Games Benchmarks - Doom 3 (OpenGL)
Doom3 is the long awaited revamp of the original Doom title from id Software. Doom3 finally brings a fully shader enabled environment to OpenGL, utilising the newer functionality of OpenGL1.3 and above. Although the title makes use of elements made available in "DirectX 9 class" hardware, Doom3's primary rendering characteristic, being its unified lighting model, requires plenty of stencil fill-rate and a lighting shader that can be achieved in a single pass per light in DirectX8.1 equivalent hardware.
| Doom3 (FPS) | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X700 XT | 94.7 | 85.5 | 63.8 | 43.8 | 30.7 |
| X700 PRO | 92.7 | 74.6 | 55.7 | 37.9 | 26.9 |
| X800 XT | 96.4 | 94.1 | 92.4 | 77.0 | 59.3 |
| X600 XT | 70.5 | 50.4 | 34.2 | 22.8 | 15.5 |
| 9800 PRO | 93.3 | 71.9 | 52.6 | 35.4 | 24.6 |
| X700 XT % Faster Than: | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X700 PRO | 2.2% | 14.6% | 14.5% | 15.6% | 14.1% |
| X800 XT | -1.8% | -9.1% | -31.0% | -43.1% | -48.2% |
| X600 XT | 34.3% | 69.6% | 86.5% | 92.1% | 98.1% |
| 9800 PRO | 1.5% | 18.9% | 21.3% | 23.7% | 24.8% |
| X700 PRO % Faster Than: | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X800 XT | -3.8% | -20.7% | -39.7% | -50.8% | -54.6% |
| X600 XT | 31.5% | 48.0% | 62.9% | 66.2% | 73.5% |
| 9800 PRO | -0.6% | 3.8% | 5.9% | 7.1% | 9.3% |
With Doom 3 we can see that all the boards with 8 pipelines or less are getting more fill-rate limited than CPU limited at fairly low resolutions. In this instance both the X700's are able to outperform the 9800 PRO, indicating the the test is more limited by factors that effect fill-rate rather than pure bandwidth. Because of the fairly intensive nature of Doom 3 we can see how the X600 XT is beginning to be left behind by these new mainstream boards.
| Doom3, 4x FSAA + 8x AF (FPS) | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X700 XT | 65.2 | 50.4 | 38.9 | 27.0 | 15.4 |
| X700 PRO | 53.1 | 40.9 | 31.7 | 21.8 | 14.7 |
| X800 XT | 92.2 | 84.7 | 70.4 | 51.0 | 37.7 |
| X600 XT | 28.7 | 23.7 | 16.9 | 11.6 | 6.1 |
| 9800 PRO | 50.6 | 38.1 | 28.5 | 20.4 | 12.0 |
| X700 XT % Faster Than: | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X700 PRO | 22.8% | 23.2% | 22.7% | 23.9% | 4.8% |
| X800 XT | -29.3% | -40.5% | -44.7% | -47.1% | -59.2% |
| X600 XT | 127.2% | 112.7% | 130.2% | 132.8% | 152.5% |
| 9800 PRO | 28.9% | 32.3% | 36.5% | 32.4% | 28.3% |
| X700 PRO % Faster Than: | 640x480 | 800x600 | 1024x768 | 1280x1024 | 1600x1200 |
| X800 XT | -42.4% | -51.7% | -55.0% | -57.3% | -61.0% |
| X600 XT | 85.0% | 72.6% | 87.6% | 87.9% | 141.0% |
| 9800 PRO | 4.9% | 7.3% | 11.2% | 6.9% | 22.5% |
With 4x FSAA and 8x AF enabled the performance differences between the boards show similar trends to normal rendering, although obviously the pure performance is lower. As we saw with CS:S Video Stress Test, the fill-rate performance drop at 1600x1200 is fairly pronounced on the 128MB boards as the frame-buffer requirements of Doom3 is fairly high in the mode we are using here.
