You are not logged in.
Pages: 1
What are the plans for the NVIDIA drivers now that it is open-source?
https://developer.nvidia.com/blog/nvidi … l-modules/
https://www.phoronix.com/scan.php?page= … rnel&num=1
There are some limitations:
> These changes are for the kernel modules; while the user-mode components are untouched. So the user-mode will remain closed source and published with pre-built binaries in the driver and the CUDA toolkit.
> Only Turing and newer GPUs will be supported by this open-source kernel driver. Pre-Turing GPUs are left to using the existing proprietary kernel drivers or the Nouveau DRM driver for that matter. Turing and newer is a hard requirement due to being dependent upon the GPU System Processor (GSP).
> So the NVIDIA Open Kernel Driver is certainly superior for GeForce RTX 20/30 series while GTX 900 / GTX 10 graphics cards will likely be left in an awkward state outside of the proprietary driver stack.
Last edited by jd8nN (2022-05-11 21:48:35)
Offline
I packaged them as nvidia-open{,-dkms} and put them into [testing]. Please give them a spin and see whether they do anything for you.
Offline
I guess no one really knows, but do you think the regular proprietary driver will still be maintained in the future? It works very well for me and my RTX2060 with OpenCL.
The OP says the new, open driver "is certainly superior". In what ways?
Offline
Here is a blog post from a developer which addresses my question:
Offline
How can I intall the open driver for my rtx 3050 ti?
Offline
Install the linux-headers or similar for your current kernel, enable the testing and multilib-testing repositories and install nvidia-open-dkms and {lib32-}nvidia-utils from the respective testing repositories.
Offline
Thanks for packing that?
Can we not also include the files under `.github` directory?
Offline
I guess no one really knows, but do you think the regular proprietary driver will still be maintained in the future? It works very well for me and my RTX2060 with OpenCL.
The OP says the new, open driver "is certainly superior". In what ways?
From my understanding the reasoning behind developing this new driver from the ground up was because the old driver's codebase doesn't conform to Linux upstream's standards. I doubt that's a surprise to anyone to be honest.
Logically I guess that means this new driver is a lot cleaner and will be easier to maintain, and for new developers to be able to jump right in without requiring intimate knowledge of the old, poorly organised driver's inner workings.
"Navigating one's life is like being a pathfinding algorithm with no paths to find." - Ghandi, probably
Offline
Pages: 1