What an end to an era. It's crazy to think she started this journey at 18 and now finished 5 years later. Not many people believed they would be able to make the GPU work in Asahi linux. Kinda curious what her "Onto the next challenge!" link means. Is she working for Intel Xe-HPG next?
Wish her the best with this. Intel staying competitive in GPUs can only benefit the consumer. Those who want a mid-tier graphic card, without paying to compete with AI use cases, may not a huge group, but we do exist! Those who use desktop Linux may be a small group among that small group, but we do exist!
Apple is too much about beeing closed and creating barriers not sure that would have been a good fit. Plus that's a good way to flee a country quickly degrading.
Although most likely she’s well compensated, and doesn’t have to waste time on useless efforts at work, this level of discipline and striving towards a goal is just very rare in general.
Possibly also no family, limited social life and no other hobbies.
For myself, when I lived on a different continent to my family, had limited social life and job with strictly set hours, it was much easier to have the time needed to make significant progress on a hobby.
However, discipline is an enormous factor too, actually using that extra available time on something “productive” is no easy feat.
Now I have kids and live in the same area as my parents and siblings again, entirely happy, but less free time.
2021 and 2022 was also when many places were only just coming out of COVID lockdowns. I remember how much dead time I had back then. I used it to watch lots of series and youtube videos. I wish I had the discipline and motivation to work like she did during that area with all that free time.
May I ask something, I want an apple silicone MacBook Air and I am probably just be running Linux on it, what are pros and cons of getting an m1 vs m2? Except for more ram or so.
Thx
They seem closely aligned with the Free Software Foundation (FSF), so I could very well imagine that being a major ideological reason not to want to work with Apple. Yes, Apple sometimes upstream patches and they do contribute to open source here and there, but they certainly are no FSF poster child. Intel on the other hand are about as open as it gets when it comes to their track record in the graphics space. I personally have nothing but admiration for Rosenzweig's work and I hope they will continue to find environments where they can flourish and do great things in the years to come.
You do have to wonder how that kind of interview would go. Hopefully it would be actual engineers that created what she reverse engineered instead of some gatekeeper trying to one up her somehow.
Inspiring stuff! I didn't even expect basic Linux support on M1 to be so good in such a short time-span, leaving graphics aside. I was very pleased when I tried booting up Asahi on M1 a couple months back and went on to get work done in it and even enjoy some games.
Thanks for all your amazing contributions Alyssa and all the best for the road ahead!
Sorry to hijack, but since the topic is related: is the development of Asahi Linux still actively ongoing, or has slowed down a lot? The progress for M1 and M2 was steady and now almost everything is done, but the M3+ work still seems to not have started. And with major contributors leaving the project I'm kind of worried for the future of Asahi (on newer Apple hardware).
The new leadership team set a short term goal of getting their existing work upstreamed, which seems to be going well.
> Our priority is kernel upstreaming. Our downstream Linux tree contains over 1000 patches required for Apple Silicon that are not yet in upstream Linux. The upstream kernel moves fast, requiring us to constantly rebase our changes on top of upstream while battling merge conflicts and regressions. Janne, Neal, and marcan have rebased our tree for years, but it is laborious with so many patches. Before adding more, we need to reduce our patch stack to remain sustainable long-term.
> With Linux 6.16, we also hit a pretty cool milestone. In our first progress report, we mentioned that we were carrying over 1200 patches downstream. After doing a little housekeeping on our branch and upstreaming what we have so far, that number is now below 1000 for the first time in many years, meaning we have managed to upstream a little over 20% of our entire patch set in just under five months. If we discount the DCP and GPU/Rust patches from both figures, that proportion jumps to just under half!
While we still have quite a way to go, this progress has already made rebases significantly less hassle and given us some room to breathe.
> With Linux 6.16, we also hit a pretty cool milestone. In our first progress report, we mentioned that we were carrying over 1200 patches downstream. After doing a little housekeeping on our branch and upstreaming what we have so far, that number is now below 1000 for the first time in many years, meaning we have managed to upstream a little over 20% of our entire patch set in just under five months. If we discount the DCP and GPU/Rust patches from both figures, that proportion jumps to just under half!
So if the discussions are true, it can take years for the developers to finish M1/M2 upstreaming with all the Linux kernel bureaucracy. That is, unless they decide to start working on M3 before finishing the upstreaming
Makes sense, every patch they upstream is less maintenance and forward-porting work that they have to do. Keeping a downstream kernel up to date is very painful, even one that's "near mainline" as with Asahi's.
I would hope not. That would mean that no other vendor has shipped working ARM hardware support for Linux or has upstream support in the kernel. Forget the hostile nature Apple has proven to possess when consumers dare treat their hardware as if paying for it makes it their own.
Qualcomm has been beating the marketing drum on this instead of delivering. Ampere has delivered excellent hardware but does not seem interested in the desktop segment. The "greatest Linux laptop around" can not be some unmaintained relic from a hostile hardware company.
As somebody that has worked in a company that did Qualcomm devices in the past - Qualcomm just cares about money grabbing, and is not any less hostile to developers than Apple.
