I'm going to begin overclocking my new Phenom II 955 BE setup soon and after trying to ask a bunch of questions to the forum I've really boiled it down to just one. I'm willing to use trial-and-error to test my ideas, answer most of my own questions and learn from them, but I'm not sure which benchmarking utility is the most appropriate for my purpose.
I am trying to maximize gaming performance but don't know how to run a repeatable, controlled experiment in-game like I see in so many of the tom's hardware reviews. It seems just loading up and playing a game while watching FPS is an unsound method. However, not all synthetic benchmarks realistically simulate the requirements of a game. What method does tom's hardware use to get average FPS results for different system configurations? What other ways are there to test system gaming performance? I intend to do a lot of tests with various configurations and I want to use a good method otherwise the data I get will be useless.
****Additional thoughts related to the question above****
The reason I want to know the answer to the above question is I intend to do more than just find the highest possible CPU speed. I would like to compare system performance from increasing the HTT vs increasing the CPU multiplier, and decide whether or not adjusting RAM multiplier, vdimm, and vcore are worthwhile or not. Maybe I'm going about this the wrong way and I should only adjust HTT or the CPU multiplier and leave everything else alone, but I think I can get more out of a system playing around with more settings. Am I right in thinking this or am I overcomplicating the situation?
Thanks RJR. I had read through that thread's link but had no idea which one was most appropriate. I'm sorting through the different versions of 3DMark and it looks like I want 3DMark Vantage Advanced Edition. Is that right or is there a different version I should be looking at?