As discussed many times on this site and others, you test games at the lowest resolution possible when probing a CPU's limits because it isolates the CPU as the limiting factor as opposed to the video card GPU limiting speeds.Originally Posted by K404
It's not about "real world" gaming frame rates in tests like these, the games are actually more like a synthetic benchmark for these purposes.
As far as the results being "much closer" at higher resolutions, that's correct, but not due to any sort of misplaced fanboy hopes regarding the Opteron holding a candle to Conroe. They're closer as the resolution climbs because the video cards have to endure more load. Eventually, say at 2560x1600, there is actually zero difference in the frame rate scores between any two CPUs...because the video card is maxed out and incapable of scaling any higher.
Bookmarks