Graphics Cards

[berrigan7563]

Monitor randomly loses signal during gaming sessions, seemingly out of nowhere and then continues but happening more easily (i.e. crashes in-game after gaming for couple of hours then doesn't let me get to the menu without crashing). Happened once this week and a couple of times last week. Its been a few days since the last it happened. Any ideas as to what the issue might be whether it might be software / hardware or how to find out would be much appreciated. Thank you in advance.

Attempted Solutions:

1. Factory reset the PC multiple times.
2. Fully uninstalled and reinstalled GPU drivers with different versions.
3. Ran Memtest86 to check for RAM issues, which passed without errors.
4. Checked temperatures, which appeared normal and in safe range.
5. GPU properly seated and connected.

Issues Encountered:

1. Random no signal black screens during gaming sessions after a few hours, requiring manual shutdown as the system doesn't shut down automatically.
2. Inability to play the game after a black screen, black screens occur more easily / even earlier (at the main menu).
3. Instant black screen when initiating a GPU stress test after black screen (hours after can run at 100% for more than 10mins).
4. Disappearance of the GPU from AMD software and appearing as disabled in Device Manager.
5. No critical errors found in Event Viewer prior to crashes.
6. Seemingly random but persistent issue.

PC Parts:

• 7900XTX SAPPHIRE PULSE
• B650E-I STRIX
• 32GB 6000MHZ GSKILL
• R7 7800X3D
• 2+4TB CRUCIAL M.2 
• 850W PLAT
• CUSTOM WATER COOLED (RUNNING FOR 3-4 MONTHS NO ISSUES)

EDIT: a few hours after the crash I re-ran after it crashed instantly when pressing run and Furmark has been running for 10 minutes now with no issues with gpu hotspot at 78c. The issue is very intermittent. 

[bluesadam]

Edit: Just realized your problem isn't a hard shutdown. In which case tells me it might be a ram issue and trust me, succeeding through memtest does not mean anything. For me, if I set XMP to mode 2 (3200mhz), I get game crashes and BSOD but with XMP profile 1 (3000mhz) I have no such problems whatsoever. Funnily, I can pass memtest just fine with either profile. So memtest does not paint the whole picture. Before trying a new PSU, I suggest trying a lower XMP profile or turning off XMP altogether just to see if it fixes the issue.

Since you state you have checked temps already, an insufficient or faulty PSU is what comes to my mind. I had a friend who had a PC shutdown if he played R6 Siege at unlocked framerates. Locking it to 60fps was okay but when unlocked pc would shutdown.

Turned out to be a PSU problem. When PC drew full power, the PSU was failing to deliver enough power and was shutting the PC down. If you have a different PSU you can try (maybe from a friend's pc), I would suggest giving it a try.

[berrigan7563]

That would make a lot of sense, I thought that might be the issue a while ago so I tried EXPO I instead of EXPO II or something but will give that a try and see if putting it on auto or off fixes the issue. I have an 850w platinum PSU that is fairly new and I run games capped so most times my GPU is 60-70% usage so I hope its not my PSU. Is there another way of testing RAM so that I can re-create the no-signal black screen? Thanks again

[berrigan7563]

Unfortunately setting RAM profile to manual which sets the values to default ,the closest to off, still led to a no signal black screen mid ranked game x_x. Weirdly it happened just as I respawned which has happened before at that exact same screen. Not sure what is significant about respawning in Apex but it seems to induce a no signal screen every once in a while. It also seems to happen after a few hours of gaming and then it keeps happening until I leave it a while.