Question Benchamark/Stress Test for GPU clock stability testing for Gaming Scenarios

[THpapi]

Hey community!
I've been using Firestrike to test GPU OC profiles and noticed that while many of these profiles are stable during testing and yield a much higher Firestrike score than no OC, they are often unstable while gaming. Is there a better tool besides Firestrike to test whether or not an OC profile will be stable in most games? I know some games are picky but just looking for one test I can use that will pretty much prove stability in many gaming scenarios.

 

[drivinfast247]

Run Prime95 ver.26.6 Blend while running Firestrike. That should do the trick.

 

[THpapi]

Haha seriously? Haven't thought of that... Will try it after work and report back!

 

[alceryes]

Prime95 blend is good test but the Prime95 small FFT test is better if you're testing for maximum load/heat related failures.
Firestrike is good but I usually use Furmark instead.

 

[TJ Hooker]

Run Prime95 ver.26.6 Blend while running Firestrike. That should do the trick

Prime95 doesn't involve your GPU at all though...

Just try some more graphics card stress tests/benchmarks. There is 3DMark Timespy, the Unigine benchmarks (e.g. Valley), etc.

 

[THpapi]

Prime95 doesn't involve your GPU at all though...

Just try some more graphics card stress tests/benchmarks. There is 3DMark Timespy, the Unigine benchmarks (e.g. Valley), etc.

Well yeah haha that is my plan... was wondering if anyone had some insight on a particular test they know to demonstrate gaming stability better than the others.

 

[drivinfast247]

Prime95 doesn't involve your GPU at all though...

Just try some more graphics card stress tests/benchmarks. There is 3DMark Timespy, the Unigine benchmarks (e.g. Valley), etc.

Because gaming requires more than just GPU. OP said he wanted to test "gaming scenarios". So running a GPU only test(I use Heaven) and P95 Blend(which stresses CPU and RAM) seems like the logical thing to do. If your OC can pass 20-30 minutes of that, then a real gaming load should be no sweat.

Of course IMO.

 

[TJ Hooker]

Because gaming requires more than just GPU. OP said he wanted to test "gaming scenarios". So running a GPU only test(I use Heaven) and P95 Blend(which stresses CPU and RAM) seems like the logical thing to do. If your OC can pass 20-30 minutes of that, then a real gaming load should be no sweat.

Of course IMO.

Sure, but if you're only overclocking your graphics card then your CPU and RAM should already be stable. A GPU overclock isn't going to make your CPU unstable.

 

[drivinfast247]

It will also stress the PSU and may reveal a weakness. For all we know the OP is trying to OC a 2080ti on a 9900k with a 350w PSU.

 

[THpapi]

It will also stress the PSU and may reveal a weakness. For all we know the OP is trying to OC a 2080ti on a 9900k with a 350w PSU.

Haha sorry I didn't Clarify! 7700k @ 4.8 (less than 1.3v if I remember correctly) and GTX 1080 OC'd pretty hard but no voltage offset. 750w gold PSU. PSU shouldn't be an issue if performing nominally.

 

[tennis2]

I use F@H actually.
It's MUCH faster at sussing out system RAM faults (5-10 minutes vs 19h for Prime95 to catch mine) and it's more stressful on GPU stability than FurMark.