[SOLVED] I ran my i9-10850K around 100 °C for nearly 4 hours. How screwed am I?

dc200

Distinguished
Jan 11, 2012
16
0
18,510
I started a video processing job and left it to run overnight. The following morning, I found that the job took 4 hours to complete, and that my CPU (an i9-10850K) had been running very close to 100 °C that entire time.

I've since figured out what caused the overheating and have fixed it to ensure that this doesn't happen again, but I'm still worried about what damage I might have done to my CPU in those 4 hours (temp graph below).

How screwed am I? Would this have significantly shortened my i9's lifespan? And is there anything else I should be checking for as well?

5kSscsL.png
 
Last edited:
Solution
You've got @ 3+ Billion transistors in that cpu. You could fry a couple tbousand and never see any difference in performance.

The only thing you Might have done is lower the physical lifespan of either the motherboard or the cpu, and since there are cpus and motherboards that still work after 40 years of service, it's pretty much a given that you've realistically done nothing to lower the Usable lifespan of the cpu or board. It'll be obsolete long before it's dead, barring any catastrophic abuse or damage.

If you are reaching 100°C with a 250w capacity cooler on an intel cpu that's supposedly correctly mounted, obviously the OC applied and respective voltages used are beyond the cooling solutions capacity. Meaning lower the...
tjMax is 100C on that CPU so you should be OK. The CPU has a built in safeguard and would throttle down (As your graph seems to indicate) or shut the system down. A diminished CPU shows signs of constantly slowing down and there's no coming back from that condition.

Nevertheless it's not a good idea to consistently run your system under load at those temps. I like to keep the temps no more than 80C when under load for longevity.
What type of cooling do you have?
 
  • Like
Reactions: CompuTronix

Phaaze88

Titan
Ambassador
The cpu went from lasting 12 years to 11 years and 10 months. You'll still lose interest in it before it actually keels over.

Just go into bios and set a lower thermal limit.
-We all know running a cpu at excessively high temperatures for extended periods of time isn't good.
-Just because the cpu has a hard thermal limit of 100C doesn't mean we have to let it run close to that.

Set a thermal limit of 85C in the bios if need be - it's not hurting anything, if helping the cpu.
And if you're still sustaining temps up to that, then something needs to change: the cooling solution, bios settings, or both...
 
  • Like
Reactions: PC Tailor

dc200

Distinguished
Jan 11, 2012
16
0
18,510
Sure, I'll look into that. My cooler is a Dark Rick Pro 4 and I've also verified that it's seated correctly.

Is there any CPU benchmarking software or something that I can run and compare to ensure that there isn't any performance degradation and also rule out any potential damage? Is it worth doing?
 

Phaaze88

Titan
Ambassador
Is there any CPU benchmarking software or something that I can run and compare to ensure that there isn't any performance degradation and thereby rule out any potential damage?
Negative. It'll do more harm than good anyway.



EDIT:
Is it worth doing?
No, because even IF you did any real damage, you wouldn't know it, because you hadn't taken any records beforehand.


I've done far worse on my 7820X. The cpu is fine.
 
Last edited:
  • Like
Reactions: PC Tailor

Karadjgne

Titan
Ambassador
You've got @ 3+ Billion transistors in that cpu. You could fry a couple tbousand and never see any difference in performance.

The only thing you Might have done is lower the physical lifespan of either the motherboard or the cpu, and since there are cpus and motherboards that still work after 40 years of service, it's pretty much a given that you've realistically done nothing to lower the Usable lifespan of the cpu or board. It'll be obsolete long before it's dead, barring any catastrophic abuse or damage.

If you are reaching 100°C with a 250w capacity cooler on an intel cpu that's supposedly correctly mounted, obviously the OC applied and respective voltages used are beyond the cooling solutions capacity. Meaning lower the workload, lower the OC or raise the cooling capacity/ability or any combination.
 
Solution