My desktop is shutting down frequently with no specific condition

I hope you've done the obvious, which is clean your fans and filters from dust.

Have you stresstested both VGA and processor?

It could also be some volatile memory (RAM or VRAM) chunk that's corrupted and fucks up everything when it's accessed by the system, which happens completely at random as far as an user is concerned.

Have some monitoring software log temperatures to a file, so you can see what was going on when you reboot after a new crash.

Godspeed.

>>342582
Actually, the shutdown become frequent after I cleaned it(I used a brush to take the dust off).The heavyloader stress test gave me nothing on both cpu or vga, but I have no idea how long the test should be. I just activated the temp log on speed fan as you said, but I don't think it's overheating. Also, no BSOD, in the event viewer it shows a kerner 41 error everytime it shutdown. If it is the ram, shouldn't the memtest spotted it?

>>342588
>If it is the ram, shouldn't the memtest spotted it?
I don't think it tests VRAM. Could be wrong on that though.

A simple way to check if it's the GPU's (or the north bridge's...) fault is to physically disconnect it and run on integrated graphics for a while.

>>342597
>integrated graphics
... and I just checked that it's got none.

Sorry anon, my help can't go further.

>>342597
My MOBO do have integrated graphics, but I don't think it's the gpu. In the post I said the the gpu temp got lower but it was actually the CPU, sorry for that. Thank you, anon.