Quake 4 Performance

Quake 4 gives us a good performance benchmark based on the Doom 3 engine. The game looks a bit better than Doom 3, but interestingly performs about the same. Without AA enabled, NVIDIA does a little better than ATI, especially in the higher resolutions. With 4x AA, NVIDIA is still a bit better in general, but the gap closes significantly, and the Crossfire setups come out ahead of the 7800 GTX 512 SLI.

Quake 4 - No AA

Quake 4 - 4X AA

Quake 4 - Maximum Quality



Splinter Cell: Chaos Theory Performance Image Quality, Feature Tests, and Power
Comments Locked

120 Comments

View All Comments

  • Live - Tuesday, January 24, 2006 - link

    Thanks for the explanation! Derek I think this merits a mention in the review.
  • NullSubroutine - Tuesday, January 24, 2006 - link

    perhaps a flash system where you can pick the card within the benchmark and it will show it on the line graph. just simply activate/deactivate feature.
  • bldckstark - Tuesday, January 24, 2006 - link

    I have to agree that a group color for the multi-GPU setups would be helpful on the bar graphs. The outline you used to denote negative gains would work well for this. Then ATI and Nvidia bars would still have a different major color, but the multi-GPU setups could have a yellow outline. E.G. ATI = red, ATI X-fire = Red w/ yellow outline, Nvidia = blue, Nvidia SLI = blue w/ yellow outline.
  • Rock Hydra - Tuesday, January 24, 2006 - link

    I don't know if you meant this or not, on the page mentioning the new crossfire board. There is url, I don't know if it was intended to be active or plain text, but I thought I would just bring that to your attention.
  • DerekWilson - Tuesday, January 24, 2006 - link

    thanks, fixed
  • emilyek - Tuesday, January 24, 2006 - link

    Good article.


    You have two typos in your article.

    In the system specs you have OZC Powerstreams instead of ...stream

    When you use the words 'eek out' as a verb that means 'squeeze out', it is spelled 'eke'-- 'eke out'.
  • DerekWilson - Tuesday, January 24, 2006 - link

    I had no idea there was a correct spelling for eke ...

    thanks
  • beggerking - Tuesday, January 24, 2006 - link

    Did anyone notice it? the breakdown graphs doesn't quite reflect the actual data..

    the breakdown is showing 1900xtx being much faster than 7800 512, but in the actual performance graph 1900xtx is sometimes outpaced by 7800 512..
  • DerekWilson - Tuesday, January 24, 2006 - link

    We didn't aggregate performance of each card under each game.

    for the percent improvment breakdown we only looked at 2048x1536 with 4xAA which clearly shows the x1900xtx in the lead.

    our reasoning is that this is the most stressful stock test we throw at the cards -- it shows what the cards can handle under the highest stress.
  • beggerking - Tuesday, January 24, 2006 - link

    umm.. what about 8xAA or higher? or lower resolution? w/wo AA?

    if you don't aggregate performance, then won't the graphic be misleading?

    isn't max quality the most stressful test ?

Log in

Don't have an account? Sign up now