I5 6600k 4.6ghz at 1.28v

[Poisonarrow]

I just reached a stable overclock with these settings. I tested it with Aida64 (2 hours), Prime95 and Asus realbench for 15 min.. temps at 60C.
Only with prime95 ut hits 70.

Should i push further? And can i trust Aida and realbench enough?
Becouse everyone advising p95 But that gives high temps..

I work with a kraken x52 (fans at 55-60%)

 

[chenuki]

I suggest prime95 v26.6 small fft setting

 

[Poisonarrow]

Ok and can you explain why?

 

[alphadogg123]

Run some games prime95 is a pain as it gets whispered around the Internet so a lot of people are happy to suggest it as a stress test for overclocking, I only ever suggest prime95 as a stress test for checking your cooling performance as you can leave it running and find out the max temps.
4.6ghz is a good overclock with a 6600k at that voltage and should mean your cpu fan is near silent, but at 1.35v you could probably get 4.8ghz or higher. If your paired graphics card is a gtx 1070 or gtx 1080 I'd try to overclock a bit higher as small increases in cpu performance with these cards can lead to significant frame rate increases in some AAA games.

 

[Poisonarrow]

I tried 4.8 But Its need more then 1.38v to make it work(stil having bluescreen at 1.38v) and Im not comfterble with the heat (up to 80c).
The weird part is that i made 4.7 work at 1.295v now. Its stable for the moment..

 

[alphadogg123]

Sounds like you're a victim of crap thermal paste under the IHS, you could delid but I appreciate it's not for everyone.
My delidded 3770k runs 4.6ghz @ 1.425v with my corsair h60 and only gets to 75oC on the hottest core. I really don't think people appreciate how little cooling these CPUs should require if they had good thermal contact as in soldered. My old 2500k system could run 5ghz 1.42v and max at 70oC with a Arctic Freezer 13 92mm cooler.

 

[chenuki]

Ok and can you explain why?

small FFT in prime95 v26.6 is a good way to test just the core and cache for stability. it doesn't generate as much heat as a blended/large FFT that also tests ram. simply a way to test core stability and not necessarily the cooling solution. prime95 version 26.6 is preferred to avoid testing with AVX instruction set, which sometimes hinder attempts for higher clock. With AVX you are testing for a more stable overclock under a less realistic load condition.

 

[Poisonarrow]

So When can i say i have a stable overclock? What do i run and How long?

 

[chenuki]

So When can i say i have a stable overclock? What do i run and How long?

I usually do quick 20 minute prime95 tests to find the max, then if a final round overnight can pass without errors then I'd qualify as stable, then move on to my RAM.

Your temps look good, I think you can push for more. GL