Problem using the NVIDIA RTX 3050 Ti GPU on Bazzite (Wayland)
Hello everyone,
I need help with a persistent problem to make my system use my dedicated NVIDIA graphics card by default. I have tried several solutions, but the integrated GPU is still the one used for all tasks.
System information:
Distribution: Bazzite (based on Fedora Atomic)
GPU: NVIDIA GeForce RTX 3050 Ti Laptop GPU
Graphics server: Wayland
Diagnostic information :
rpm-ostree status:
State: idle
Deployments:
● ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-deck-nvidia:stable
Digest: sha256:d7ab4f616e00f44aa1ef32b404bdf1132a4c0a30e637bce2e93e2f99ee3655fe
Version: 42.20250621 (2025-06-21T21:19:24Z)
LayeredPackages: bridge-utils qemu-kvm virt-manager
ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-deck-nvidia:stable
Digest: sha256:d7ab4f616e00f44aa1ef32b404bdf1132a4c0a30e637bce2e93e2f99ee3655fe
Version: 42.20250621 (2025-06-21T21:19:24Z)
uname -a
Linux bazzite 6.14.6-109.bazzite.fc42.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Jun 20 08:53:55 UTC 2025 x86_64 GNU/Linux
Steps we have already tried, and their results:
Run export-gpu: The command ran successfully and I selected the NVIDIA GPU, but the problem persisted after restarting.
Try prime-run: The command was not found (bash: prime-run: command not found).
Configure with nvidia-settings: The “PRIME Profiles” option does not appear in the interface, possibly due to the use of Wayland.
Add the kernel parameter with grubby: I ran the command sudo grubby --update-kernel=ALL --args=“nvidia-drm.modeset=1”, but the problem persists.
Check with nvtop: The GPU monitor shows that the NVIDIA GPU remains at 0% usage, while the integrated GPU handles all the workload.
Could this be a problem with the driver or with a specific Bazzite configuration on my hardware? I would greatly appreciate any guidance or solution.
Thanks in advance!
I need help with a persistent problem to make my system use my dedicated NVIDIA graphics card by default. I have tried several solutions, but the integrated GPU is still the one used for all tasks.
System information:
Distribution: Bazzite (based on Fedora Atomic)
GPU: NVIDIA GeForce RTX 3050 Ti Laptop GPU
Graphics server: Wayland
Diagnostic information :
rpm-ostree status:
State: idle
Deployments:
● ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-deck-nvidia:stable
Digest: sha256:d7ab4f616e00f44aa1ef32b404bdf1132a4c0a30e637bce2e93e2f99ee3655fe
Version: 42.20250621 (2025-06-21T21:19:24Z)
LayeredPackages: bridge-utils qemu-kvm virt-manager
ostree-image-signed:docker://ghcr.io/ublue-os/bazzite-deck-nvidia:stable
Digest: sha256:d7ab4f616e00f44aa1ef32b404bdf1132a4c0a30e637bce2e93e2f99ee3655fe
Version: 42.20250621 (2025-06-21T21:19:24Z)
uname -a
Linux bazzite 6.14.6-109.bazzite.fc42.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Jun 20 08:53:55 UTC 2025 x86_64 GNU/Linux
Steps we have already tried, and their results:
Run export-gpu: The command ran successfully and I selected the NVIDIA GPU, but the problem persisted after restarting.
Try prime-run: The command was not found (bash: prime-run: command not found).
Configure with nvidia-settings: The “PRIME Profiles” option does not appear in the interface, possibly due to the use of Wayland.
Add the kernel parameter with grubby: I ran the command sudo grubby --update-kernel=ALL --args=“nvidia-drm.modeset=1”, but the problem persists.
Check with nvtop: The GPU monitor shows that the NVIDIA GPU remains at 0% usage, while the integrated GPU handles all the workload.
Could this be a problem with the driver or with a specific Bazzite configuration on my hardware? I would greatly appreciate any guidance or solution.
Thanks in advance!
