A lot of system crashes on Ubuntu 25. Help needed

Graphics / Linux Linux
,
1

Hello! I’ve recently bought a new PC, it has RTX 5060.

I experience frequent crashes, and those crashes, according to logs, are because of GPU/Driver problems.

The seller told me they stress tested the PC(on windows), so I’m unsure if this is a hardware or software issue(I use driver 580.95). Can someone help?

2

nvidia-bug-report.log.gz (399.1 KB)

Here’s the result of nvidia bug reporter

3

have you tried other driver versions? v575 is currently considered the most stable by users and v590 is the newest, which may contain some bug fixes.

Please be more specific so it’s easier for others knowing where to start from when viewing your logs ;-)

4

Thanks for the quick reply! I haven’t tried other drivers to be honest, will give it a shot.

I run journalctl --since “2025-12-20 00:00:00” --until “2025-12-21 23:59:59” --no-pager | grep -iE ‘nvidia.*error|nvidia.*fail|GPU.*error|drm.*nvidia.*ERROR|NVRM|Xid|modeset.*ERROR|RmInitAdapter’ and it showed my plenty of stuff, I assume important stuff is this

Dec 20 23:49:55 machine kernel: NVRM: GPU at PCI:0000:01:00: GPU-24de1a3f-4fa0-5ce0-604c-dc4a3ad4fcab
Dec 20 23:49:55 machine kernel: NVRM: GPU Board Serial Number: 0
Dec 20 23:49:55 machine kernel: NVRM: Xid (PCI:0000:01:00): 79, GPU has fallen off the bus.
Dec 20 23:49:55 machine kernel: NVRM: GPU 0000:01:00.0: GPU has fallen off the bus.
Dec 20 23:49:55 machine kernel: NVRM: GPU 0000:01:00.0: GPU serial number is 0.
Dec 20 23:49:55 machine kernel: NVRM: kgspRcAndNotifyAllChannels_IMPL: RC all channels for critical error 79.
Dec 20 23:49:55 machine kernel: NVRM: _threadNodeCheckTimeout: API_GPU_ATTACHED_SANITY_CHECK failed!

Full output of this command:

gpu_errs.txt (65.7 KB)

5

As per Nvidia engineers:

So if other driver versions don’t help, then check your thermals and if those look good, then try another PSU.

6

I see, thanks for the help!