one of the PS 1.x versions). That's why GF3 and 4Tis can use them. The "ATI Shaders" is actually the ARB path. It was renamed such because of the addition of the NV Shaders with Mac Halo version 1.03. There's nothing ATI specific about it in terms of the shaders. The ARB/ATI path uses ARB Fragment Programs for the "shader" effects, so cards that can handle ARB Fragment Programs should technically be able to handle the ATI Shaders. I've heard that the FX 5200 owners had problems with the ARB shaders due to driver issues so that's probably why those cards can no longer use them.
     
  8. a2daj macrumors member

    Joined:
    Sep 14, 2004
    #8
    Heh, I always get a chuckle when I see something along the lines of "let's wait for the barefeats results!"

    They never run the tests I want to see even after asking nicely. They only run the tests they think 'real gamers' would use. I'm a gamer. I don't use the settings he uses. Don't get me wrong, he's a nice guy but his tests don't answer the questions I'd ask. Particularly the 9200 tests.

    When I do my benchmarking i run numerous tests across various screen resolutions. Depending on the game, I run with different options enabled and disabled. It's very time consuming but I like a large enough data set to be able to make thorough analysis and conclusions.
     
  9. Jigglelicious macrumors 6502

    Jigglelicious

    Joined:
    Apr 25, 2004
    Location:
    NYC
    #9
    Wow, I had no idea the scores for the UT botmatch would be so low. Even the lowly Radeon 9600XT is getting the same scores as the Geforce 6800Ultra! That means even the dual 2.5GHz processors can't push those video cards to their limits. Someone needs to raise the resolution and turn on AA and AF to see what those cards are really capable of.
     
  10. Mr. Anderson Moderator emeritus

    Mr. Anderson

    Joined:
    Nov 1, 2001
    Location:
    VA
    #10
    Excuse my ignorance, but is this card supposed to be a gaming card or is it used more for 3D apps?

    Thanks,

    D
     
  11. Converted2Truth thread starter macrumors 6502a

    Converted2Truth

    Joined:
    Feb 6, 2004
    Location:
    Hell@HighAltitude
    #11
    The G5 CAN push these cards to their limits. Remember that when you get the same reading from all the different vcards, then it's evident that the game bench is CPU bound. The G5 is fully capable of sending data to the 6800Ultra... all the compontents are there: SATA, DDR400, Fastest FSB available, etc...
     
  12. Converted2Truth thread starter macrumors 6502a

    Converted2Truth

    Joined:
    Feb 6, 2004
    Location:
    Hell@HighAltitude
    #12
    I really don't understand why people like you get a kick out of saying , Ah mate, i know that! Good for efff'n you! But there's plenty ppl out there who don't, and that's why i posted it. I also posted it minutes after he sent the info to Mike. So it's not like it matters that you know, because i posted it DAMN near the same time you (and I as well for that matter) read it on the insidemacgames forums.

    And i could care less what Rob-Art has to say about this, because he's not the one performing the tests. He's doing the vicarious act right now, which IMHO is the same damn thing as the stats i've posted.

    But to be polite, I am really happy that YOU KNOW. Good for you.
     
  13. Converted2Truth thread starter macrumors 6502a

    Converted2Truth

    Joined:
    Feb 6, 2004
    Location:
    Hell@HighAltitude
    #13
    It's the best card out for gaming right now. It gets the best frames per second anyway. But it turns out this is for the purpose of pushing 30" apple cinema displays. It's also the best card for Motion. I don't think apple intended it to be used for gaming... I still don't know what Steve meant when he huffed at it's introduction saying ,"yeah". I'd venture to say that this card is the best for everything right now... unless perhaps you have an ADC monitor, and you don't want to fork out 600 + ADC->DVI adappter cost.
     
  14. invaLPsion macrumors 65816

    invaLPsion

    Joined:
    Jan 2, 2004
    Location:
    The Northlands
    #14
    I seem to remember the same test on a 3.2GHz P4...

    The site used all of the top graphics cards and they all achieved around 53FPS.

    The "bot" test is very CPU bound...

    None of the tests up on Barefeats right now effectively show the power of the 6800 or the dual 2.5GHz Powermac G5...
     
  15. Jigglelicious macrumors 6502

    Jigglelicious

    Joined:
    Apr 25, 2004
    Location:
    NYC
    #15
    You are right - the bot match test is very CPU bound. But even so, there should STILL be some score differences between a 9600XT and a Geforce 6800Ultra, even if only slightly. I have checked the Anandtech review of the 6800Ultra, and it got 13fps more than a Radeon 9700Pro which is a good bit faster than the 9600XT. Those FPS results should not be so level, leading me to believe that the 2.5ghz G5 cannot push the cards to their max potential.
     
  16. applekid macrumors 68020

    Joined:
    Jul 3, 2003
    #16
    Bad day?
     
  17. Converted2Truth thread starter macrumors 6502a

    Converted2Truth

    Joined:
    Feb 6, 2004
    Location:
    Hell@HighAltitude
    #17
    lol

    Yeah, it was. Sorry about all that blah action
     

Share This Page