GPU on it's way out?

[agnes]

I posted a little while ago that my wifes 1080ti started randomly crashing games to desktop. We swapped cards (she got my 2080 ti) and we tested, I could, with consistency, get her card to crash if I upped the power target from 100% to 105%. It would crash almost instantly. So I ran it at 95%, it stopped crashing, swapped back, it did it a couple of times so we dropped it to 90%. It's solid for now, but I have a funny feeling it will eventually show it's problem again.

The time she had my 2080 ti, we maxed the power target and stuck a max OC on that I knew it was stable for. It ran rock solid without a hiccup, so I doubt it's anything to do with the PSU/motherboard.

I've never experienced this, as the crash reports back as the driver stopped working and resumed. Temps are sub 80C, it can easily be replicated by increasing the power target.

So, is it on it's way out? Has anyone seen this and if so, how fast does it start to go until it's essentially a brick?

I can't RMA as I bought in the UK and now live in the US. We want to keep it as she'll be getting my 2080 ti when I get a 3080 ti (when/if they come out). But if that's not for a while, I'm tempted to just buy her a 2080 ti to do her until the new cards come out.

 

[agnes]

no one seen this before then?

 

[Panos]

I posted a little while ago that my wifes 1080ti started randomly crashing games to desktop. We swapped cards (she got my 2080 ti) and we tested, I could, with consistency, get her card to crash if I upped the power target from 100% to 105%. It would crash almost instantly. So I ran it at 95%, it stopped crashing, swapped back, it did it a couple of times so we dropped it to 90%. It's solid for now, but I have a funny feeling it will eventually show it's problem again.

The time she had my 2080 ti, we maxed the power target and stuck a max OC on that I knew it was stable for. It ran rock solid without a hiccup, so I doubt it's anything to do with the PSU/motherboard.

I've never experienced this, as the crash reports back as the driver stopped working and resumed. Temps are sub 80C, it can easily be replicated by increasing the power target.

So, is it on it's way out? Has anyone seen this and if so, how fast does it start to go until it's essentially a brick?

I can't RMA as I bought in the UK and now live in the US. We want to keep it as she'll be getting my 2080 ti when I get a 3080 ti (when/if they come out). But if that's not for a while, I'm tempted to just buy her a 2080 ti to do her until the new cards come out.

What are the temps, card precise model and PSUs involved? It might need a good heatsink cleaning, replacing thermal paste & pads and go back to working.

 

[agnes]

What are the temps, card precise model and PSUs involved? It might need a good heatsink cleaning, replacing thermal paste & pads and go back to working.

Thermals are fine, never goes above 80C, I replaced all the thermal pads and paste earlier this year. It's a EVGA 1080 ti SC2. PSU is a 750 Corsair (forget which one, but a good one). As mentioned, my EVGA 2080 ti XC Ultra with heavy overclocks was rock solid in her PC. Yet her card in my machine displayed the exact same symptoms. I'm convinced it's the card on it's way out. I'm just curious to see if others have had the same and if so, how fast did it **** it's self? We're down to 90% power target but I imagine soon we'll have to reduce if it starts crashing again.