Rinku

I switched from 24 GB Ram to 16 GB and my CPU clock limit under GPU load went from 2.6 Ghz to 2.7 Ghz. Kinda weird but I assume the Ram power consumption has something to do with that. It was BS, sorry. Edit : I just used Coollaboratory Liquid metalPads and I gained something like 10-15C !

I was just thinking about giving CLU a try. And yea, max fans is too loud for me to play constantly with. I'll keep looking for undervolt solutions and try CLU. EDIT : I figured what caused me premature throttling. It was the thermald package, installed by default in Ubuntu 16.04. I just removed it and my clock stay at 3.3 Ghz until something like 98C. Then it varies from 3 Ghz to 3.3 Ghz. Though I'll still repaste with CLU or MetalPads and i'm still interested in overvolt under Linux. And when both CPU + GPU are in load, my CPU clock stay at 2.6 Ghz, as you mentioned. Is the 180w PSU guilty ? Thanks !

Thanks for the reply. Isn't there any way to undervolt the CPU through the BIOS as Prema Bios is unlocked ? Or if anyone knows how to undervolt on Linux, I'd appreciate it. I already tried to repast (using Noctua NT H1), but it didn't really help, the temps still reach 90 - 95C way too fast.

Hi all ! I have a temperature problem with my Clevo P670SG. My processor heat a lot and the frequencies start to get lower when the temperature reach about 85C. I use the prema bios and I'd like to know if there is any tweak in the bios to reduce the temperatures (without lowering the turbo clocks) or to prevent the clock scaling by setting a higher limit for temperatures. I'm using Ubuntu but the problem is the same on Windows 7. I join a screenshot of the temps during a cpu test. The weird thing is that the temperatures get really high within a second after the test start. And then the clock is auto adjusting to keep the temperatures 'cooler' (the green sensor is my GPU). I have a 4710HQ. Thanks ! Edit : A second pic showing my temps during a quick gaming session (Skyrim through Wine) :