Nvidia-powerd using lots of cpu on fedora 35 running gnome on wayland

Nothing obvious in the logs.

$ rpm -qf /usr/bin/nvidia-powerd

$ systemctl status nvidia-powerd.service
● nvidia-powerd.service - nvidia-powerd service
Loaded: loaded (/usr/lib/systemd/system/nvidia-powerd.service; enabled; vendor preset: enabled)
Active: active (running) since Mon 2022-02-14 09:48:58 EST; 3h 29min ago
Main PID: 1038 (nvidia-powerd)
Tasks: 3 (limit: 76842)
Memory: 134.2M
CPU: 27min 32.308s
CGroup: /system.slice/nvidia-powerd.service
└─1038 /usr/bin/nvidia-powerd

Feb 14 09:48:58 rackn.timbos.se systemd[1]: Starting nvidia-powerd service…
Feb 14 09:48:58 rackn.timbos.se /usr/bin/nvidia-powerd[1038]: nvidia-powerd version:1.0(build 1)
Feb 14 09:48:58 rackn.timbos.se /usr/bin/nvidia-powerd[1038]: Dbus Connection is established
Feb 14 09:48:58 rackn.timbos.se systemd[1]: Started nvidia-powerd service.

$ rpm -qa | grep nvidia

$ uname -r

Let me know if there is anything else I can provide to help troubleshoot.

Opened from 6220 – nvidia-powerd using lots of CPU

nvidia-bug-report.log.gz (320.3 KB)


Same problem in Fedora 35 X11 plasma session. After system start all was fine, but after I played some games nvidia-powerd started to consume 100% of one thread and stopped only after I restarted it (games were not running already, but daemon kept consuming cpu until restart).
nvidia-bug-report.log.gz (1.3 MB)
UPDATE: it also eats 130MB of memory.

1 Like

Same problem in Fedora 35 X11 gnome session. At some point nvidia-powerd starts to consume 100% of one thread. Can be worked around by restarting the nvidia-powerd, but sooner or later it degrades again.

nvidia-bug-report.log.gz (387.3 KB)

1 Like

Same problem for me as well, using Fedora 35 with the i3 window manager. I can restart nvidia-powerd but after some time it goes crazy again…

Does anyone have any temporary fix to suggest? E.g. a bash script that monitors whether the process gets stuck and restarts it automatically?

Same problem with Fedora 35 on Lenovo Thinkpad P1 G4. I “solve” it by killing nvidia-powerd after each general reboot.
From root I do “killall nvidia-powerd” . Then suspend and hibernate work as expected and CPU usage is no longer crazy.

Problem persists in Fedora 36, very annoying.

Same problem with Ubuntu 22.04, latest drivers from NVidia repos (see package details below) and using Xorg.
It starts consuming 100% of the CPU until I restart the services, but after sometime it starts to consume 100% of CPU again.

# dpkg -l | grep nvidia | grep ^ii
ii  libnvidia-cfg1-520:amd64                   520.61.05-0ubuntu1                         amd64        NVIDIA binary OpenGL/GLX configuration library
ii  libnvidia-common-520                       520.61.05-0ubuntu1                         all          Shared files used by the NVIDIA libraries
ii  libnvidia-compute-520:amd64                520.61.05-0ubuntu1                         amd64        NVIDIA libcompute package
ii  libnvidia-compute-520:i386                 520.61.05-0ubuntu1                         i386         NVIDIA libcompute package
ii  libnvidia-decode-520:amd64                 520.61.05-0ubuntu1                         amd64        NVIDIA Video Decoding runtime libraries
ii  libnvidia-decode-520:i386                  520.61.05-0ubuntu1                         i386         NVIDIA Video Decoding runtime libraries
ii  libnvidia-encode-520:amd64                 520.61.05-0ubuntu1                         amd64        NVENC Video Encoding runtime library
ii  libnvidia-encode-520:i386                  520.61.05-0ubuntu1                         i386         NVENC Video Encoding runtime library
ii  libnvidia-extra-520:amd64                  520.61.05-0ubuntu1                         amd64        Extra libraries for the NVIDIA driver
ii  libnvidia-fbc1-520:amd64                   520.61.05-0ubuntu1                         amd64        NVIDIA OpenGL-based Framebuffer Capture runtime library
ii  libnvidia-fbc1-520:i386                    520.61.05-0ubuntu1                         i386         NVIDIA OpenGL-based Framebuffer Capture runtime library
ii  libnvidia-gl-520:amd64                     520.61.05-0ubuntu1                         amd64        NVIDIA OpenGL/GLX/EGL/GLES GLVND libraries and Vulkan ICD
ii  libnvidia-gl-520:i386                      520.61.05-0ubuntu1                         i386         NVIDIA OpenGL/GLX/EGL/GLES GLVND libraries and Vulkan ICD
ii  nvidia-compute-utils-520                   520.61.05-0ubuntu1                         amd64        NVIDIA compute utilities
ii  nvidia-dkms-520                            520.61.05-0ubuntu1                         amd64        NVIDIA DKMS package
ii  nvidia-driver-520                          520.61.05-0ubuntu1                         amd64        NVIDIA driver metapackage
ii  nvidia-kernel-common-520                   520.61.05-0ubuntu1                         amd64        Shared files used with the kernel module
ii  nvidia-kernel-source-520                   520.61.05-0ubuntu1                         amd64        NVIDIA kernel source package
ii  nvidia-prime                                                        all          Tools to enable NVIDIA's Prime
ii  nvidia-settings                            520.61.05-0ubuntu1                         amd64        Tool for configuring the NVIDIA graphics driver
ii  nvidia-utils-520                           520.61.05-0ubuntu1                         amd64        NVIDIA driver support binaries
ii  screen-resolution-extra                    0.18.2                                     all          Extension for the nvidia-settings control panel
ii  xserver-xorg-video-nvidia-520              520.61.05-0ubuntu1                         amd64        NVIDIA binary Xorg driver

It is a Lenovo P1 Gen 4 with the following NVidia card:

NVIDIA T1200 Laptop GPU
01:00.0 VGA compatible controller: NVIDIA Corporation TU117GLM [T1200 Laptop GPU] (rev a1)

I have the Intel GPU disabled, but it also happens with the Intel GPU enabled.

I’m using the laptop monitor and two external monitors with the Lenovo Thunderbolt 4 Dock (also tested with Lenovo Thunderbolt 43 Dock): Screen resolutions (basically 1920x1080 and 1920x1200 on the laptop monitor):

$ xrandr --listmonitors
Monitors: 3
 0: +*eDP-1-1 1920/345x1200/215+1920+0  eDP-1-1
 1: +DP-0.1.5 1920/477x1080/268+0+120  DP-0.1.5
 2: +DP-0.2 1920/477x1080/268+3840+120  DP-0.2

The power settings, from Nvidia-settings application are set in “Auto” mode.

Same issue with Mint 21.1, I addressed it with a cron ( */1 * * * * /path/to/script.sh > /path/to/script.log 2>&1 ) calling this script :

nvidia_usage=$(ps -C nvidia-powerd -o %cpu | grep -v C | awk '{printf $1}' | cut -d "." -f 1)
if [ "$nvidia_usage" -gt 5 ]; then
    systemctl restart nvidia-powerd
    echo "nvidia-powerd service KO => restarted"
    echo "nvidia-powerd service OK"

I hope that helps, waiting for a fix as well.
I noticed this behavior since cuda install a few days ago.