UT2004 Benchmarking Guide + 6800GT Performance Results

[burningrave101]

nV News has a guide up showing how to accurately set up and benchmark UT2004.

http://www.nvnews.net/articles/benchmarking/ut2004/index.shtml

I'd like to see some performance results from you guys out there with 6800's and X800's. The X800's SHOULD be a little faster because they tend to fair better in UT2004 in the reviews i've seen.

You can see the performance of the cards in UT2004 here:

http://www.anandtech.com/video/showdoc.aspx?i=2113&p=13

And here are MY results in the rather intensive ONS_Dria demo using an XFX 6800GT stock and overclocked to Ultra speeds. This is on a fresh install of UT2004 without any tweaks. All visual settings in UT2004 are set to the absolute MAX.

I ran each test 3 times to confirm results.

6800GT 350/1000

6800GT 1600x1200 no AA no AF

UT2004 Build UT2004_Build_[2004-03-03_02.42]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3018 MHz with 1023MB RAM
NVIDIA GeForce 6800 GT (6172)

-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200

39.955928 fps rand[4118]

1600x1200 w/ 4xAA + 8xAF

UT2004 Build UT2004_Build_[2004-03-03_02.42]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3018 MHz with 1023MB RAM
NVIDIA GeForce 6800 GT (6172)

-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200

37.131844 fps rand[7862]


6800 Ultra 400/1100

6800u 1600x1200 no AA no AF

UT2004 Build UT2004_Build_[2004-03-03_02.42]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3018 MHz with 1023MB RAM
NVIDIA GeForce 6800 GT (6172)

-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200

39.692918 fps rand[16966]

6800u 1600x1200 w/ 4xAA + 8xAF

UT2004 Build UT2004_Build_[2004-03-03_02.42]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3018 MHz with 1023MB RAM
NVIDIA GeForce 6800 GT (6172)

-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200

37.379567 fps rand[24971]

6800u 1600x1200 w/ 4XAA + 8XAF w/ High Quality IQ

UT2004 Build UT2004_Build_[2004-03-03_02.42]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3018 MHz with 1023MB RAM
NVIDIA GeForce 6800 GT (6172)

-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200

33.875658 fps rand[27361]

I included a Full Trilinear test there at the end to show the performance drop going from Quality to High Quality. The X800's dont support High Quality.

 

[burningrave101]

Noone has any comments on my results or further benchmarks of their own in UT2004???

 

[FrgMstr]

No offense but this game is really not a challenge to any of the newest hardware out in real gameplay situations, this this is the reason we have moved away from it being in our reviews. Sort of like Quake3, once you are going so fast, going faster really does not matter.

On that note, I was playing Quake2 at 1600x1200 the other night. Nice game.

 

[ZenOps]

UT2004 is fairly intensive if you run "holy shit". Yeah, even the best high end cards have trouble hitting a 60fps average. Honestly though, I don't really think its the cards fault for being generally slow, its the engine that has be hack and slashed so much to add new visual enhancements that it was never designed for in the first place.

Quake3 is pretty easily getting 300fps average, so there definitely is no point in benching it anymore.

I've noticed in UT2004 that if you turn off just a few things, the peformance hike is usually worth it. For ATi, that means turning shadows to "blob" and projectors "off" ATi has always had weak speed in DX7 shadows and I think thats what UT2004 uses. ATi also has weak fog. Nvidia has weak lighting and sometimes turning off dynamic lighting gives a good performance boost.

 

[jaffy]

Just wanted to throw in some ATI scores. My 9800pro didnt do too bad. Everything is maxed (holy shit). BTW, I should be getting a 6800u soon

(1600x1200, no AA/AF)
UT2004 Build UT2004_Patch_[2004-06-11_16.51]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3593 MHz with 1023MB RAM
RADEON 9800 PRO (6458)
-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200
32.698811 fps rand[2804]

(1600x1200, 4XAA, 8XAF)
UT2004 Build UT2004_Patch_[2004-06-11_16.51]
Windows XP 5.1 (Build: 2600)
GenuineIntel Unknown processor @ 3593 MHz with 1023MB RAM
RADEON 9800 PRO (6458)
-exitafterdemo -benchmark -ini=UT2004.ini -exec=..\benchmark\stuff\demNS_dria.txt -nosound -d3d -1600x1200
21.849189 fps rand[23976]

 

[Thermite Paste]

i ran mine on a pIII 900, and a radeon 7000, with GREAT results i.e. 640 x 480 low everything and no options and still lag

 

[Brackle]

Isnt ut2k4 also a TWIMTBP game?...optimizations?.....nvidia?....hmmmm

DASHlT

P.S. Whats the point of showing us your benchmarks?...didn't Hardocp do a good review of the cards themselves?

 

[burningrave101]

Isnt ut2k4 also a TWIMTBP game?...optimizations?.....nvidia?....hmmmm

Ya thats why UT2004 is usually one of the games where the ATI cards perform better.

And ATI has optimizations out their ying yang so dont give me that lol.

P.S. Whats the point of showing us your benchmarks?...didn't Hardocp do a good review of the cards themselves?

Oh so now we dont need to post benchmark results if someone else has already reviewed the card? Whats the point in buying a $400-$500 video card and having a video card forum if your not going to post benchmark results. That makes no sense.