"I didn't do proper science and try one at a time so I don't know which fixed it."
Sometimes even then lol.
Much of the info on ntkrnlmp.exe via links is non-explaining really. They don't say why or how its involved in a good way.
But, I see Microsoft docs can slowly help make a connection between c-states, CPU, and ntkrnlmp.exe at least eg (don't view this as a fix here. Its just used to link the 3 mentioned points) "The Ntkrnlmp.exe file that is included in hotfix 840987 includes a sequence of instructions that reveal this bug.", "when a processor resumes from a C1 idle state" https://learn.microsoft.com/en-us/troubleshoot/windows-server/performance/stop-0x0000000a-error-processor-c1-idle-state
I know from my last motherboard with this computer I'm on that c-states and older PSU's with newer CPUs can be used in disabling the low power state that will render the computer with an issue randomly. The motherboard I'm on now deals with all that automatically though while the other one didn't which is quite amazing. But, that certainly can apparently cause an issue with ntkrnlmp.exe.
That's far more likely than the power cable and outlets :)
But, it goes to show sometimes people wrongly view problems. If the problem is with the computer actually working improperly then the software isn't to be done for a fix (most times). There can be exceptions.