[SOLVED] GPU gone bad?? Game crash and fans fullspeed...

soRa_21

Prominent
Aug 10, 2019
47
1
535
Hi all. I have rx580 4gb armor oc for about 2 years now. It worked amazing, no issues etc. Temps were decent since i have good airflow and good case. I even undervolted to 1.050 stable and set custom fan curve and temps were even better. Few days ago i noticed that my temps got higher for some reason. Everything was the same, room temp was fine etc, all of a sudden it started to go from 68-69 full load to 78-79 full load and it was undervolted. First i thought msi afterburner crashed and my fan curve and undervolt didnt work but everything was fine. Then after a day i was playing fortnite and my game crashed. My screen was black and gpu fans went full speed. I had to shut down pc and start it again and it worked fine. I uninstalled driver with ddu and reinstalled them and its still happening. I have no idea what is causing this. I tried with other games and it only happens with high intensive games when its under 90-100% load such as gta v, warzone etc. Sometimes it works fine for a few hours and sometimes it crashes after 5min. I'm starting to suspect that gpu has gone bad but i'm not an expert on this. It might help, i changed thermal paste few weeks ago. Since it was quarantine i used some cheap paste to last until i buy something high end. Maybe chepa paste is causing issues? It worked fine untill now.

Note: I changed paste a lot of times on gpu's and cpu's so it wasnt some kind of rookie mistake where i damaged something.

If anyone has some kind of advice or solution i would be very thankful. I really dont want to spend money on a new gpu if i dont have to. Also sorry if there are any msitakes in my writing cause its not my native language.
 
Solution
I didnt do any overclock. Card is running stock clocks. I also already did what you said and issue keeps happening. Tried heaven and 3d mark, temps go to around 78-82 with everething on stock and more aggressive curve. Before it got to 75 max with no undervolt and 67-68 with undervolt.
Then it's likely the paste has used up its lifespan. Try opening it up and repasting it even if it's the cheap paste.

soRa_21

Prominent
Aug 10, 2019
47
1
535
CPU: r3 2200g
RAM: Adata 2x4gb 2400mhz
Mobo: a320 hdv r.4
PSU: modular sama armor 650w 80+ gold
SSD: PNY 480gb
Running: windows 10 pro 64bit.

Pc worked like a charm untill few days ago. My guess was gpu.
 

soRa_21

Prominent
Aug 10, 2019
47
1
535
I didnt do any overclock. Card is running stock clocks. I also already did what you said and issue keeps happening. Tried heaven and 3d mark, temps go to around 78-82 with everething on stock and more aggressive curve. Before it got to 75 max with no undervolt and 67-68 with undervolt.
 

poorbugger

Distinguished
I didnt do any overclock. Card is running stock clocks. I also already did what you said and issue keeps happening. Tried heaven and 3d mark, temps go to around 78-82 with everething on stock and more aggressive curve. Before it got to 75 max with no undervolt and 67-68 with undervolt.
Then it's likely the paste has used up its lifespan. Try opening it up and repasting it even if it's the cheap paste.
 
Solution

soRa_21

Prominent
Aug 10, 2019
47
1
535
I will thanks. I'll pick up some high end paste and try that. I did notice that heatsink isnt really touching vrms on the board at the end, where the 6pin is. There is like maybe 1-2mm gap when i press it. Could that be potentially causing the issue?