Hey all! I've been grappling with a peculiar problem for the past few weeks that I just can't get to the bottom of. My system will randomly crash, sometimes while completely idle, other times when loading an intensive scene in a game, but also sometimes it'll work completely fine for hours. After checking event viewer I saw that there were two warning and errors that appear every time the system freezes/crashes like this. Either it's event 4101, Display "Display driver nvlddmkm stopped responding and has successfully recovered." and/or event 13 with a different exception each time. And just yesterday, i was rendering something in blender when cycles crashed (but not blender itself) and i got the error "Illegal address in CUDA queue copy_from_device". Furthermore, ocasionally upon crashing windows will generate a minidump file which I've attached below, which seems to state "Arg1: 0000000000000002, The driver failed upon the submission of a command."
At first I believed that this was due to me using an experimental BIOS on my MSI B350 Tomahawk to use a R7 5800x as this issue began to occur after I changed the CPU from a 2600x, however upon buying a different board (MSI b550 Tomahawk) the issue persisted. Furthermore, it continues to crash with a different set of RAM, and yes, I have updated drivers using DDU first to 511.79 and again to 512.15 with no change.
If anyone has any insight to these issues I would be highly grateful for any help.
The system specs are:
Ryzen 7 5800X
2x16gb 3200mhz CL16 trident Z RGB
MSI B550 Tomahawk
ASUS ROG STRIX GTX 1080 Ti GAMING OC
WinDBG of minidump: https://pastebin.com/7NxNB8BN
minidump: https://drive.google.com/file/d/14Hl2DbTAFO0wCyK2KriDOEEfa8Z1Gsii/view?usp=sharing