Originally Posted by jncunha
Is there a max safe core speed? With 1.2v and 1200Mhz of memory clock, I can get 970Mhz of GPU clock and running for some time in Heaven DX11 without showing artifacts. I've tried 980Mhz but it started to crash. So since I can't go any higher on the clock I would need to higher the voltage but I don't want to risk the card even more.
With these results on Heaven DX11, playing a game like BF3 without any problems and with temps at maximum ~60 ºC am I safe to leave it like this? Should I try to bump the memory to at least 1300Mhz?
Still I haven't found the best way to monitor the highest temperature while testing or playing. I can't use MSI Afterburner because it won't let me tweak the voltage and GPU-Z has a ton of sensors with temps!
Okay, calm down!
GPU-z will indicate what every sensor are reading. This is the best monitor to make sure the card isn't over heating.
Everything saying GPU Temperature/Temp (#1,2,3) try to keep bellow 80-85.
If you have VDDC reading temperature, these are the VRM if I am not mistaken and they are fine up to 115 but I wouldn't run that hot.
You should start GPU-z on background, do a mild overclock and play a bit. Click on the temperature reading numbers so you can set it to show the max value and show us the screenshot!
I usually monitor my FPS via MSI as posted above so I can see if the overclock is really helping to pull more FPS.
Furmark pulls way more power from your card than any other game/benchmark. It is possible a little component is overheating and making the card crash. It could explain why you don't see that happen in game. There can be a ton of other explanation! As long as it is running fine on the game you want to play, then why do you want to find a error that isn't hapenning? Forget Furmark! This program will just damage your card more you use it.
Hope this help!