Specs: I7 8700k Asus maximus X hero G skill trident z rgb 3200 1080ti
I've been working on over clocking my cpu for the past couple days. Due to my cooling limitations I'm keeping it at 4.8 ghz. I was able to actually tweak the voltage all the way down to 1.155 and it would boot into windows and run real bench for an hour with no errors. I was getting an odd screen flicker though so I decided to up it to 1.168 vcore.
I was able to pass an hour of each real bench, p95 blend, Intel burn test, and 10 runs of Cinnebench. I even played a few games. Temps peaked at 76 C and everything seemed stable. A friend recommended I try OCCT, so I did and failed within 1 minute. I figured well if I failed that it must be unstable. So I upped the vcore to 1.184. Everything passed again for an hour except OCCT. It was able to make it 10 minutes this time. Now I'm running at 1.200 and was able to get OCCT to pass for an hour so far.
My question now is since it all felt really stable and tested well outside of OCCT at 1.168 is it worth going back? I know this question is really open to interpretation. So is OCCT generally a reliable test as I've never used it before. I figured prime would have hammered my cpu harder as well as extreme Intel burn test. Is it better to stick with 1.200 and do more extensive testing. I was thinking 3 hours of real bench, etc.
OCCT is a more stringent test than Realbench. The voltage difference you needed agrees fairly well with my experience. Cinebench is not as good as Realbench. My current 24/7 CPU passed a long Realbench test, but crashed at a critical point of a Creative Suite workflow. No such crashes since my stability criterion has been OCCT.
Lately I've been using y-Cruncher. Look it up on HWBOT. It chooses from several algorithms to use the set of AVX instructions in the CPU under test. It also needs very good memory stability. Most important, it checks the computations.
Prime 95 is more harmful than useful. Prime 95 is all stress and no test in that calculation results are not checked. When passing Prime 95 you know only that it did not crash.
I haven't heard of y cruncher either. I'll definitely give it a try. I was able to run 2 hours of occt under large data set but it failed almost immediately in small data sets. Perhaps it's still unstable. I'll for sure give the y cruncher a shot and see what comes up.