Gaming Benchmarks

We had to bypass 3DMark Vantage for this review because it didn't want to play nicely with the HD3870X2 we used for this review. Honestly, I am not a fan of Vantage at this point as it has just caused headaches for me up until now. Needless to say, 3DMark 06 is still a very viable tool for testing a systems performance.

The results really speak for themselves as the increase in score is very significant. This is a result of both the processor tests but also the increased system performance allowing the HD3870X2 to be fully utilized. This may indicate a bottleneck in the video system that will show again in the last to gaming benchmarks we are about to look at.

Having very little to do with the actual game play, the Company of Heroes in-game benchmark is a very consistent way to test 3D performance of a system. All detail levels are maxed out and the resolution was set to the very popular 1680x1050. These settings are quite common but a hefty video card will be needed in order to not be the bottleneck at these settings.

At this point it is painfully obvious that the limitation of these results were the HD3870X2 and not the system. There is absolutely no gains to be had going from stock to the overclocked settings with detail levels turned up as high as they were. Obviously lowering details would present a phantom performance increase in the system but these settings are more than playable and there is no need to adjust the settings just to show a difference in results between system clocks. Adding a second HD3870X2, which is completely possible with the Rampage Formula, would result in a much different set of results showing up between the two setups.

The detail levels are again set to the maximum allowed and resolution to 1680x1050. This game runs flawlessly at these settings and many hours were spent playing Call of Duty 4 in this setup. For the benchmark, we have recorded a ten minute free for all match on the Crash level with fifteen users. There is plenty of heavy gun fight and action involved and fairly accurately portrays what type of results we would see on a regular basis playing multi-player COD4. The demo is then played back and FRAPS is used to record the frame rates shown below.

Again, we find out that at these detail levels, the system has absolutely no bearing on the results of the time demo. Lowering detail levels again creates a gap between the two sets of results but is completely un-necessary as COD4 runs very well at maximum details in this setup.