Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login

Does libhybris support X11 with GPU accel now? It didn't when I tried to use it in the past


sort by: page size:

I think Xwayland GPU acceleration won't work with NVIDIA still, though.

Really? Where is the GPGPU support of X?

Last time I checked it was a X extension, if at all.


That's a driver; has nothing to do with X itself.

Consider that the nVidia driver works on Solaris, FreeBSD, and Linux. Therefore accelerated X works there as well.


Understandably it can use libhybris. I was asking what GPU is used in the poster's watch. If it has native drivers, libhybris isn't necessary.

It certainly used to work fine with llvmpipe; I tested it when it first came out and it worked with no problems. Was pretty fast, too; it wasn't crippled by it though the stress tests (that do horribly in Firefox and Chrome) didn't do that well in it either (of course). This may have changed and broken it on llvmpipe.

In general we've not smoothed out this stuff so that you can fall back cleanly when a GPU doesn't exist.


I'm using it on an AMD GPU with the clblast backend.

Still no GPU accel for the GUI?

Is the 2D X11 acceleration done via glamor using the closed source driver?

(Congrats btw!)


Iterm2 also supports GPU accel. FYI

It now supports AMD GPUs since 3 weeks ago, check the latest commit at the repo:

https://github.com/vosen/ZLUDA

The article also mentions exactly this fact.


There's a patched version of Chromium called chromium-vaapi (in Fedora's RPM Fusion repositories) which supports GPU acceleration.

X on Raspbian is not GPU accelerated. Since the CPU (but not the GPU!) is a bit anemic even by current ARM SoC standards, that's a big deal.

Firefox OS is accelerated, running on top of OpenGL ES. :)


Was it impossible to add GPU acceleration while keeping the old API?

Accel uses the LLVM PTX backend, which is unusable for serious projects and doesn't work on windows, i have more about it here https://github.com/RDambrosio016/Rust-CUDA/blob/master/guide...

It's not officially supported, but it does seem to work, don't know if there are any bugs.

I agree that AMD is dropping the ball.


yeah, but it's no longer current/supported as far as I can tell.. the actual project appears to be dead and I can't find it on the backports at all. Shame, because it looked promising for a while.

I'm a bit amazed that Nvidia isn't supporting this more, given y'know, SteamOS just around the corner.


Doesn't XeSS work on non-Intel GPUs as well?

I used this half a year ago, love the UX but it was not possible to accelerate the workloads using an AMD GPU. How's the support for AMD GPUs under Ollama today?

Tried it but it seems it only works with Cuda 11 and I have 12 installed. Not really willing to potentially screw up my Cuda environment to try it.
next

Legal | privacy