[SOLVED] Did I kill my GPU?

Nov 29, 2020
4
0
10
Hi guys,

Last week I transferred my mobo and gpu to a new case. During the process, somehow the motherboard died (confirmed by a local pc repair shop) so I got a replacement and now it's working fine. But just a day before i got the new mobo, I stupidly tried to deshroud and reapply thermal paste for the gigabyte 3080 gaming oc. The gpu now runs but everytime I try to run a game (or push the GPU somewhat) it will just crash. I put my old 980Ti back and it seems to run games fine meaning that the motherboard is OK. I did the render test (putting the card the highest power state) and it immediately crashed as well

What could be the problem and what should I do now? It's running and showing display etc but it just crashes

Thanks
 
Why would you try to replace the thermal paste on a brand new graphics card? This makes literally zero sense. Thermal paste, regardless of what it is going on, doesn't generally need to be replaced for YEARS, unless the heatsink had to be removed for some other reason anyhow.

So yes, you likely DID kill your graphics card. It worked before, it doesn't now, seems pretty conclusive to me.

Your best bet would be to simply take it back apart and figure out what you did wrong. Perhaps you applied too much thermal paste, or not enough, or forget a step in the process, or damaged one of the components on the graphics card PCB. Or, take it to a repair shop and let them take a look at it. Returning it might also be an option, but I wouldn't hold out much hope that they will honor the warranty since the damage will likely be found to be user caused. That's without even considering any moral imperatives.

Are you saying that when you do a render test with the 980 ti it crashed, or only with the 3080?
 
  • Like
Reactions: horsesample
Nov 29, 2020
4
0
10
thanks for the reply

I know I screwed up pretty badly; I thought of it because I already had the shroud removed and thought why not reapply the paste as I've read some stories about stock thermal paste being mediocre

it crashed when I did the render test with the 3080, I haven't really tried it with the 980ti
 
Ok, so what are your FULL hardware specifications including model or series numbers for everything including CPU, motherboard, power supply (Exact model or series), memory kit, CPU cooler, etc.?

Are you positive you didn't forget to reconnect the fans when you reassembled it? If the fans are not connected, the card may not function properly and may overheat. Might also simply not be installed correctly and there may be thermal issues. Have you monitored thermals to see what the temperatures are doing?
 
Nov 29, 2020
4
0
10
Ok, so what are your FULL hardware specifications including model or series numbers for everything including CPU, motherboard, power supply (Exact model or series), memory kit, CPU cooler, etc.?

Are you positive you didn't forget to reconnect the fans when you reassembled it? If the fans are not connected, the card may not function properly and may overheat. Might also simply not be installed correctly and there may be thermal issues. Have you monitored thermals to see what the temperatures are doing?

AMD 3600
Noctua U12A
Gigabye B550-I mobo
Corsair 2 x 16GB RAM
Corsair SF750 750W PSU

I tried connecting the fans back on with the shroud and it was the same.

The 980ti I'm running now is deshrouded, 2 x 120mm fans connected to a controller/fan curve

That said, I just checked and the 3080 seemed a tad warmer than the 980ti with the fans not spinning at around 38 deg C as opposed to 30 deg on the 980ti. It could possibly be the contact/thermal paste issue, and I hope it is because that's easier to fix than sending back to Gigabyte for repair