Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- Wouldnt it make sense to just implement a Vulkan driver and then use Zinc on top of that for OpenGL?by preisschild
- I think to reduce ewaste part it should be mandatory with open firmware and open hardware specificationsby acd
- Damn, I did exactly this already for the M3 and got it working. Was hoping to make a post about it but this kinda scoops it. Bravo!by superdisk
- Share your code please.by tiahura
- What did you use for making those rather sexy diagrams/figures?by prier
- I am also looking for what was used, I really like the design of these!
- What is the state of Nvidia and AMD GPU drivers on Linux? Maybe these techniques can finally give us reasonably good open source drivers which can be supported for a long time.
AFAIK Nvidia drivers are proprietary and AMD has a tendency to drop support quickly
by shipp02 - amen to that. amdgpu is a shitshow for years. im sure the firmware plays its part in the shitshow, but amdgpu does as well.
- AMD driver support is great due to community effort, even for 10+ year old cards. Nvidia also now has (partly?) open drivers.by Azantys
- Maybe I’m getting really old, and I know we live in changing times but the fact this works at all is insane, as in black magic insane.
I assume I’ll get used to it, but my mental model of what’s possible hasn’t been fully rewired yet.
by Agingcoder - I still can't quite believe it either!
- And this kiddo is probably barely 22yoby afr0ck
- To the developer: it doesnt matter whatever you can upstream it or not. Just share the code and documentation of how this process can be reproduced.
There plenty of us not working at Apple who can reuse the same approach for M5 and next generations of hardware.
by SXX - For the kernel driver it's literally just: run the hypervisor, wait for the kick, save the UAT state, replay the whole thing, make sure the output pages change, then just follow the pointer to initdata until you've built the whole tree from scratch. LLM handles the rest.
- All this work is tainted since the poster is ex-Apple[1]. There is no way Linux will take that code, since people from Apple also contribute to Linux and this creates a major conflict.
There's also ongoing issues around Codex for this sort of thing, given Apple is suing OpenAI over stolen trade secrets[2]. And nobody knows whether training data in other models is similarly tainted.
[1]: https://codyho.dev/documents/cody_ho_resume.pdf
[2]: https://www.theguardian.com/technology/2026/jul/10/apple-sue...
- There were and are many important pieces of Linux kernel that live out of tree; ZFS is a big example.
The problem with this driver is not licensing or code quality; I assume it's under a threat of receiving C&D letter, or maybe also a legal suit for breaking some NDA.
by nine_k - I don't give a flying fuck.
Ignoring the idiocy of Imaginary Property is how China is winning.
It's time we caught up and then leap ahead. Everything is a derivative work.
by userbinator - I don't know if that holds.
Neither OpenAI or Apple "owns" the Linux project. As long as there is enough evidence that code is real contribution and not stolen property (if you can even prove it), there is no reason to deny it.
by nicce - Why do you hide behind throwaway accounts instead of just using your identity ? its harder to listen if you dont want to say who you areby kanwisher
- > All this work is tainted since the poster is ex-Apple[1].
So if I worked as a janitor at Apple, I couldn't contribute anymore?
by cromka - Resume says : "Machine Learning Engineer".
Exactly how does that equate to knowing details of Apple's GPU implementation? Apple is a massive company.
by outworlder - Linking to another HN comment: https://news.ycombinator.com/item?id=49719480
- Now that this work is done and out in the open what in theory would prevent the next generation of LLMs from ingesting it and eventually a non-Apple employee creating yet another driver “from scratch” but with the knowledge of how the hardware works acquired from an LLM or with the help of one?
Basically, aren’t we facing a situation where LLM-assisted knowledge laundering is becoming a thing?
by IgorPartola - This is super great. The biggest pain point of Asahi Linux is how it doesn't have GPU acceleration on M3 and newer, especially now that M6 is out!
However, Asahi Linux has a strictly no-AI policy [1]. So this great work can't be upstreamed. I expect to see a bunch of AI-assisted forks that get things working smoothly on newer hardware to dominate as most people just care about getting stuff working, while only a handful of purists stick to the non-AI version running on ancient hardware.
by porphyra - I imagine either they or others can take their discoveries and write a real driver now, though. The hard part was always the reversing the black box system.by smith7018
- Asahi's long term goal is to get everything possible merged into the (actually) upstream projects anyways so any distro can just work. It'd be nice to see that continue rather than have forks on forks for the sake of singular differences (and it looks like proper upstreaming is what they are going after per the Remaining Work section).by zamadatix
- :) I think we have a surprise in store here. Asahi don't have a monopoly over Linux for Apple Silicon, and upstream Linux absolutely does *NOT* ban LLMs.
- It's extremely impressive that they were able to make a working driver so quickly. I think this is one of the best use cases for LLMs. You don't need someone to spend years reverse engineering undocumented hardware anymore. It will interesting to see how good the driver the LLMs came up with is, and whether it can be upstreamed into the Linux kernel.by ndiddy
- I’m sure it’s really easy when you can get help from the people who made the chip or write the Apple drivers.by MBCook
- I can’t help wondering if the whole process will be automated being an LLM.
Apple releases the M15, and using everything leaned while building drivers for the M4 - M14, the LLM just does the whole job and we have the drivers a day later.
by testing22321