Hello, apologies if this has been discussed before. However, I’m new on here & my search-fu isn’t bringing up any results, or I’m just not seeing it.
I am using a laptop with AMD Ryzen 5 4600H & NVIDIA GeForce 1650 Ti w/Tumbleweed 20240524 & Kernel 6.9.1-1-default.
Reading about SUSE Prime & wondering if that’s still useful for me because I’d like to have some apps run on the GeForce rather than the in-built AMD.
Sorry if my question isn’t clear, not good at writing.
I believe the concept is the same as Intel iGPU with NVIDIA dGPU (open-source drivers on iGPU), and there’s an official method to have on-demand GPU selection with the proprietary NV GPU (prime render offload maybe?). I preferred this a lot more than the default run-everything-always-on-NVIDIA-GPU method, and that prime render offload option worked basically the same as DRI_PRIME with all open-source drivers.
I haven’t tried it in years, but I think NV in driver notes has a xorg.conf snippet that was copy/paste into oS TW no problem. You’d install the NV driver and likely start with the render-only-on-NV method, do xorg.conf, and then reboot to the gpu-select method. I don’t believe it was Intel-specific, but you can likely replace intel with amdgpu if it’s mentioned.
@bionic Look at using switcherooctl? I ran AMD RX550 and Nvidia GPU’s as Prime Render Offload. I only need the AMD gpu to run my displays… Now this is on GNOME with dbus integration of switcherooctl, don’t think Plasma has that yet, so you have to run manually.
No Xorg or Wayland configuration should be necessary. I have two Nvidia GPU’s running on Xorg without any configuration on GNOME, I also have Aeon running on a dual AMD gpu laptop with switcherooctl running fine…
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.