Originally Posted by Rollo
Yes, like a "we've removed the GPU limitation from the benchmark so we could test the cpu" bottleneck?
If you crank up the res and aa you may be testing system performance under normal gaming conditions, but you've shifted the majority of the workload being tested to the GPU.
And? It's not like you're using different GPU's for each system. Real world is what matters. Sure, finding the CPU limitation may tell you what future games will do (arguable), but it misleads greatly as far as current games go. Sure, if under a more GPU limited situation all systems perform the same, that's somewhat useless information, although it should still be taken into account. If one system is slightly faster even at these high settings then this to me is far more important information than whatever system is faster than 1024x768 low settings.