Caporegime
1.3v go for 4.5ghz mate. If it passes some stability tests of that you have a good chip. Mine did 4.4ghz at just over 1.3v.
---------------------------
IntelBurnTest - Success
---------------------------
Success! Your system was able to maintain its stability while
running IntelBurnTest. This is usually an indication that your
CPU is fully stable, if at least tested with the "Standard"
stress level. If you are testing with memory size below Standard,
please re-test with "Standard" or higher stress level.
Test completed successfully in 121.08 seconds.
---------------------------
OK
---------------------------
What do people class as stability tests these days?
Using IntelBurnTest it gets the CPU up to high 50's. All cores max'd but task manager shows the default clocks? Contradicting CPU-Z.
Hmm, with 4.5 set with 45 multi, 1.3v on the CPU and memory all as auto and not XMP, it crashes as soon as I start the bench.
EDIT I see it hammers the GPUs as well. 4.4 is running fine, so will try again afterwards.
Will do mate once I've got things stable, far from that yet Lol! Seems my cpu just does not like 3200mhz so may have to dial back to 3000mhz. Even with vcssa at 1.225v no dice.
Understood mate, thanks!
I've been so busy I've not had chance to play any more with mine. Does seem like I have done OK and I have more to tweak yet. What is everyone using to monitor temps? Have had to use Aida64 because HWmonitor causes instant BSOD and CoreTemp grinds my computer to a halt.
So much fun buying into early tech
Barely had time for any clocking yet and I'm still not sure what to use to stress stability (using Rog Realbench for now), but I've discovered that VCCSA needs quite a healthy bump (1.2v) from stock to get the board to boot at 3200Mhz 16-16-16-36-1T.
So are the 3200MHz RAM any better than the 2400 version? (other than the risk to malfunction as in Kaap's case ). And would I lose OC potential buy getting the 2400MHz version? (About to pull the trigger on the same spec as OP but RAM is out of stock at the moment in OcUK..and I can't find anyone else who sells it lol)
You couldn't of done much testing then, as the difference in bandwidth can be as much as 20gb/s between those two speeds.