I was hopeful this would help since we have the same card, but my s2idle was already [deep] and adding the parameter anyway didn’t help.
In case you didn’t downgrade yet, do you have any access to the OS after the system hangs? Sometimes spamming Esc or causing the nvidia driver to crash by turning displays off and on again lets me log in and use the system (with a crashed plasmashell) just enough to run dmesg
on a previously open terminal.
I’m asking because I get a CPU soft lockup [nvidia-sleep.sh:3805] upon sleep/resume, and I’m curious whether it’s possibly connected to the issue. My dmesg.