Quake 3: Arena is one of our primary gaming benchmarks here at SE, due to its highly repeatable results and de facto standard as a game benchmark. Quake 3 testing is performed using High Quality detail and 1024x768 resolution and the program has also been updated to release 1.30, along with a natural progression to the standard "demo Four".
In another big non-surprise, the EPoX 8K3A+ finished first in Quake 3 testing, followed by MSI. The AOpen AK77-333 surprisingly pulled into third, followed by ASUS and ABIT. There is a definite pattern starting to form in the performance benchmarking end, as both EPoX and MSI seem to be holding down the top two spots the majority of the time. Quake 3 tends to make use of every ounce of system power, not to mention it scales very well, and we see an almost-20 fps difference between the top and bottom boards
Return to Castle Wolfenstein is another Quake-based game, but with some notable differences. The basic game engine may be the same, but the graphics, gameplay and stress it puts on a processor are very different. Until the next Quake game appears, RtCW is the next best way to determine high-end Quake engine performance. We have used the Checkpoint MP demo using the default High Quality detail settings and have upped the resolution to 1024x768.
This trend is becoming a veritable pattern, as once again EPoX and MSI take the first two spots and are followed by ASUS and AOpen. The actual framerate differences are smaller than under Quake 3, but we're still seeing a healthy 10 fps differential between the top and bottom performers.
The Comanche 4 benchmark from Novalogic gives us an opportunity to use an actual flight sim for performance testing. Flight sims are notorious for their CPU-dependence, and this makes the Comanche 4 benchmark potentially a better CPU test than it is for 3D video cards. The reliance on the CPU shows itself off in the benchmark, and even the slightest difference in framerates could pay off in significantly enhanced game framerates. For our comparison, all testing has been performed at 1024x768, 32-bit with audio disabled.
Even though the actual performance differences are quite small with Comanche 4, we're once again seeing the exact same pattern as other game benchmarks.
