PC Processors

[linxiaopi]

I just installed a Ryzen 5 7600 CPU and am running a CPU stress test (C-Ray, 5k). All cores are at 100% load. The core temps are mostly in the 30s C. However, one core is at 60 C. That's a 30 C difference between the coolest and hottest core. The next hottest is at 40 C. I understand that 10 C difference between cores is normal, but 20 to 30 C seems way off.

I am using a Thermalright Peerless Assassin Cooler with the provided TF7 thermal paste, which I spread with a spreader, so I know I got full paste coverage on the CPU.

I returned my first 7600 for the same issue, but the replacement is behaving the same. Is this normal or am I just having bad luck?

Here's a screenshot of the temps and CPU load:

The left window shows the 12 threads at 100% load with a dip between stress test runs. The right window shows a graph of the 6 CPU core temps and current, min, and mx temperatures for each core on the right panel. CPU 3 is the hot one, although CPU 2's temp never goes below 41 C at idle, so I'm a bit concerned with that as well. The rest of the cores (including 3) hover around 30 - 34 C at idle (not shown here).

[FunkZ]

What software is displaying those core temperatures?

It shows 5 of the cores min/max only 1-2°C variance, and 1 core min/max 13°C difference.

It is highly unlikely those are accurate running a 100% load.

Have you tried Core Temp or HWiNFO?

[linxiaopi]

Thanks for your input!

I'm using Mission Center to display the load, and Psensor to show the temps. These are Linux software. The 1-2 C min/max difference is probably because I started the monitor after the benchmark had already started. The 13 C difference is on the core that is running way hotter, and the min temp may be reached when between when between benchmark runs.

I agree that the min temp diff is suspicious as they should probably drop more than 2 C between runs as well.
I'm going to reapply thermal paste and try again. If I don't see a difference, I'll install windows and try HWiNFO and/or Core Temp.