I am starting to undervolt my CPU with Throttlestop.
At the same time I am monitoring with HWinfo64.
When I make offset -100mV and start prime95 for stresstest, then HWInfo showing me for each core 1,08-1,09V. Before offset it was ~1,11V on each core @3,1GHZ during prime95.
So it is just -20mV during 100% load, why no -100mV like I did in Throttlestop?
Is it reading error?
2nd question: Why is Volt higher on cores while there is no load at all?
when I start prime95 the Volt goes down on each core?
-
-
Don't use prime95, use OCCT large data set with infinite mode.
Core voltage variances are normal, it's ok to see one core requesting a different amount of voltage compared to the other.
Try undervolting with stress test OCCT, I'd say -35mv on core and cache, then work your way down.
Stress test should be run at least for 8-10 hour for checking, 24 hour if you're confident enough.TomJGX likes this. -
ty for reply. But my question was, why do the cores only take 20mV less, when my offset is -100mV.
Is this the normal case?
I will try OCCT, thank you ! -
-
I'd always test for at least 10h OCCT, short benchmarks like that don't demonstrate proper usage scenario -
I used to run 10 minutes stress max, passed it with -180mv, so I decided to go with -145mv. This -145mv is also very decent, stable and never had any issue for two months now.TomJGX likes this. -
-
I use Intel XTU and I have it set at -175.xxx mV (I think it's mV anyway). I remember only encountering problems beyond -190, so overall I'm happy with it. Slashed around 2-3C from my stress tests. Never ran any tests longer than a few hours.
Questions Undervolting i7 6700HQ
Discussion in 'Hardware Components and Aftermarket Upgrades' started by DaReference5754, Aug 19, 2016.