Sometimes people aren't in the mood to read through dozens of pages, that just so happened to be the case for me. In case you hadn't noticed.
I also recommend that you update the first post so that it's easier on everyone.
Anyway, I'll check it out.
I take it that this doesn't work for AMD does it? I do have up to SSE3 but nothing beyond it.
Well, my builds should work on AMD systems too, but generally, ICC builds run better on Intel systems.
Edit: First post updated wih information about ICC.
I tried starting the latest up and I get a error about "libiomp5md.dll" not existing.
oops, i forgot to bundle the *.dll with the package, sorry.
The problem is now fixed
I used this patcher on another ICC build from someone else and performace increased a bit. Gonna try that with yours.
You mean the Intel Compiler Patcher? If yes, i'll try it out
(07-02-2012, 07:05 AM)DefenderX Wrote: [ -> ]You mean the Intel Compiler Patcher? If yes, i'll try it out
Yes, that's the one. Though it didn't really do too much for me with your build.
(06-23-2012, 01:52 PM)Streeter Wrote: [ -> ]I'm tempted to try and see how much of a difference the SSE instruction sets make. If someone else is willing to do some objective benchmarks then that would be swell, as I'm not about to switch off the official builds until I see the numbers.
If no one's up to the task, and next time I lurk the forums none have taken the challenge, then I might have to do some thorough benchmark testing myself. I implore anyone else who's interested to give it a shot, as I've got other projects on my hands right now.
I guess no one wants to do any objective measurements. *sigh*
Okay, so what are some recommendations to ensure consistent speed benchmarking?
I know I'm going to be turning off the max fps lock, that's for sure. It'll be much easier to compare the numbers.
And considering my processor can take advantage of about every instruction set these builds utilize, this'll be interesting.
Is FRAPS a good way to record FPS?