I think my 980 Ti may be going bad. Help please.

[Jumpem]

I was playing The Grim Adventures of Van Helsing and the screen would turn a solid color (white, black, yellow, pink, etc.) just after the intro screens. A small number of times I was able to play for half a minute before it would happen. I have to power the PC off, and back on, in order to recover.

Diablo 3 plays fine. Fallout 4 is also doing it right when loading a game.

How else do I verify that it is the card? I am really out of the loop on utilities.

Thanks.

 

[indydude345]

I myself had a 980Ti and I started having issues with it before it died. Not saying yours is going to die, mine was because of user error, for another discussion. Anyways, older cards after a lot of usage start to need more voltage to maintain their clocks. Degrading circuits and such. (@StefanR5R gave me this info for my 980.) It's really a shame but that may be what is happening to you. Try increasing the voltage and see if the problem persists.

 

[UsandThem]

You can always try it in a different PCIe slot to rule that out, and possibly another PC if you have friends/family willing to let you do that. The only other thing I would look at would be your PSU. Many times when they begin their slow death spiral, they start causing odd issues like this one.

 

[bbhaag]

Have you verified the fans are spinning to keep it cool? I know that may seem like a "duh" suggestion but I've noticed on my RX480 that when it gets hot I start to see video corruption and blackouts similar to what you are experiencing.

 

[StefanR5R]

older cards after a lot of usage start to need more voltage to maintain their clocks. Degrading circuits and such. (@StefanR5R gave me this info for my 980.)

Well, I don't have direct experience with 9-series cards though; also I don't game. (the other discussion was in the context of GPGPU computing.) Also, what I probably haven't expressed is that I would lean more towards lowering clocks (or at least stepping down on overclocks, including factory overclocks) rather than dialing up the voltage, in order to try to regain stability.

I was playing The Grim Adventures of Van Helsing and the screen would turn a solid color (white, black, yellow, pink, etc.) just after the intro screens. A small number of times I was able to play for half a minute before it would happen.

I wonder, couldn't such things also be caused by a bad driver update?

 

[Jumpem]

I myself had a 980Ti and I started having issues with it before it died. Not saying yours is going to die, mine was because of user error, for another discussion. Anyways, older cards after a lot of usage start to need more voltage to maintain their clocks. Degrading circuits and such. (@StefanR5R gave me this info for my 980.) It's really a shame but that may be what is happening to you. Try increasing the voltage and see if the problem persists.

Interesting. I wouldn't even know to increase the voltage. I settled for stability and stopped overclocking a long time ago.

 

[Jumpem]

You can always try it in a different PCIe slot to rule that out, and possibly another PC if you have friends/family willing to let you do that. The only other thing I would look at would be your PSU. Many times when they begin their slow death spiral, they start causing odd issues like this one.

I could try the other PCIe slot. I don't have spare hardware or anyone else nearby to use to troubleshoot. My PSU is a Seasonic SS-660XP2 so I imagine it wouldn't be an issue. It seems the only way to verify these things is to buy parts locally just for testing, and then return them. That doesn't seem like a good thing to do though.

 

[Jumpem]

Have you verified the fans are spinning to keep it cool? I know that may seem like a "duh" suggestion but I've noticed on my RX480 that when it gets hot I start to see video corruption and blackouts similar to what you are experiencing.

I have not physically checked that they are. According to GPU-Z temperatures never got that high just before it would freeze.

 

[Jumpem]

I wonder, couldn't such things also be caused by a bad driver update?

I updated the drivers and chose a clean install. It didn't make a difference.

 

[UsandThem]

I could try the other PCIe slot. I don't have spare hardware or anyone else nearby to use to troubleshoot. My PSU is a Seasonic SS-660XP2 so I imagine it wouldn't be an issue. It seems the only way to verify these things is to buy parts locally just for testing, and then return them. That doesn't seem like a good thing to do though.

If that's the case, maybe you have a local PC repair shop near you where they could test it for you for a few bucks.

Even as good as Seasonic's quality control is, they still produce some duds every once in a while.

 

[StefanR5R]

I wonder, couldn't such things also be caused by a bad driver update?

I updated the drivers and chose a clean install. It didn't make a difference.

Actually I was thinking of a regression in a new driver version. GPU driver regressions happen every now and then. I am not sure at all though whether the issue which you describe could be caused by the driver in the first place. But still, just one game being affected makes me suspicious of a software bug.

The hard part may now be to recall which driver version was installed before the problem occurred first.