Although the Quake 3 engine has been around for a number of years now, it is still being used in a number of today's games. Thanks to strategic and creative coding, developers have been able to create games with graphics which are well beyond the original game the engine was developed for. With a crop of games utilizing the engine in a multitude of different ways, we can benchmark the card and make an educated assumption of its strong points and weaknesses.

For each benchmark, every visual setting will be maximized. This includes running at a 32-bit color depth at all times. Given the power of the latest generation of cards, there seems to be little need to sacrifice image quality by selecting 16-bit color.

With Quake 3, I utilized the first map with a single opponent. Here, frame rates were sampled until five frags were recorded. Despite the insane frame rates registered with stock settings (no antialiasing or anisotropic filtering), the enabling of these features still managed to bring frame rates below 30 at the highest resolution.

In order to test Return to Castle Wolfenstein, I loaded the initial single player mission. Here, I would proceed through the castle until I crossed the bridge and knocked off the guard blocking the entrance to the next building. Turning on every feature uncovered graphics that were some of the most visually stunning scenes I have ever seen. For those who are not yet convinced of the benefits of antialiasing, this game will make you a believer!

As in the case of the two previous benchmarks, the initial level was used to test Jedi Knight II. In this case, I would proceed through the level until the first checkpoint was reached.

One aspect of this test, which I found to be extremely interesting, was the comparison between stock and overclocked speeds with maximum settings at 1024x768. Such an extreme variance in performance indicates that the graphics card is acting as the bottleneck. Fortunately, the XFX has headroom to spare and the card can be clocked high enough to break to 60fps barrier.

Breaking the pattern of using the initial level, I picked a specific level to test Solider of Fortune 2. Given the presence of a large amount of detail, I chose kam12.bsp to sample frame-rates. On its own, the level is very complex and should prove to be a good test for the system. However, given the presence of so many raw-edges and objects, the maximum-detail settings should really stress the system and show-off its true abilities.