It is rare, but CPUs can have all sorts of issues that show up as random instability, from the somewhat recent Intel oxidation problem that impacted supposedly only 13th & 14th gen, to poor contact being made on a bunch of different LGA style CPUs & sockets (sometimes caused by cooler cold plate shape or poor install/ bad design), to burned contact pads/ pins on AM5 from some fault.
I had a brand new from Microcenter AMD 7960x Threadripper have a bad memory controller or other internal fault, it had one stick of memory not be usable, swapped RAM around and the issue followed the socket, replaced the motherboard and the issue was the exact same, I had reseated the CPU a couple dozen times, very closely inspected the pads & pins. Once I swapped the CPU there hasn’t been a problem.
Every single peripheral, PCIe lane, audio chip/ chipset or anything else connects to the CPU, if there is an internal fault in the CPU it will show up on a new motherboard. If the motherboard’s path to the CPU has a problem then a motherboard swap should resolve it.
Interesting. Are these all the SMART metrics? Normally NVMEs have metrics such as these that tell how many times the controller reached warning and critical temperatures:
Warning Comp. Temperature Time: 0
Critical Comp. Temperature Time: 0
And thats a cheapo Samsung.
I do have some WDs that one time got really toasty during a large copy operation:
SMART/Health Information (NVMe Log 0x02, NSID 0xffffffff)
Critical Warning: 0x00
Temperature: 46 Celsius
Available Spare: 100%
Available Spare Threshold: 10%
Percentage Used: 4%
Data Units Read: 192,088,587 [98.3 TB]
Data Units Written: 439,978,154 [225 TB]
Host Read Commands: 1,687,095,272
Host Write Commands: 5,518,285,987
Controller Busy Time: 12,555
Power Cycles: 845
Power On Hours: 23,208
Unsafe Shutdowns: 164
Media and Data Integrity Errors: 0
Error Information Log Entries: 1
Warning Comp. Temperature Time: 18
Critical Comp. Temperature Time: 0