My 1080Ti has a memory failure causing visual artifacts and eventual gpu hang under load:
$ ./cuda_memtest --stress
[11/13/2020 10:53:39][0f3d62ba338f]:ERROR: the last 4 error addresses are: 0x7f1ee4178e48 0x7f1ee3135fa0 0x7f1ee2dd22e8 0x7f1ee61db7d8
[11/13/2020 10:53:39][0f3d62ba338f]:ERROR: 0th error, expected value=0x3be5a39152a03ca5, current value=0x3be5a39152a03ea5, diff=0x200
Any Linux GPU alternatives to BadRAM?
https://github.com/envytools/envytools/ looks interesting. I might try out nouveau first before playing with the bios. That said, https://www.phoronix.com/scan.php?page=news_item&px=Nouveau-2020-Early-Status does not sound too optimistic.
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!