For some reason the critical temperature is being set to 77 C. This makes the GPU basically unusable because when it is reached, the GPU goes back to the 300 MHz base clock. My kernel version is 6.8.0, and my driver version is 550.67. Any advice would be appreciated.
nvidia-bug-report.log.gz (496.4 KB)
Does that notebook support platform profiles that might change the temperature settings?
https://docs.kernel.org/userspace-api/sysfs-platform_profile.html
It doesn’t have /sys/firmware/acpi/platform_profile, but it has /sys/firmware/acpi/pm_profile.
That’s something unrelated, doesn’t help.
Is there any other way to adjust it? It looks like the value is called NVML_TEMPERATURE_THRESHOLD_GPU_MAX in the nvidia-ml library.
I don’t think so, it should be set by the system’s firmware.