The games were run at stupid resolutions (my 2 19" 1280 LCDs will never see 640x480).
Come on, BaronMatrix! It's been said over and over again that the games were run at a low resolution so that the graphics card does not create a bottleneck; the test therefore stresses the CPU as opposed to the graphics card. Notice how once the resolutions are cranked up high, the frames per second for all the CPUs tested do not deviate much from each other? YES, that's the effect of the GPU bottlenecking the CPU! Yay, let's not use this argument ever again for putting down a CPU benchmarker, yeah??
SO what you're saying is that if the numbers aren't as great at 1280 it's the GPUs fault. GPUs don't become a bottleneck until 1920 or 2560 and things like SLI begin to stretch their wings at those resolutions.
Again Primordia I believe is CFD and
Structural dynamics soit's worth something to people. Engineers dont spend their days timing PI calculations. At least I hope not.
Don't tell me you didn't know that lowering the resolution will bring out the cpu true potential. Higher resolution means the gpu is working. It is limited by how fast the graphics card can handle the info. Dahh!
Do more research or I will ask Action Man to tell you to get a new keyboard.