Re: The XDNA NPU stack
On Sat, Aug 2, 2025 at 12:33 AM M. Zhou <cdluminate@riseup.net> wrote:
>
> Hi Cordell,
>
> Thanks for the summary on AMD's NPUs.
>
> The current situation of NPU on linux looks somewhat messy and not
> mature. For GPUs, there are just three major vendors, and their software
> stack are relatively more mature than NPUs. In contrast, the number of
> NPU vendors is too many, and the support level on linux differs. Even
> worse, each vendor seems to keep one copy of their own fork of LLVM that
> is unlikely to merge upstream.
>
> I will personally keep an eye on the news, and I do not have plan to
> work on NPU packages myself since I see no signal of ecosystem maturity.
> Of course, potential contributors are welcome to push things forward.
>
With my "idiot user" hat on, I'm quite interested on NPU support on
Linux platforms, especially whether there are use cases that may
change the day-to-day experience a lot, for example:
1. Run a small LLM model locally with NPU
2. Traditional applications with AI support, e.g. image processing
(like denoise/upscale), video playback (frame interpolation) and
editing (like using Blender).
There could be a lot more cases, and it will potentially also help
industrial users with even more powerful NPU hardware.
Cheers,
Aron
Reply to: