Gaming performance
We've purposely benchmarked the gaming performance using an NVIDIA GeForce 8800 GTX and a modest 1,024x768 resolution. The premise here is to see how the subsystem throughput - and the processors' in particular - affects performance. Real-life gaming would be reflected by a higher resolution that includes image-quality enhancements. Going down that route would put the onus on the graphics card, however.
No real benefits for the new architecture in Company Of Heroes. AMD Athlon 64 X2 6400+ performance (over three runs) is strange, to say the least, and we attribute the below-par benchmark result to problems associated with the motherboard correctly applying AMD's Cool 'n' Quiet technology for this particular processor.
Extra and smarter cache, in the main, is, we reckon, the driving factor behind the QX9650 handily beating out the 65nm QX6850 in the low-resolution Quake 4 benchmark. Again, we note that the AMD Athlon 64 X2 6400+ and 6000+'s results appear to be erroneous. Our AMD testing platform died shortly afterwards and we couldn't re-run numbers in the limited time available. We also note that AMD sampled us with the '6400+ processor at the very last minute, holding up testing. Extrapolating from the '5600+ results we reckon that the 6400+ would still benchmark at the sub-200fps level.
Penryn > Kentsfield > Conroe > K8, evidently.
