INTEL NUC 8i5BEB 8i5BEH -some cpu cores died?!?!

[d_ne0]

Hello to everybody!
I am asking some experienced members opinion on the following problem.The NUC was working fine until new win 10 install and while doing some updates just freeze restarting on bios splash screen.At first I thought is a corrupted bios,took a dump with ch 341,and asked for help on winraid forum,got the bin cleaned...etc and rewritten on the chip,same situation,freeze booting at splash screen.Trying to install ubuntu server failed also but,freezing when ubuntu said "bringing all cores up" something like that...That got me thinking might be something wrong with the cpu.Disabled from bios 4,3,2 cores and at 1 last left the pc was booting again into windows without problems.I was having a little hope that might be something wrong with power to the cpu.vcore ,vrm;I identified 3 coils that at boot time have 0.89V all 3 that seem normal,reported also in bios.I dont have schematic or service manual dont know what I can try further.Now what do you guys think?Are the cores dead?
Thank you very much for your replies in advance!Every opinion is valuable !
L.E. the voltage on the 3 coils drop and rise when in windows,I guess is it normal?Boost or back when needed...I am not sure how it is supposed to be.

[DynaxSC]

I bought some time ago a cheap Intel 8 core CPU. It was said it is partially broken. And yes, I have tried to boot Windows with all 8 cores, and it was crashing. Then I have switched off 2 cores in the BIOS and voila, the board booted correctly. Sometimes also I get some issues booting with 6 cores and cold CPU, but after the CPU get somewhat warmer it boots correctly. So to say, it is possible that individual cores of a cpu die. On a NUC its difficult to test the cpu on another board, difficult but not impossible. So to go sure you would have to transplant the cpu on another same or similar board, meaning desoldering, reballing and soldering the cpu on another board. This is probably the only way to test if the cpu is good or not. Other approach could be to buy a new cpu and exchange it.

Theoretically the issue could also be the ram, not the cpu. This would be worth to evaluate, as ram has according to my experience a higher probability of failure than cpu.

Another possibility is a bad contact between cpu and the board pads. Reflow or better reballing might help. Same can apply to ram chips.

[d_ne0]

Thank you very much for your answer! I will do a ram test and try to reflow the cpu,replacing is not worth I guess because of the price of the cpu you probably can buy another same unit used.

[d_ne0]

So I did memtestst86 on mb.Enabled all cores in bios and fired up the test.Memtest recognized cpu with all 4 cores and reported 8 cpus (threads) and said using 4 of them for test so 2 cores.The test passed!to my surprise without cpu crashing.I saw temp on cpu rising on some test to 85°C despite the fan was set in bios to 100%before test.So the test ran with all cores enabled,still when trying to boot in windows,it works only with 1 core enabled.Next is to try a reflow.

[DynaxSC]

An idea, disconnect power and CMOS battery for 2-3 minutes. Sometimes only disconnecting the battery can make a full CMOS reset. If chipset goes into an undefined state, normal CMOS reset might not be fully working.

BTW a failure of the chipset is much more probable compared to CPU failure. So the PCH can also be the reason for such strange errors.

If this not help maybe it would be a good idea to give a last chance and reinstall Windows before reflow. The OS is by now so complicated, that it might be it has issues with non-typical hardware. Maybe try Windows 11 ?