how bad is this?

T

TroyX

[H]ard|Gawd
Joined
Jan 17, 2012
Messages
1,461
during a cpu stress test with OCCT core #0 is running 20c hotter than the core core #3

and temps from core #1 and #2 are in between.


normal or something is wrong?

im using an aftermarket heatsink and fan.


right now im idle temps are:

core #0: 49c

core #1: 42

core #2: 40

core #3: 39
 
your heatsink is probably cocked

tiny particle of grit in the TIM, one side tightened down harder than the other, uneven TIM application

I would try a remount
 
Although I see nothing wrong at all with your idle temps, remember that idle temperatures are not very accurate.
 
during a cpu stress test with OCCT core #0 is running 20c hotter than the core core #3
Click to expand...

During the test were all 4 cores pegged at 100% for at least a few minutes and you measured the temp during this? I mean if one of the cores was 25% while the other was 100% the 20C temp difference can be reasonable.
 
drescherjm said:
During the test were all 4 cores pegged at 100% for at least a few minutes and you measured the temp during this? I mean if one of the cores was 25% while the other was 100% the 20C temp difference can be reasonable.
Click to expand...

yes, all cores were at 100%

i just finished re-adjusting the heatsink brackets and other things and got it more stable now.

core #0 stressed is now 68c

core #3 stressed is now 58c

core 1 and 2 were somewhere in between but much closer to core #3


alot better but i still think they should be 3-4c apart, core #0 must be doing other processes while OCCT is running, that would be my theory.
 
Last edited:
TroyX said:
yes, all cores were at 100%

i just finished re-adjusting the heatsink brackets and other things and got it more stable now.

core #0 stressed is now 68c

core #3 stressed is now 58c

core 1 and 2 were somewhere in between but much closer to core #3


alot better but i still think they should be 3-4c apart, core #0 must be doing other processes while OCCT is running, that would be my theory.
Click to expand...
I see something similar on my 3570k. During IBT I get 60 on the first core and 70-72 on the other three cores. I have redone the TIM and remounted my H100 multiple times and it hasn't changed anything. Maybe the issue is with Intel's TIM application under the IHS or maybe one core just runs cooler than the other three?
 
Uneven TIM application or a problem with the TIM? Definitely remount with a good cleaning and new TIM.
 
night_2004 said:
Uneven TIM application or a problem with the TIM? Definitely remount with a good cleaning and new TIM.
Click to expand...
I would redo it just to make sure it is even. Another weird thing about my setup is that even though the first core runs the coolest and the other three run about the same at load, the third core is 5-6c cooler than the other three at idle.
 
If this is an Ivy CPU, I've seen this exact core temp difference on everyone I've built or helped build. More and more people are commenting on this also.
It's not your cooler mounting, it's either the Intel Tim on the die or there's something else amiss with how Ivy distributes loads to each core.
 
TroyX said:
yes, all cores were at 100%

i just finished re-adjusting the heatsink brackets and other things and got it more stable now.

core #0 stressed is now 68c

core #3 stressed is now 58c

core 1 and 2 were somewhere in between but much closer to core #3


alot better but i still think they should be 3-4c apart, core #0 must be doing other processes while OCCT is running, that would be my theory.
Click to expand...

Are you running a chip with on die graphics/memory controller such as sandy bridge? If so, don't forget that it can act as a 'heatsink' of sorts for the 2 nearest cores. In theory, increasing their respective surface areas. http://images.bit-tech.net/content_images/2011/01/intel-sandy-bridge-review/sandy-bridge-die-map.jpg
 
You must log in or register to reply here.
Back
Top