Question Display randomly dying. Why and how to troubleshoot?

[A Grandma]

This problem has persisted for quite a while and it's extremely annoying.

The hardware:

• ZOTAC GAMING GeForce RTX 3070 Twin Edge OC
• AMD Ryzen 7 5800X
• Corsair Vengeance RGB Pro SL DDR4 (Part number: CMH32GX4M2D3600C18)
• ROG STRIX B550-A GAMING
• Corsair RM750x (2018) (Power supply in case you need it)
• Samsung SSD 970 EVO Plus 1TB (Windows installed here)
• ST2000DM008-2FR102 (Secondary HDD)
• Asus VG248QG with G-Sync (Monitor attached directly to GPU thru display port)

The problem:
Screen randomly goes black (monitor reports no input) and the display won't return back unless I restart my PC by performing a hard shut-down using the power button.
This happens randomly, especially when running games (apparently the game being run doesn't matter; display can still die in very light games). Display sometimes does when sharing my screen.

What I've tried:

• Increasing the fan curve to keep the GPU cooler. Display would sometimes die even when temperatures are well below 65C.
• Win + Ctrl + Shift + B. The beep is still heard but it does nothing.
• Ran Heaven benchmark on the highest setting to stress-test for 20-30min intervals. Temperatures would max out at around 70C but the display wouldn't die.
• Ran various other benchmarks; same result as Heaven benchmark.
• Re-seating the GPU.
• Installing the GPU in the other PCI slot.
• Updated Nvidia drivers.
• Tried installing various older drivers.

Edit:
I'd really appreciate any help on this. I'm afraid of doing any work or playing any competitive games because of this. Methods of troubleshooting are very much welcomed as well.

 

[Koekieezz]

This problem has persisted for quite a while and it's extremely annoying.

The hardware:

• ZOTAC GAMING GeForce RTX 3070 Twin Edge OC
• AMD Ryzen 7 5800X
• Corsair Vengeance RGB Pro SL DDR4 (Part number: CMH32GX4M2D3600C18)
• ROG STRIX B550-A GAMING
• Corsair RM750x (2018) (Power supply in case you need it)
• Samsung SSD 970 EVO Plus 1TB (Windows installed here)
• ST2000DM008-2FR102 (Secondary HDD)
• Asus VG248QG with G-Sync (Monitor attached directly to GPU thru display port)

The problem:
Screen randomly goes black (monitor reports no input) and the display won't return back unless I restart my PC by performing a hard shut-down using the power button.
This happens randomly, especially when running games (apparently the game being run doesn't matter; display can still die in very light games). Display sometimes does when sharing my screen.

What I've tried:

• Increasing the fan curve to keep the GPU cooler. Display would sometimes die even when temperatures are well below 65C.
• Win + Ctrl + Shift + B. The beep is still heard but it does nothing.
• Ran Heaven benchmark on the highest setting to stress-test for 20-30min intervals. Temperatures would max out at around 70C but the display wouldn't die.
• Ran various other benchmarks; same result as Heaven benchmark.
• Re-seating the GPU.
• Installing the GPU in the other PCI slot.
• Updated Nvidia drivers.
• Tried installing various older drivers.

Edit:
I'd really appreciate any help on this. I'm afraid of doing any work or playing any competitive games because of this. Methods of troubleshooting are very much welcomed as well.

follow this step by step in order (read till end):

• Try to boot normally to windows with the procedure you did
• Disconnect from internet
• Uninstall every gpu driver using DDU (clean and do not restart, also tick all option in Nvidia option in the ddu settings).
• Uninstall all the processors (is a must, should be 16 on yours since it's 16 threads, also when it asks for restart, click on no and keep uninstalling all processors) on device manager like this:
  
  
  
  
  
• Uninstall AMD Chipset Software in control panel (if there is none, skip it.)
  
  
• reboot the PC to bios, disable AMD fTPM and secure boot (if enabled by default), save and exit, go to bios again, flash to the latest bios (do it again if you're currently using the latest bios) ver 2803 the one with agesa 1.2.0.7, go to bios after finished updating, then load default or optimized settings, enable fastest xmp profile, then save and exit.
  
  Optional: disable CSM, enable Above 4G Decoding and Resizable bar option (these 2 options wont be available unless CSM is disabled). If gpu supports, why not XD
  
  
• if successful, boot up to windows and install the latest Chipset driver (should be ver 4.08.xx as the time i write this), then reboot.
  
  
• Install the latest nvidia driver, reboot, and then connect to internet.
  
