Question PC randomly shut down, restart, uninstall Amd drivers itself

[tapdig]

For months the PC shuts down by itself and uninstalls the video card driver. Sometimes it won't turn on, it makes 1 long 3 short beeps. It opens later, does not shut down for a max of 4 days. I tried the DDU method, bios update, XMP off/on, different rams, I removed the card many times, tried different driver versions. As a result, nothing has changed. I think the card is faulty but I can't be sure. I need your help.

Saphire Rx 580 Nitro+ 8Gb
Ryzen 5 3600
Corsair Vengeance RGB pro 16 gb (2x8) 3200 mhz ||| Kingston value (8+4)
Corsair vs650
GA-AB350M-D3V (Gigabyte)
win10,win11 (same result)

dump files: https://gofile.io/d/LKLIbx

 

[tapdig]

Ok. So I'd first try disabling the automatic driver updates by following the steps in the guide that I linked, and see if that stops it from doing that.

I changed the PSU with the other PC in my room. No problems with my Corsair. Not in the other. So my main problem is not the PSU. I did memtest, I already tried other rams. I did GPU tests, everything is normal there. Only the motherboard remains. So how can I test it, where could it be broken? I have so many questions in my mind.

 

[Darkbreeze]

No, PSU can't do anything except supply power. Did you say you tried this graphics card in another system, and by that I mean put it in there and run it through some use, few restarts, etc., and did NOT have the same problems with it?

Because right now it's looking to me like the problem is pointing towards the graphics card. Sure, could definitely be a motherboard issue but every time I've seen these problems with being recognized and removed in the past it's generally been related to the card itself.

Also, check the user manual and such on your card. Some of these older AMD graphics cards had buttons on them to change them from UEFI compatible cards to legacy non-UEFI cards. My Sapphire R9 290x had this and if the button was inadvertently changed it would cause problems with detection since I had CSM disabled in the BIOS.

 

[tapdig]

No, PSU can't do anything except supply power.

I've never done an OCCD test before. I finally found that the source of the problem was the GPU. The result is the same with different tests with different hardware. GPU has a problem. This is my third AMD graphics card that just crashed out of nowhere. If anyone has the same problem, I would recommend checking the GPU. A topic can be closed.

 

[Darkbreeze]

See my last post prior to you making that post. I was likely typing while you were typing. Specifically where I said

No, PSU can't do anything except supply power. Did you say you tried this graphics card in another system, and by that I mean put it in there and run it through some use, few restarts, etc., and did NOT have the same problems with it?

Because right now it's looking to me like the problem is pointing towards the graphics card.

 

[tapdig]

See my last post prior to you making that post. I was likely typing while you were typing. Specifically where I said

There is only compute mode, it is not turned on either. Unfortunately, it's not a good time to buy a new GPU. As far as I have experienced and researched from my old AMD cards, it seems like it may be a VRAM problem. But I can't fix it or test it. Thus, a new GPU is necessary.