Question BSOD, System Reboots, Driver Crashes

[kobitz]

Hello,

About a year and a half ago I built this system:

Fractal Meshify C
MSI B450 Tomahawk MAX
Ryzen 5 3600 (non-x)
16GB G-Skill 3600Mhz
Gigabyte 5700XT Gaming OC
650W EVGA G5 PSU

The system is running it's third fresh install of Windows 10 with newest AMD drivers (without their GUI).

Originally, it had the stock wraith prism cooler and some random case fans. I suspected my temps may have been an issue (or at least part of the problem), so the system now has 2x140mm noctua fans in the front, 1x 120mm noctua fan in the rear, and a noctua NH-D15 tower cooler (new thermal paste, temps are now great at idle and both under load).

Since i've built the system, I have had crashes. At the time I built it, I played mainly warzone and chocked the crashes up to warzone being the issue. Since then, it has crashed in nearly every game i've played including Destiny 2, Apex, and most recently the New World closed beta and open betas. The crashes in the New World betas have been erratic, sometimes crashing graphics drivers (requiring a system restart to be able to run the game again), BSOD (black) with the windows "ba-doop" noise as if a device had been disconnected then reconnected, as well as a single occurrence of a hard system power off with my GPU disabled on reboot. This makes me strongly lean towards the GPU being the issue, but i'm unsure of how to nail that down.

Here are things I've done:

Reinstalled windows multiple times
Uninstalled and reinstalled drivers, including older drivers (Current install was wiped with DDU before installing new drivers)
Updated motherboard BIOS
Updated GPU BIOS
Increased GPU power limit, underclocked, slightly overclocked, increased fan curve, set fan speed manually at 80 through afterburner
Changed Windows power plan
Ran memtest86, twice -- original settings, no failures
Reseated RAM
Reseated GPU and PCI power connectors on card side

Since building the system I've read that the 5700XTs are sort of notorious for crashing, but I'd like to nail down the issue. Unfortunately, I don't have another system or card (or access to either, at the moment) to test the scenario. What can I do?

Thanks in advance

Edit 1: 9/16/21

System Reliability History shows Live Kernel Events 141, 144 and 117.

 

[Colif]

you getting BSOD or just black screens?

how long you had problems for?

there are benchmarks but they only really used to see if it crashes during them.

when it crashes, what shows in system reliability history? does it show driver is d/c constantly?

 

[kobitz]

you getting BSOD or just black screens?

how long you had problems for?

there are benchmarks but they only really used to see if it crashes during them.

when it crashes, what shows in system reliability history? does it show driver is d/c constantly?

I've had BSODs (blue) on occasion, but mostly black screens. Some accompanied by the device connect/disconnect noise but mostly just black screens with audio continuing to work (I can hear the game and the discord call, even talk and people are able to hear me, but no display. Sometimes I can alt-tab out of the application and get to the desktop to reboot, sometimes I can't.) On occasion, the drivers have crashed and the rest of the system continues to work fine. I've had some form of issue since the build (DEV errors in warzone, which is why I blamed warzone to begin with, then began to see problems across most if not all games).

System Reliability (which I did not know existed; I scanned event viewer and while it showed some critical errors it did not provide any useful information to me personally) shows a multitude of hardware errors, all LiveKernelEvent 141, 144, 117.