  *do this all offline until reboot after installing chipset driver, put ram on slot A2 and B2 (slot 2 and 4) if you only use 2 ram sticks, also you may reboot to bios after all of this to set the XMP (and previous settings you did). Download needed files (highlighted word) before doing step 1, do the step by orders.
  
  
• Run cmd as admin, then do chkdsk /x /f /r, after that do sfc /scannow
  
  
• And check windows update if there is any and install them (except optional update).
  
  
• Make sure the psu connected to the gpu is 1 pcie cable per 1 slot (use main cable, not the branches/split) like this:
  
  

 

[A Grandma]

I have followed your instruction step-by-step, in order, and without any problems. What I did next:

• Ran Heaven benchmark for about 15mins with no problems at all. Temps reached a max of 72-74C.
• Ran CS:GO and played an unranked long match, again, with no problems at all. Temps were around 56-58C throughout the whole game.
• Continued on to play a ranked long match. Display died within the first few rounds of the match.
• Pressed and held the off button to force a shut-down and rebooted the PC. The boot was successful.
• Started up CS:GO again to reconnect to the match, but display died soon again. (Here's what that looks like)

This shut-down and rebooting process continued on for a while, until it suddenly couldn't boot. POST light was stuck on "VGA."

• Tried physically reinstalling the GPU to the second PCI slot. The PC didn't want to boot with the POST light stuck on "VGA" again.
• Reinstalled the GPU to the first PCI slot (where it was originally seated), and PC booted successfully again.
• Ran Heaven benchmark again with no problems at all.

Is there any possibility that it's just a display port cable problem? Is this really a fault GPU problem? I'm starting to consider bringing the GPU (or the whole PC) to a professional repair service, but I have never done that before and I don't really know where to look, or who to ask to diagnose and fix a GPU (or the whole PC).

Edit:
Sometimes, this would happen during reboots. Not too sure if this is related to the big issue, or if it's just a DisplayPort cable problem.

 

[Koekieezz]

I have followed your instruction step-by-step, in order, and without any problems. What I did next:

• Ran Heaven benchmark for about 15mins with no problems at all. Temps reached a max of 72-74C.
• Ran CS:GO and played an unranked long match, again, with no problems at all. Temps were around 56-58C throughout the whole game.
• Continued on to play a ranked long match. Display died within the first few rounds of the match.
• Pressed and held the off button to force a shut-down and rebooted the PC. The boot was successful.
• Started up CS:GO again to reconnect to the match, but display died soon again. (Here's what that looks like)

This shut-down and rebooting process continued on for a while, until it suddenly couldn't boot. POST light was stuck on "VGA."

• Tried physically reinstalling the GPU to the second PCI slot. The PC didn't want to boot with the POST light stuck on "VGA" again.
• Reinstalled the GPU to the first PCI slot (where it was originally seated), and PC booted successfully again.
• Ran Heaven benchmark again with no problems at all.

Is there any possibility that it's just a display port cable problem? Is this really a fault GPU problem? I'm starting to consider bringing the GPU (or the whole PC) to a professional repair service, but I have never done that before and I don't really know where to look, or who to ask to diagnose and fix a GPU (or the whole PC).

Edit:
Sometimes, this would happen during reboots. Not too sure if this is related to the big issue, or if it's just a DisplayPort cable problem.

well if your gpu warranty is still appliable, i would prefer RMA it. but before that the last thing to try is to use the gpu on somewhere else a whole day use, say if you have a 2nd pc or a friend with good psu on his pc, if it didnt have any problem then it might be your pc that causing it, but if the problem presists, then its gpu fault

also try to reflash the gpu bios, but before that could you send a gpuz screenshot here? just to make sure i could find the correct one from techpowerup site. or if you know how to use nvflash and could find the same exact vbios for your card then try it yourself.

 

[A Grandma]

also try to reflash the gpu bios, but before that could you send a gpuz screenshot here?

Here is a screenshot of my GPU-Z. I don't really know how to flash GPU BIOS nor use NVFlash, but I'll take a look on that meanwhile. Also, on my GPU's warranty, I'm not too sure so I'm gonna have to dig more on that. I'll give updates whenever I find something new.

 

[Koekieezz]

Here is a screenshot of my GPU-Z. I don't really know how to flash GPU BIOS nor use NVFlash, but I'll take a look on that meanwhile. Also, on my GPU's warranty, I'm not too sure so I'm gonna have to dig more on that. I'll give updates whenever I find something new.

try to learn how to use nvflash
here try this bios