Originally Posted by Cakewalk_S
How are you guys determining 'stability'? Furmark and Kombuster is great yes, but other than seeing a ton of artifacts that I can notice by the eye, is that the only way of checking for 'stability'? I'm playin bc2 and I'm not getting much over 50% GPU usage when I've got vsync on... so theres not much of a need to run higher clocks but when bf3 comes out I bet I will need OC.
I like OCCT test but I get a mixed pool of people saying its bad or it stresses the gpu too much. Its great it tells me how many errors I've got but I also get weary with it when my GPU nears 80C @ 1.075v @ 940MHz. In real gaming situation my GPU hasn't seen a temp on bc2 above 52C...
I just want to know where I can accurately get a 'stable' definition for gaming without running OCCT test persay when my GPU is freaking on fire, which I consider anything 70C over to be my danger zone...
Its just we say, well find your max clock @ stock V. then up the volts. Well I find 920MHz is max, up the volts from 1.025v to 1.075v and still get errors @ 940MHz...so how much higher volts should I go?
Edit: Well after realizing OCCT is a good 'stability' test. I realize that 920MHz is my limit to my GPU. Upping the voltage from 1.025v to 1.075-1.086v doesn't grant me 20MHz but certainly adds a good 8C to the temps. 920MHz @ stock volts it is...
I use gaiming for testing clocks. My GPU is ment for gaming so if it's gaming stable then it's stable for me. Simple as that. No need for extreem burn tests.
Anyway. I'll join the club when I'll get the time to take some photos and upload them.