Caporegime
why didnt intel make the heatsink screw off
I'm curious - CPU's never used to have IHS' fitted- how do they perform WITHOUT an IHS vs with?
I know it's possible to crush the die, which is why Intel fitted the IHS in the first place, but having 2 junctions with thermal paste vs 1 sounds like a further opportunity to reduce the temp...
Has anyone tried?
Nice
But 1.3v for 4.5Ghz
What LLC setting do you use?
Okay, I think I've got it stable at 4.7 now. The difference in temperatures is amazing though.
Before, if I ran Intel Burn Test at 1.3v, I would get as high as 101 degrees. CPU-Z just showed my core voltage reaching 1.45v whilst doing the burn test (10 cycles) and my average max temperature was 93.5.
...I'm guessing that voltage is a little high though? LLC is now set to Extreme.
Yeah that is a little high, i'm doing 4.4 at 1.220v
I could do 4.4 easily at 1.26, it just seems to require exponentially more voltage the higher it goes. I got it stable at 4.5ghz doing 1.285v, but to get 4.7ghz the lowest I can go is setting it to 1.385.
Here's the CPU-Z screenshot, running Prime95 getting around 90 degrees:
I've set it to 1.385 but the LLC takes it up to 1.44 anyway. When not on full load it's doing 1.4.
The difference in performance your going to see is not worth it from 4.5 to 4.7 in my opinion, especially at the voltage needed
I've put it back to 4.5ghz and it's happy on 1.28v now.
I've set the LLC back to Optimized, after reading through a couple of threads. Also dropped some of the other voltages and settings a bit. Now hitting about 70 in Prime95.
That guide would be very handy . My target is 4.8ghz, but I understand I might struggle to reach that, especially on air cooling.
I did find it interesting that rather than high temps just being a result of overclocking, they actually cause it to be less able to overclock. Do you think with water, or a bigger heatsink I'd be able to hit 4.8ghz more easily? Or should I keep fiddling around with my voltages and settings some more...