PDA

View Full Version : Testing the stability of an OC?



WillyWonka
05-17-2007, 01:46 PM
This is what I normally use:

12 hours orthos small fft on both cores
2 hours orthos blend on both cores
overnight memtest default options
2 hours sandra burn-in suite

Am I missing anything :confused:

Neon Biker
05-17-2007, 01:52 PM
Give this (http://www.testmem.nm.ru/snm.zip) a try. Just make sure you set it to 100% load. ;)

WillyWonka
05-17-2007, 01:58 PM
Give this (http://www.testmem.nm.ru/snm.zip) a try. Just make sure you set it to 100% load. ;)

What does that offer that memtest doesnt? Its rather sparce on the documentation side of things :shocked:

burningrave101
05-17-2007, 02:05 PM
If you want to really test your memory then run Windows Memtest and run enough iterations that you use up almost all of your available memory. Windows Memtest is more stressful on the memory than running Memtest in DOS is. The memory becomes less stable when the OS files are loaded into it and running on it while you're testing. That's why Memtest can be pretty easy to pass at times and why your system may still fail a heavy load on the FSB/Memory.

Bail_w
05-17-2007, 02:19 PM
download OCCT and run it with 12 hrs.

Neon Biker
05-17-2007, 02:29 PM
What does that offer that memtest doesnt? Its rather sparce on the documentation side of things :shocked:

Well, there probably is some documentation on the official website, but since it's in russian I wouldn't know. Anyway, documentation or not, in my personal experience:
the CPU FPU test is the best CPU stability tester that I know of
the memory tests 2-4 are the best RAM stability testers that I know of
Temps and Voltage monitoring

I'd take an hour of that FPU test on 100% load over any amount of time in Orthos or anything else and the same goes for those RAM tests, but don't take my word for it, try it for yourself. ;)

WillyWonka
05-17-2007, 03:28 PM
Right. I'll give those tools a whirl.