[SOLVED] Computer screen turns off while gaming ?

[Tenet123]

Hello, I've recently upgraded my graphics card and there's an issue that I'm facing, it happened with the previous card as well.

My computer goes black while playing games. The LCD gives a no signal warning and then its off, I still hear the sound of the game for a while after which there's just noise. It doesn't start until I power it off and turn it on again. Another observation is that on newer games (Watch Dogs 3) it happens within 10-20 minutes of the gameplay, whereas on older games (CSGO), it takes about 40-50 mins to happen.

I've monitored CPU and GPU temperatures, GPU temperatures are around 85-86 C just before the crash and CPU temperatures are within the normal range.


What could be the possible issue? I've ruled out the card since it happened with the old card as well.

Specs
CPU- Core i5-10400
Board- ASRock Z490 Phantom Gaming 4
GPU- RTX 3070 8GB
RAM- 16GB DDR4
also it's a 4k monitor

 

[hedwar2011]

Have you tried removing and updating the GPU drivers?

 

[Tenet123]

Have you tried removing and updating the GPU drivers?

the drivers are updated

 

[hedwar2011]

Does it happen if you run any benchmarking programs?

 

[Tenet123]

Does it happen if you run any benchmarking programs?

I haven't tried that, will do.

 

[hedwar2011]

I haven't tried that, will do.

Might also try just letting it sit and see if it happens with the PC at idle.

 

[Tenet123]

Might also try just letting it sit and see if it happens with the PC at idle.

It doesn't happen on idle. But it did happen within 3 minutes of 3DMark's time spy

 

[Tenet123]

It doesn't happen on idle. But it did happen within 3 minutes of 3DMark's time spy

tried installing new windows, it's still the same.

 

[DSzymborski]

What's the exact PSU you're using? You left out that information and if temperatures are fine, it's by far the most likely cause of symptoms like these while gaming.

 

[Tenet123]

What's the exact PSU you're using? You left out that information and if temperatures are fine, it's by far the most likely cause of symptoms like these while gaming.

It's a cooler master silent pro semimodular 1000W PSU. I changed it when I upgraded the card, previous one was 600W. so I don't think its the PSU since it happened on the previous card and psu as well

 

[hedwar2011]

It doesn't happen on idle. But it did happen within 3 minutes of 3DMark's time spy

Is the GPU new or did you buy it used?

 

[Tenet123]

Is the GPU new or did you buy it used?

it's new

 

[hedwar2011]

It's a cooler master silent pro semimodular 1000W PSU. I changed it when I upgraded the card, previous one was 600W. so I don't think its the PSU since it happened on the previous card and psu as well

I just realized that you said it did this on the previous card and PSU as well. There has to be a common item in there somewhere that is either getting re-added, etc. when you setup the new setup from the old setup or that has been involved all along that is causing it. Figuring out what that variable is may be harder than you think.

For kicks and giggles, try disabling your AV software and then either try playing a game again or run 3dmark again and see if it happens again.

 

[Tenet123]

I just realized that you said it did this on the previous card and PSU as well. There has to be a common item in there somewhere that is either getting re-added, etc. when you setup the new setup from the old setup or that has been involved all along that is causing it. Figuring out what that variable is may be harder than you think.

For kicks and giggles, try disabling your AV software and then either try playing a game again or run 3dmark again and see if it happens again.

Some additional information:

I did a GPU stress test, the display went off in like 20 seconds, tried it again with the case open, same result.
How can I know for sure if its a PSU problem or an overheating problem?

I'm gonna use the GTX 1060 with the same PSU and board, Do a stress test and then see what happens.

 

[hedwar2011]

Some additional information:

I did a GPU stress test, the display went off in like 20 seconds, tried it again with the case open, same result.
How can I know for sure if its a PSU problem or an overheating problem?

I'm gonna use the GTX 1060 with the same PSU and board, Do a stress test and then see what happens.

The GPU temps you initially reported aren't necessarily GPU melting temperatures but could be a source of a problem potentially.

Let us know how it goes with the other GPU.

 

[Tenet123]

The GPU temps you initially reported aren't necessarily GPU melting temperatures but could be a source of a problem potentially.

Let us know how it goes with the other GPU.

The other gpu worked well. No black screen.

my RTX 3070 temperatures are going till 88,89 C and I think that's a lot, how do I reduce them?

secondly, it seems that this (https://linustechtips.com/topic/1473076-black-screen-on-rtx-3070-high-setting/) is the problem since I have the exact same Tuf gaming card

 

[hedwar2011]

You can try undervolting the card like the link suggests but do so at your own risk.

 

[Karadjgne]

Sounds like the cards are overheating. The 'GPU' is the actual processor itself, and for many cards that's the only registered temp. However, that's not including the VRAM and VRM's, which are not part of the 'GPU' as such, but do sit under the heatsink and generally a different area of the fan coverage.

If there's little actual airflow over the gpu area, it'll sit in its own heat, getting hotter by the second with no relief. It's entirely possible for either the VRM's or the VRAM to exceed 99°C+ and the 'GPU' never see north of 85°C.

Black screens or loss of signal but retaining audio is almost always a VRAM issue, but can also be a driver conflict.

 

[Tenet123]

Sounds like the cards are overheating. The 'GPU' is the actual processor itself, and for many cards that's the only registered temp. However, that's not including the VRAM and VRM's, which are not part of the 'GPU' as such, but do sit under the heatsink and generally a different area of the fan coverage.

If there's little actual airflow over the gpu area, it'll sit in its own heat, getting hotter by the second with no relief. It's entirely possible for either the VRM's or the VRAM to exceed 99°C+ and the 'GPU' never see north of 85°C.

Black screens or loss of signal but retaining audio is almost always a VRAM issue, but can also be a driver conflict.

Solved. It was after all an overheating issue (85-89°C). I changed the thermal paste, it was all dried up but it still didn't work. Turns out the paste was bad. Then I contacted a technician who applied a new "better" paste which lowered the temps down to 70s under 100% load. I also got 3 more fans installed, and then the temperature went down in the 60s. And the problem is resolved now. Thanks everyone for your input.