this post was submitted on 12 May 2024
475 points (86.0% liked)
linuxmemes
21434 readers
1132 users here now
Hint: :q!
Sister communities:
Community rules (click to expand)
1. Follow the site-wide rules
- Instance-wide TOS: https://legal.lemmy.world/tos/
- Lemmy code of conduct: https://join-lemmy.org/docs/code_of_conduct.html
2. Be civil
- Understand the difference between a joke and an insult.
- Do not harrass or attack members of the community for any reason.
- Leave remarks of "peasantry" to the PCMR community. If you dislike an OS/service/application, attack the thing you dislike, not the individuals who use it. Some people may not have a choice.
- Bigotry will not be tolerated.
- These rules are somewhat loosened when the subject is a public figure. Still, do not attack their person or incite harrassment.
3. Post Linux-related content
- Including Unix and BSD.
- Non-Linux content is acceptable as long as it makes a reference to Linux. For example, the poorly made mockery of
sudoin Windows. - No porn. Even if you watch it on a Linux machine.
4. No recent reposts
- Everybody uses Arch btw, can't quit Vim, and wants to interject for a moment. You can stop now.
Please report posts and comments that break these rules!
Important: never execute code or follow advice that you don't understand or can't verify, especially here. The word of the day is credibility. This is a meme community -- even the most helpful comments might just be shitposts that can damage your system. Be aware, be smart, don't fork-bomb your computer.
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
You should consider passing through your Nvidia GPU to a virtual machine in order to do compute tasks on; that way, your host machine won't be infected with proprietary Nvidia drivers (I'm assuming you need CUDA for your compute tasks). The only performance differences you'll notice is less available system RAM (you will have access to all of your VRAM), and very slightly less CPU performance, due to running two operating systems at the same time (barely even noticable, TBH). This is the option that I would personally recommend.
If you want to try a super hacky solution which might not work for everything you need, you can try using the open source, recently released ZLUDA translation layer to perform CUDA tasks on your AMD GPU.
https://github.com/vosen/ZLUDA
The reason Hyprland doesn't work with proprietary Nvidia drivers is due to Nvidia refusing to implement the accepted Wayland standard in favor of their own, home-rolled solution which is incompatible. AFAIK, only GNOME and KDE implement that standard.
Hyprland works fine on Nvidia, I've been using it for about a year now. It's only going to improve now that Nvidia hired people from the Nouveau team to work on Nouveau and Nvidia is making the open drivers the default in version 560. Can't wait for the 555 drivers they've been working on with the Wayland team and most of the major desktops to implement explicit sync etc.
An option would be to only install the CUDA toolkit without the drivers but distros like Ubuntu just don't support it. You could also switch display managers to sddm because Hyperland recommends it, might work better. Hyprland prints information in the tty if you launch it with
Hyprland. I'm just thinking it's gdm being weird tbh.Ah, I like this solution. Thanks for the suggestion! I set up GPU passthrough for a VM on a build years ago with QEMU. I'm sure I'll be able to figure that out again.