Vega 64 keeps crashing, mostly BSOD but a few other crashes
I got a Gigabyte Gaming OC Vega 64 recently, and it's been problematic. First day, I plug it in, using two PCIe power connectors from the same rail, and it "crashed" a few hours into Assassin's Creed: Odyssey. The "crash" consisted of the screen going black, Windows not responding, the game audio still playing (or maybe just looping) and the fans on the graphics card shooting up to 100%. I had to hold the power button to turn the PC off. I went to Google. Thinking it was the PSU issue everyone talks about, I unplugged one of the 8 pin power connectors, and switched to using one from a different rail on my PSU.
Since doing this, I haven't gotten the 100% fan "crash", but I have gotten a lot of "THREAD_STUCK_IN_DEVICE_DRIVER" BSODs. All of them have ntoskrnl.exe in common, and the bug check codes always point to a graphics problem. I went back to using two 8 pin PCIE cables from the same rail, and still got the same BSOD.
I am on the latest Vbios, I've tried messing with drivers, I've used DDU, undervolt, power limit +50%, all that jazz. After my most recent crash, (about two hours ago as of this post) I tried setting HBM state 2 as minimum, and it hasn't crashed yet, but it's too soon to tell.
Windows 10 64
Ryzen 2600 at stock
MSI B350 Gaming Pro Carbon
2 x 8GB Memory at 2933 MHz
Enermax SGalaxy Evo 1250W PSU
Gigabyte Vega 64 Gaming OC
1. Asus MG279 via DP, 90hz, Freesync, 1440p
2. Asus MX279 via HDMI, 60hz, 1080p
3. Samsung T24B350 via HDMI, 60hz, 1080p
I've seen quite a few other people with similar issues, and I saw a comment from amdmatt saying that there might have been a bad batch of GPUs. If so, I'd be more than happy to open up my card and see when it came off the line.