Question Is my GPU on its last legs?

[kkan]

Hello,

One and a half years ago I purchased a used ASUS TUF 3080 10GB. Its been worked well. Two days ago I started crashing in Witcher 3. after about 30-50min of playtime I though it might be due to some memory leak within the game as it was the only game that I was crashing in. After few more crashes I went to the Event Viewer and noticed that Nvidia driver is causing the crash.

The description for Event ID 153 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Error occurred on GPUID: 800

Does anyone know what exactly is GPUID 800 or event ID 153?

I DDU my drivers and installed the latest ones, but I was still crashing, now in other games. I ran FurMark, Superposition and 3DMark stress test to try to replicate this behavior, but everything ran without issues. However if I got into more intensive game, I would crash - sometimes after 10 min, sometimes after 3 hours. No pattern whatsoever. I tried every potential fix. I decided to underclock my GPU to 1670Mhz Core Clock. So far, I haven't crashed yet, time will tell if helps. It seems that my GPU core is unstable and unable to work reliably on the factory clocks (not sure why the stress tests are not crashing it) I'm loosing about 15% of performance (Based on Superposition benchmark) with my underclock.

I filtered all errors from event logs for this year and it turns out I have been experiencing this as far as logs go, I just never thought of looking into them. Turns out every crash I had within a game was caused by this error. Is this behavior a sign that my GPU is giving out? Is there anything I can test/try?

Appreciate any help!

 

[mikemid]

try to lower the power to 80% and -100MHz of VRAM clock via MSI afterburner. and limit FPS to 60fps. may the GPU is the problem of not running stable

 

[Shay Green]

Hello,

One and a half years ago I purchased a used ASUS TUF 3080 10GB. Its been worked well. Two days ago I started crashing in Witcher 3. after about 30-50min of playtime I though it might be due to some memory leak within the game as it was the only game that I was crashing in. After few more crashes I went to the Event Viewer and noticed that Nvidia driver is causing the crash.

The description for Event ID 153 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Error occurred on GPUID: 800

Does anyone know what exactly is GPUID 800 or event ID 153?

I DDU my drivers and installed the latest ones, but I was still crashing, now in other games. I ran FurMark, Superposition and 3DMark stress test to try to replicate this behavior, but everything ran without issues. However if I got into more intensive game, I would crash - sometimes after 10 min, sometimes after 3 hours. No pattern whatsoever. I tried every potential fix. I decided to underclock my GPU to 1670Mhz Core Clock. So far, I haven't crashed yet, time will tell if helps. It seems that my GPU core is unstable and unable to work reliably on the factory clocks (not sure why the stress tests are not crashing it) I'm loosing about 15% of performance (Based on Superposition benchmark) with my underclock.

I filtered all errors from event logs for this year and it turns out I have been experiencing this as far as logs go, I just never thought of looking into them. Turns out every crash I had within a game was caused by this error. Is this behavior a sign that my GPU is giving out? Is there anything I can test/try?

Appreciate any help!

What PSU do you have? If it has been caused as long as logs go it could be the cause of a poor power delivery from your PSU

 

[kkan]

What PSU do you have? If it has been caused as long as logs go it could be the cause of a poor power delivery from your PSU

I have Thermaltake Toughpower GF1 850W 80+ Gold. I do not expect it being the issue as its a very good power supply and if it was a power delivery issue, you would expect for it to be far more consistent in its behaviour?

 

[SkyNetRising]

Is this behavior a sign that my GPU is giving out? Is there anything I can test/try?

What gpu temperatures are you getting. when crash occurs?

With prolonged exposure to elevated temperatures, thermal paste would dry out and
would need to be replaced.

What pc case are you using? Does it have proper ventilation?
Can you show a photo of your system with side panel removed?
(upload to imgur.com and post link)

 

[kkan]

To think about it. I do use a single cable that goes into 2 x 6 pin and 2 x 2 pin which make up two 8 pins. I will swap that out for a separate 8 pin cables.

What gpu temperatures are you getting. when crash occurs?

Around 70C I have Lian Li LANCOOL 216 with 7 fans so air flow is not an issue. Everything including GPU stays cool. So far, with the lower clocks, nothing is crashing.

 

[SkyNetRising]

Around 70C I have Lian Li LANCOOL 216 with 7 fans so air flow is not an issue.

7 fans seem excessive.
How are they oriented?
Have you followed proper intake/exhaust orientation for your fans?

 

[kkan]

Yes, they are 4 intakes and 3 out. They are all Noctua fans and work on low RPM for low noise operation. I don't believe this is the issue. 70C temp on a GPU under load is nothing special and the crashes would follow some pattern, instead of what I have been experiencing so crashes after 5 minutes or sometimes after 3-4 hours.

 

[kkan]

Update.

It was the 8 pin cable. I was using a single 8 pin to 2x 6 pin + 2x2 pin cables. I connected a second one and didn't see a crash in 6 hours.