UT2003 - Botmatch

Here we'll look at the Botmatch element of the UT2003 benchmark.




 640x480 800x600 1024x768 1280x960 1600x1200
Normal 25.3 25.2 24.3 22.2 18.4
3X FSAA 21.4 18.9 14.3 11.0 6.8
% Diff -15% -25% -41% -50% -63%

Unlike the Flyby benchmark, the Botmatch element of UT2003 is more inline with the results we've seen before, if a little on the low side even for a low-end board. Enabling 3XFSAA brings the performance down by as much as 63% in high resolutions.




 640x480 800x600 1024x768 1280x960 1600x1200
P4 2.53GHz 25.3 25.2 24.3 22.2 18.4
P4 1.8GHz 18.2 18.1 17.7 17.1 14.9
% Diff -28% -28% -27% -23% -19%

The performance difference on the two different CPU's again highlights the dependence on the CPU in the Botmatch benchmark, with this board. The 1.8GHz P4 looses 28% to the 2.53GHz P4 at 640x480, sinking to 19% deficit at 1600x1200.


 640x480 800x600 1024x768 1280x960 1600x1200
AGP4X 25.1 25.0 24.1 22.5 18.8
AGP8X 25.3 25.2 24.3 22.2 18.4
% Diff 0.6% 0.7% 0.6% -1.5% -2.1%

Here we can see that in the lower resolutions, where the CPU/system performance is tested more, hence a wider bus between the system to graphics board is more beneficial, we can see that AGP8X is indeed providing a small gain in performance. Oddly, though, this once again sinks to a slight deficit at the higher resolutions where the system is more limited by the fill-rate of the graphics board.