Hyprland Multi-GPU Renderer Selection Issue (AMD iGPU/NVIDIA dGPU) #12284
Unanswered
boxboy523
asked this question in
Bugs - Renderer
Replies: 1 comment
-
|
Check #8679. I added this to my env vars and it seems to be working rn:
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hello, I'm a Hyprland user setting up a hybrid graphics system. I'm facing difficulties forcing the Compositor (Hyprland itself) to use the power-efficient AMD iGPU as the primary screen manager.
Current Issue Status
My goal is for the AMD iGPU to handle basic rendering to save power, with the NVIDIA dGPU only being used for high-performance tasks.
However, based on
nvtopoutput, the Hyprland process itself runs exclusively on the NVIDIA dGPU, while regular client applications correctly use the AMD iGPU.Configuration (Hardware):
nvtopObservation:Configuration Attempts (Referencing Wiki)
I followed the instructions in the [Hyprland Multi-GPU Wiki](https://wiki.hypr.land/Configuring/Multi-GPU/) to set up device priority and achieved a stable, but incorrect, configuration.
udevsymlinks for both GPUs and set the priority with the AMD iGPU first. This is the only combination that allows Hyprland to boot without freezing.env = AQ_DRM_DEVICES,/dev/dri/amd_igpu:/dev/dri/nvidia_gpuCore Question
The current evidence suggests that the AMD iGPU is perfectly capable of rendering client apps, but it fails to initialize as the Compositor's primary rendering backend (as seen by the freezing when set alone). It requires the NVIDIA dGPU path to be in
AQ_DRM_DEVICESto successfully start the Hyprland session, which then causes Hyprland to default to NVIDIA.AQ_DRM_DEVICESto boot, what is the best way to force Hyprland itself to render on the iGPU while still including the NVIDIA dGPU path for output synchronization?Any insights into why the AMD iGPU alone causes freezing but works perfectly when the NVIDIA path is present would be greatly appreciated. Thank you!
P.S. I utilized an AI (Gemini) tool to ensure the clarity and grammatical correctness of this post.
Beta Was this translation helpful? Give feedback.
All reactions