HiVizMan wrote:
Does this happen when your system is at default? If no then your OC is not stable for that game.
Stress tests do not mean your system is stable, they only mean your system can pass that stress test. 🙂
Could this be incorrect OC of my cards? I think @ 1137 MHz I have no such problems. During 12 hrs under FC3 all was OK, then another "tests" for 2-3 hrs.
Arne Saknussemm wrote:
I would normally say look to memory for this kind of thing first......but it might be drivers crashing etc. any clues from event viewer? Any flickering lights in the house....you having power outs or anything of that sort?
Though, If you have tried replacing everything maybe does point to cards....try running them at stock does this still happen? maybe try running them PCIe 2.0 instead of 3.0.....there are signalling issues with 3.0 sometimes.....
I can say only one thing about RAM - Memtest few passes (at 2400 MHz), 15 hrs under P95 Blend and 3 hrs under LinX. But sure it's OC'ed (2400 MHz, like it should be).
In event viewer - nothing.
PSU is connected to APC SUA2200I. And I even tried P95 + 3DMark (Fire strike) together, so it was even higher load - nothing, even with higher GPU clocks.
So if it's incorrect GPU OC, why motherboard will cut power off for a few secs? What about black screens etc.? I mean normal things for that kind of situations.
Or maybe it's not GPU? Not enough VTT? VCCSA? But Why @ 1137 MHz works OK? Or maybe this is a coincidence?
I've also connected power cable to onboard PCI-e connector (like always).
Melting Point wrote:
Try using PCIe 2.0 and see if that fixes it. It may be related to the know PCIe 3.0 issue.
There is one strange thing - I had that problems earlier even with one card (at stock) and even in PCI-e 2.0. But after that I've changed PSU, CPU, MB and MEM, and also started my CPU/MEM OC from scratch.
