Furmark and GPU overclocking

Toggle sidebar Toggle sidebar
B

Barney-

Honorable
Aug 3, 2014
42
1
10,545
HI. New to overclocking, hope this isn't a dumb question!

I'm trying to find a good stable overclock for my new GPU (EVGA 1060 3Gb), and I'm using Furmark to stress test it.

My understanding is that with a benchmark like Furmark, the GPU will be pushed far, far further than any game (present or future) could push it. Therefore I'm happy with an overclock setting which pushes the GPU to "dangerous" temperatures, i.e. above 84 deg C, justifiied by the notion that any real-world usage will be so much more lenient than Furmark that I would always stay 10 deg C (or more) below what I get on Furmark.

So, obviously, my question is whether my assumptions are good - is an 87 deg C (say) result on Furmark always likely to translate to a much safer temperature in real world gaming?

Thanks
 
Sort by date Sort by votes
feelinfroggy777

feelinfroggy777

Glorious
Dec 13, 2016
5,721
135
45,990
Furmark is more of a stress test than benchmark. I would not recommend running a Pascal card at 87C. 84C is a good top temperature target.

I run my benchmarks through Unigine Heave and Superposition. Fire strike is another good benchmark. If you can run through Heaven for a couple loops and you are not getting any artifacts or driver crashes, your overclock is stable.

Run a bench of Heave at stock settings and write down the score for your baseline. Then run heave in windowed mode with MSI afterburner or EVGA Precision X on the side. Move your base clock up in 5-10 increments until it crashes. When it crashes, back down to the previous setting. Do the same for the memory. Then you can increase the power to 120% and see if you can get more. Additionally you can increase the voltage for even more. For daily use your GPU should not be running over 80C.
 
Upvote 0 Downvote
B

Barney-

Honorable
Aug 3, 2014
42
1
10,545


Wow, thank you for that post, I realised Furmark places far more stress on a card that any real game would, but I didn't realise it could actually damage my card, so thanks for the heads up. I suppose I should be pleased that my card was able to reach 106 FPS on a test so extreme that I'm lucky that it didn't burst into flames!

I'll stick to using the benchmarks recommended by the other poster, thanks to both of you.

 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom