It depends on the test (what bench/game/etc) you run, what settings you run that test at, and what your own personal definition of 'bottlenecking' is.
If by 'will it bottleneck?' you mean 'could I ever
lower FPS than I would see if I had infinite CPU power?' then the answer is most definitely yes. But the bottlenecking you get in general running that CPU with that GPU is going to be very minimal, and very infrequent, esp. running at settings appropriate for the gfx card (high resolution/AA/IQ levels).
Generally-speaking single GPU solutions won't be CPU-bottlenecked to any significant
degree unless you have a ridiculous imbalance in GPU vs CPU power.
Originally Posted by CountChoculitis
I have a older Phenom II 920 @ 3.7ghz/NB at 2400 and my GTX 570 is not really bottle necking me. Only Starcraft 2, WoW, and other poorly coded games that do not use all 4 cores create any sort of cpu bottleneck. I was going to buy a i5 2500k/mobo/ram but I cannot justify it for $470.
Some benches that make me think I am not CPU bottlenecked.
(notice graphics scores are the same)
3DMark11 @ 3.6ghz P5333-
3DMark11 @ 3.7ghz P5434-
3DMark11 Extreme graphics scores are within 5 fps of each other also.
Unigine Heaven at 3.7ghz gave me a whole .1 fps better from 3.6ghz.
I would think if I had a CPU bottleneck my scores would go up by a lot if I increased my CPU frequency.
The part of the equation you're maybe not understanding here is that you do not either 'have' or 'not have' a CPU bottleneck based on your hardware alone
What you've observed and tested here is whether or not you have a bottleneck on THOSE TESTS, at THOSE SETTINGS you've run them at. And it appears you don't in those particular situations/tests. But that means nothing in terms of other testing scenarios you might construct. I.E. you can HAVE a bottleneck in Starcraft 2, and NOT in Heaven and 3dMark11 ... even though your hardware is the same Edited by brettjv - 1/16/11 at 10:51am