3070 Random Black Screen when Underload

[Dizi]

Hello fellow Carbies,

A friend of mine is experiencing an odd issue with his current setup. When stress testing the machine, a GPU Stress Test does not cause the screen to go black, yet stress testing both GPU and CPU does cause the Black Screen. Software used for the street test is MSI Kombuster. The screen just goes black after about 5-15 min (this can vary when gaming), however the PC does not shut down. The pc will then automatically restart after about 1-2 min after the black screen has occurred.
The fixes we have thus far tried with no luck:

1. Reinstalling windows to eliminate any software issues.
2. Changed display cables (DP to HDMI).
3. Both Under-clocked and Over-Clocked the GPU.
4. Temps reach no higher than 85 Degrees.
5. Resetting the monitor settings seemed to help for a week, but now the issue still occurs.
6. We even Cleaned the system.

His current setup is below:

i5-9400
Asus Prime B360m-A
16 GB
Asus Dual 3070
Cooler Master 700w
1TB nVME
Dell S2719DGF Monitor

Another person on Reddit has the exact same issue, and you can review their post here -

https://www.reddit.com/r/pcgamingtechsupport/comments/ygnnko

Is anyone perhaps able to provide any input as to what could be causing this issue?

 

[Tapsumbong]

Case and number of fans? In the reddit post he mentions Hotspot of 105 which for gddr6 is pretty hot, that's more in line with 6x memory. What are your hotspot temps?

 

[Dizi]

Case and number of fans? In the reddit post he mentions Hotspot of 105 wChich for gddr6 is pretty hot, that's more in line with 6x memory. What are your hotspot temps?

Case is a Antec, not sure on model. 2 Fans in front, 2 on top and one behind.
HW Monitor Reads GPU Hotspots as below:

72.6 - 62.8 - 107.5

 

[goldfritter]

107.5

That is very hot. That is quite likely to be your problem.

What is the GPU core temp when the hotspot is 107.5? I.e. what is the difference between core and hotspot? If it's around 20 C or more then it could be a bad cooler mount.

Are the GPU fans running at 100% at this point? If not, try set a custom fan curve in Afterburner and see if that helps.

If they are at 100% then you urgently need to look at better case cooling.

 

[Dizi]

That is very hot. That is quite likely to be your problem.

What is the GPU core temp when the hotspot is 107.5? I.e. what is the difference between core and hotspot? If it's around 20 C or more then it could be a bad cooler mount.

Are the GPU fans running at 100% at this point? If not, try set a custom fan curve in Afterburner and see if that helps.

If they are at 100% then you urgently need to look at better case cooling.

Update on this. Your feedback helped a lot.
The side panel was opened, and the extra airflow has allowed for the hotspot to not go higher than 96 C, which has now seemed to stop the black screen.

The difference between the core temp and hotspot temp was 22.3 C, yet I am not sure what the fan speeds were at.
I am having a look at the PC this weekend, and will reposition fans, or even possibly use another case.

Thank you so much for the assistance!

 

[goldfritter]

Update on this. Your feedback helped a lot.
The side panel was opened, and the extra airflow has allowed for the hotspot to not go higher than 96 C, which has now seemed to stop the black screen.

The difference between the core temp and hotspot temp was 22.3 C, yet I am not sure what the fan speeds were at.
I am having a look at the PC this weekend, and will reposition fans, or even possibly use another case.

Thank you so much for the assistance!

Cool yeah that makes sense.

22.3 is probably borderline for a bad mount. Some people would re-paste with that difference, but I think if the card is still under warranty and you can get temps under control with better case cooling and/or a more aggressive fan curve, it's probably OK to leave it as-is.

 

[Dizi]

Cool yeah that makes sense.

22.3 is probably borderline for a bad mount. Some people would re-paste with that difference, but I think if the card is still under warranty and you can get temps under control with better case cooling and/or a more aggressive fan curve, it's probably OK to leave it as-is.

The card is still under warranty, so I am going to first see if I can get the Temps under control first.
Thank you for your assistance. It is highly appreciated!