[SOLVED] Will temperature differences cause problems?

Feb 22, 2020
15
0
10
Hello everyone,

I have a asus Z10-p8 running with two E5-2690 v3.

If I run prime 95 for ~30 min the temperatures stabelize at around 71 °C for package 1 and 66 °C for package 2 (according to HWInfo). Within the different CPUs the tempertures vary by ~5 °C, with the package temp representing the max temperature (so no core is hotter than that).

Now I was wondering, if this could cause performance issues, if the cores have such different temperatures (since between the coldest and hottest core there is a 10 °C difference).

I already renewed the thermal paste, but the package temperatures remained the same. Since the temps themselfes are with their allowed limits, I am only worried about the difference. This workstation is intended for FEM calculation and I would hate for this to be the cause of unexpected errors or things alike.

Thanks in advance :)
 
Solution
CPUs generally don't suffer performance degradation due to heat as long as they are operated in their acceptable temperature zone.
I wouldn't expect their to be any difference in performance between the two as long as you stay in that zone with both.
Feb 22, 2020
15
0
10
Ok thanks. I thought, that maybe the increase of the electrical resistance would case a slightly reduced performace which then interferes with multicore application.

But you are saying a core at 40 °C has the same performance as one at 80 °C?
 
Ok thanks. I thought, that maybe the increase of the electrical resistance would case a slightly reduced performace which then interferes with multicore application.

But you are saying a core at 40 °C has the same performance as one at 80 °C?
If 40C and 80C are both considered acceptable for that model CPU and are below the temp the CPU throttles itself....then I wouldn't expect them to perform differently.