Ghost Recon: Advanced Warfighter
Since Ghost Recon doesn't have an embedded performance test or timedemo feature, we played through the same 90-second section of the game on each card with FRAPS collecting data in the background. This test was repeated five times on each card in an attempt to quell some of the inherent variability associated with playing through the game manually. Again, we've provided average and median low frame rates in addition to a timescale graph. Note that the points on this line graph represent averages, too.

Call this a beatdown, folks. The Inferno and X-Meridian are a good 11 frames per second off the X-Fi with the Athlon 64 X2 5000+ and even further off the pace with the X2 3800+. What's even more embarrassing for the Oxygen-based cards is how they're trounced by our lowly ADI 1988A codec.

More striking than the size of the frame rate gap is how it affects gameplay. Ghost Recon ran noticeably smoother on the X-Fi and the ADI 1988A than it did on the Inferno and X-Meridian.

Battlefield 2
Like GRAW, Battlefield 2 lacks in-game benchmarking features, so we resorted to another set of 90-second FRAPS sessions on the Strike at Karkand level.

The Oxygen-based Inferno and X-Meridian cards don't fare quite as poorly in Battlefield 2 as they did in Ghost Recon, but they consistently trail the X-Fi and the ADI 1988A. You probably won't notice the difference on the Athlon 64 X2 5000+, but the Inferno and X-Meridian definitely felt a little more sluggish on the X2 3800+.

Frame rates aside, we experienced a slight audio glitch with the Oxygen-based cards in Battlefield 2. In-game radio voices came across a little garbled and distorted, although that didn't affect our ability to play the game. This was the only positional 3D audio problem we experienced with any of the configurations throughout our testing.