• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

12600K core-to-core temperature variance

S

SundayOverclocker

New Member
Joined
Feb 3, 2023
Messages
26 (0.03/day)
12600KF, stock (in a B660 motherboard so not many possibilities to screw around with it), at 188 watts in the OCCT CPU stress test. Is this too big of a difference?

1749653776825.png


The cooler is IDcooling SE-206-XT, rated for 250W (probably an overstatement), 24 C ambient. Built the system a year ago, recently noticed the CPU spikes a bit high under all-core load. Today I installed the Thermalright contact frame, cleaned the old paste as thoroughly as I could using 96% ethanol, applied fresh out of the shop Arctic MX-4 (previously was IDcooling paste bundled with the cooler). The screenshot above is now, after the procedure.

The one hot core was there before the rebuild too, with about as much variance, but I don't know if it was the same or a different core.
 
did you tighten down and secure the cooler ontop of the chip correctly?

23C diff is quite a lot i feel.
 
I should hope that I did secure it properly, there's just 2 screws that compress a spring, and the screws hit a hard bottom so no guesswork as to how tight to screw them in.
I will double-check that I used the correct stand-offs between the motherboard and the cooler mount bracket.
 
How flat is the plate?

plate.jpg
 
I don't actually have a good tool to test that, like a steel ruler/calipers. But visually it looked fine. Thanks for the tip though, it didn't cross my mind that this could be the problem.
 
There's probably not a problem with the cooler or the mount. The IHS plate is soldered and it could have imperfections under the lid causing a large variance. Id say try and push an RMA on the cpu honestly.
 
Being its a 1700pin are you using stock intel mounting frame(ILM) or you using one the aftermarket ones (ex. https://www.thermal-grizzly.com/en/cpu-contact-frame/s-tg-cf-i ) ?
Stock is the frame that holds the cpu in the socket tends to over time cause cpu bend make temps go up.

Stock is left image, Right is after market.
1749955352338.webp
1749955363101.webp
 
OP stated they are using the Thermalright contact frame.

When you attached the mounting frame to the motherboard, the part the actual heatsink attaches to, you ensured it is fully down? Did you remove the plastic off the bottom of the heatsink??

That to me seems like there is really bad contact.
 
Last edited:
The heatsink mount is separate from the ILM/contact frame, they're not in contact. And yes, I did peel the protective film off the heatsink. Would have been an awfully easy fix if that was the reason!
And it is dissipating 190 W, even if one core is at 100 degrees.
 
SundayOverclocker said:
The heatsink mount is separate from the ILM/contact frame, they're not in contact. And yes, I did peel the protective film off the heatsink. Would have been an awfully easy fix if that was the reason!
And it is dissipating 190 W, even if one core is at 100 degrees.
Click to expand...
I still stand by my previous post.
Because.
8th gen was infamous for this exact issue. The difference is the IHS plates where not soldered and they used some really bad paste. A re-paste would often times drop load temps by 15-20c making them acceptable for overclocking.
 
Thanks for the advice, but RMA is not an option, I did not purchase this CPU officially. Never had a problem with a CPU, didn't think I'd need warranty.
But it's not a common issue with the 12th gen, is it?
 
SundayOverclocker said:
Thanks for the advice, but RMA is not an option, I did not purchase this CPU officially. Never had a problem with a CPU, didn't think I'd need warranty.
But it's not a common issue with the 12th gen, is it?
Click to expand...
No, not really a common issue. 20c is a large difference and not typical of soldered IHS plates.

You could try a de-lid, of course with risks...
 
Interestingly, core #3 is only the hottest one under the most extreme synthetic load. Here is 5 minutes of CPU+RAM test at steady 130 W, which is still higher than I've seen in any useful tasks, but not by much:

1750236421261.png


Cores #1 and 2 are always the coolest, though.
 
SundayOverclocker said:
Interestingly, core #3 is only the hottest one under the most extreme synthetic load. Here is 5 minutes of CPU+RAM test at steady 130 W, which is still higher than I've seen in any useful tasks, but not by much:

View attachment 404161

Cores #1 and 2 are always the coolest, though.
Click to expand...
And why does this screenshot differ so much from the 1st one at the top of this page?
 
You must log in or register to reply here.
Back
Top