Question How do you validate a thermally limited OC?

Oct 13, 2024
61
22
35
This is mostly a theoretical question: Lets say I have a CPU that I want to overclock. I mainly use Prime95 to test for stability because, in my experience, if an overclock will fail it will probably do so pretty quickly if not immediately. The problem is that p95 with small ffts pushes the CPU to the absolute limit in terms of power draw (which is kind of the point i guess), so an oc that may or may not be stable in normal workloads (meaning normal power draws and temps) can't be tested properly because the CPU is at TJmax and throttles down.

What do you guys think?
 
Last edited:
This is mostly a theoretical question: Lets say I have a CPU that I want to overclock. I mainly use Prime95 to test for stability because, in my experience, if an overclock will fail it will probably do so pretty quickly if not immediately. The problem is that p95 with small ffts pushes the CPU to the absolute limit in terms of power draw (which is kind of the point i guess), so an oc that may or may not be stable in normal workloads (meaning normal power draws and temps) can't be tested properly because the CPU is at TJmax and throttles down.

What do you guys think?
I think you would need to get a better cooling solution.
 
I think CPU-Z bench/stress test comes closer to a realistic workload.
Monitor the temperatures with hwmonitor.
You may see one or two cores touch 100c. in red, indicating temporary throttling, but that is normal.

For certain, first update your bios to currency to avoid the dreaded 13/14th gen issues.

Then, if you are a gamer, reconsider the urge to overclock.
There will be very little stable headroom.
Chips are binned, and a higher quality chip would have been used in an i7-13700K
Games do better to rely on the thermal boost which may well get you a better clock for a couple of cores than an all core overclock.