NVIDIA Says Built-In Crysis GPU Benchmark Not Good To Use

[Apoptosis]

With the upcoming launch of the GeForce 9800 GX2 video card in the near future it seems NVIDIA is a bit worries about how sites are testing with the game Crysis. It seems now that NVIDIA has determined what many of us already knew... The built-in benchmark in Crysis is not a good measure of GPU performance. Go figure right? This is something that we spotted back in November and just to let you our readers and community know all of our Crysis numbers since day one were gotten by manually running FRAPS and not using the built-in game demo that Crysis has.

Here is part of the E-Mail that one of the NVIDIA engineers sent Legit Reviews and all other review sites.

I've been testing Crysis quite extensively over the past few weeks. It turns out that the benchmark_GPU isn't quite good at measuring graphics due the speed of the camera moving across the map. It causes a lot of data streaming and ends up reducing SLI performance by 15-20%.

The more accurate way to test Crysis is to record your own timedemo and play them back. I recorded a few timedemos on the island map and performance is much better than what is reported by the benchmark.

The ice map is also very good for testing graphics due to its lower CPU utilization.

There you have it folks... Conformation of what we already knew... The timedemo in Crysis sucks and isn't realistic.

[Zertz]

It looks like it's true, but then again, nVidia says alot of things

[JuneBug]

It would be nice if we could take the lag out of thinking for the major corporations, might just get some products out on time that work like they are supposed to.