navis995 wrote:
Regarding FurMark. Some people reported it destroyed their GPU's. So I would advice not to use it. Instead try Heaven's benchmark with high/max preset. I actually reached higher temperatures faster than using FurMark. While for CPU. If you have a program like Sony Vegas to test the temperatures that would be fantastic as I've tested mine with Intel Burn Test and the temperatures were a lot lower than when I tested them with Sony Vegas.

navis995, although we appreciate the heads up, that is a really old issue, furmark downclocks when it detects an Nvidia GPU, and Nvidia downclocks when it detects furmark/OCCT - safely running under the limits of the Nvidia GPU. We have been using furmark for a long time after those incidents of many years ago.

Does Furmark really kill video cards?
http://www.overclockers.com/forums/showthread.php/677416-Does-Furmark-really-kill-video-cards

GTX580 throttling question
http://www.overclockers.com/forums/showthread.php/659870-GTX580-throttling-question

Don't do this 🙂

Disable GeForce GTX 580 Power Throttling using GPU-Z
http://www.techpowerup.com/134460/disable-geforce-gtx-580-power-throttling-using-gpu-z.html

That overheating problem was solved by the benchmarks long before the Nvidia hardware solution was implemented in the GTX580.

Unigine benchmarks are good too 🙂

navis995 wrote:
Regarding FurMark. Some people reported it destroyed their GPU's. So I would advice not to use it. Instead try Heaven's benchmark with high/max preset. I actually reached higher temperatures faster than using FurMark. While for CPU. If you have a program like Sony Vegas to test the temperatures that would be fantastic as I've tested mine with Intel Burn Test and the temperatures were a lot lower than when I tested them with Sony Vegas.

I ran FurMark to burn in my GPU for one hour at three difference times for a total of three hours.
(Do not run these burn ins consecutively, give your GPU a good cool down time, at lease one hour, or even better the next day)
FurMark is perfectly safe, that's if you run the program correctly.
Set the temp alarm within spec for your GPU, and most of all, keep an eye on the burn in process.
You don't want to just run FurMark burn in and leave it.
Also running a program like HWiNFO64 (sensors) at the same time will help confirm temps.
Running HWiNFO64 is good for confirming if any thermal throttling has occurred during the burn in.

Hope this gives you some insight on the process.