gpu PC keeps loosing signal if the GPU is put under load

[SurpriseApricot]

Hi Everyone

So I'm kinda at my wits end.

My PC looses video signal ( the display goes black, sometimes thje screen just goes black once, other times it does it a few times in quick succession before completely going black, with the only resoution being to restart my pc) if i try to put any sort of load on the GPU, what ever was open that caused strain on my gpu also closes

does been happening with any game  I play as well as when trying to test via Kombustor.

 

I have tried re-downloading my drivers and it hasn't resolved the problem,

I have also tried removing my GPU and just running my display from my CPU's display which seems to work.

 

is there anyway to determine what piece of hardware is going?

 

Windows 10 and Windows 11 both give this problem

i7 8700

2 x 16 Corsair Vengence 3200

MSI Armor GTX1080

Antec EAG Pro 750W

 

[FrowningHippo]

You should not put your load on your GPU. It may cause shortages. Especially if it's just out of the washingmachine./s

 

Have you checked the Windows Logs to see if it displays any errors at the time the GPU loses signal? Those errors might help identifying the problem. It could also be related to the GPU temperature. Have you checked if it gets to hot? No overclocks enabled? 

[SurpriseApricot]

but how else are the fans supposed to dry the laundry...

 

uhm i haven't

honestly don't know what to make of it.

 

temps seemed fine, and no overclock I've even tried a under-clock but it still gives me the same problem.

[FrostX]

"I have also tried removing my GPU and just running my display from my CPU's display which seems to work."

 

That immediatly leaves the GPU. When its under load = pulls much more power. This means we have three potential points of failure; The power supply failing to supply correct wattage (unlikely as in that scenario the GPU will downclock to compensate), GPU die overheating (Solved by new Thermal Paste, I would check temps are safe) and finally, failing VRM on the GPU itself (worst case scenario.)

 

To eliminate the last option, run the system with a different PSU or try it in a friend's build to see if the problem persists.

[SurpriseApricot]

Thank you, I'll see if I can find someone who would be able to help me out, 

I recently moved to a new city and I don't know to many PC gamers yet, 

 

Is putting new thermal paste on a gpu easy? I've never done it before I'm hoping to try every possible solution before having to buy a new part. 

[FrowningHippo]

12 hours ago, SurpriseApricot said:

but how else are the fans supposed to dry the laundry...

 

uhm i haven't

honestly don't know what to make of it.

 

temps seemed fine, and no overclock I've even tried a under-clock but it still gives me the same problem.

Ok. So at least we know the driver is crashing. The logs below the one you have selected are from the NVIDIA driver. Could you share the other error screens as wel? Not all 5 the nvlddmkm. Just one should be enough.

[SurpriseApricot]

Hi Sorry for only responding now
It has been a busy weekend, getting some stuff sorted, I attached the driver error log,
since I had done a fresh install before the last fail might it be that my SSD is causing the error?

[SurpriseApricot]

http://speccy.piriform.com/results/yhYsgag4yNahQjbOIg5Gut2

 

hi i dont know if this might help someone

 

[SurpriseApricot]

I see the Gpu seems to only have 4GB of Vram? 

 

[SurpriseApricot]

Hi so i tried DDU but now it is so bad there is no video output when i turn on the pc with the GPU in

 

[stonymason69]

The random errors and deteriorating stability seem like RAM type errors to me, so I'd hazard a guess that your VRAM might be dying. Either way, if your system can't reliably get an output with the card anymore then the only thing you can do is test another card on it or test this card on another computer.

 

If you can make the card output then testing the VRAM specifically to try inducing a crash might be able to tell you something.  Temps seem unlikely to me and the fact that it still works at any level means that the GPU is probably ok, which leaves board components of some sort.

[SurpriseApricot]

Hey sorry for only responding today
but wont the fact that it works without the GPU plugged in not eliminate some of the other parts?

[Whatisthis]

Do what stony said. Your graphics card ram is the most probable cause. Test the card in another system or take it to a repair shop to see if it replicates the error.