Originally Posted by Nunz
There is no ****ing way they ran that test 16 times and every single time matched exactly. There's no variance? Margin of error? Nothing?
Most likely each CPU/resolution was run multiple times. Any variance will just average out if you run the test enough times. It seems like Steve from Hardware Unboxed runs an ungodly amount of benchmarks, so I doubt he'd bother with this if he wasn't going to do enough runs to pretty much eliminate random fluctuations.
If you lower the settings enough in any game you can make it CPU limited. It's just at realistic settings you're going to be GPU limited almost all the time.
For some reason people just don't want to accept that your CPU doesn't matter that much for gaming.