If you want to do a device, and your only chip option is Qualcomm I'd recommend not doing a device at all.
FLOSS stacks for Qualcomm devices are actually a lot more feature complete than some other brands like MediaTek or Exynos. Still nowhere near any kind of "daily driver" status but at least getting somewhere, whilst others haven't even gottten started.
I'd pay easily let's say $100-200 a year to have linux running on modern apple laptops with full features. I'm sure I'm not alone. Their hardware, "our" OS would be perfect. Well, except notch and lack of OLED - but, reportedly that's in the works too.
Most likely, they have more mini-leds and/or more ability to independently control them. Of course the localized "blooming" of mini-leds is a lot better on the eyes regardless than the all-around bloom of a backlit display.
(Better for the battery, too, if you can keep most of the screen dark.)
The M3+ GPU is also very different. So while it may be true that the driver development for M1/M2 is now more or less complete as OP says, future work along the same lines will very much be needed.
I never understood this project. Maybe I'm missing something, but the timescale is such that by the time they're done the product isn't even being sold anymore
At least with Panfrost it made more sense bc it still being used
M1 chip laptops can only be bought second hand at this point
I believe Walmart has a deal[1] with Apple to sell[2] M1 MacBook Airs. This has been the case for a year or so, so I don't think it's old stock. They have been in stock for since that date, and slowly getting cheaper.
Why would the product have to be available new for the project to be worth it? There are still many M1 chips out there, and this helps prolong the usefulness of those chips.
> M1 chip laptops can only be bought second hand at this point
New M1 Macbook Airs are still available at Walmart (maybe elsewhere). But even if not, who cares? People are still writing code for computers that haven't been sold since the 1980s.
Im glad she stepped away from Asahi linux. Its absolutely great from a techincal perspective and the progress that team has made, but talented people like her shouldn't be trying to reverse engineer software/hardware from shitty anti-consumer company that can make the entire project work in a heartbeat by publishing documentation, in lieu of building better stuff from the ground up.
Reverse engineering requires a different mindset and somewhat different skill set than “forward” engineering. I’ve met people who were happy to only do reverse engineering (to figure out what make things “tick”) without building anything new.
If it was up to me, 2 years of successful reverse engineering (of a variety of projects/products) would be a requirement to be called an engineer. You learn a lot from working things that you can’t learn from a book (and without having to do the mistakes yourself first…)
Just to make it clear: I am not implying anything about Alyssa - just stating an observation based on my own experience.
> in lieu of building better stuff from the ground up
To be fair, even if you have the best CPU and GPU designers, it's not as if you can call up TSMC and have them do a run of your shiny new processor on their latest (or even older) process. You can't fab them at home either.
Even with proper documentation, there still would have been loads of work to get M1/M2 GPUs working on Asahi Linux. Writing GPU drivers worth a damn is about as difficult as targeting a compiler to a new CPU architecture. It would not be "in a heartbeat".
Honestly kind of heartbreaking to see her leave asahi Linux. She has done insane work building the vulkan driver from scratch. I wish her well working at Intel. If I ever buy an Intel GPU I can rest much easier it will work well on Linux. If she is working on the Linux driver stack that is.
There isn't really anything left to do for her - everything missing (including work on the newer graphics chips) can be somewhat easily done by less talented people, building on her work.
She did the challenging stuff she cares about. One aspect of nerd brain often is that you can hyperfocus on challenging stuff, but can't get the motivation to work on stuff you don't care about - and even what would be a 20 minute task can end up taking days because of that. It's great that she has the self awareness to set goals, and step away once they're done.
I didn't have that in that age - and still sometimes struggle. I was lucky enough that my employer back then recognized my issues, and paired other people with me for doing the stuff I was not interested in, and now usually manage to load those issues onto other co-workers by myself.
This is of course great as long as you can find enough "challenging" work to perform, but any successful project is going to involve a whole lot of seemingly "boring" work. A big part of true maturity and professionalism is being able to find the interesting challenge even in these more run-of-the-mill tasks and successfully engage with them.
(Mind you, I'm not talking about a matter of inborn temperament or character, much less a moral flaw! Rather, finding the compelling challenge even in "boring" tasks is a valuable skill and situational tactic that anyone should explicitly learn about and aim to acquire as part of becoming a mature professional, not a matter of morality or somehow being dismissed as "lazy"!)
M3, M4 and soon to be M5 are ready to be cracked open :). From what I understand they are actually different somewhat hardware wise. So it's really not like there would not be a continuations of this work. But of course it's natural to want something else after years of working on the project.
Take into account that she's focusing on the 3D stack, not the overall hardware. Even with hardware differences there's a good chance it's not different enough to make it an interesting new challenge.
Given the features that have been advertised for the M3+ graphics and compute stack, there's rather a good chance that it is different enough to create big, new challenges for third-party support.
Yes I meant the GPU specifically not the hardware in general. An example is the support for hardware ray tracing in M3 and beyond. In some now deleted french fediverse post Alyssa indicated M3 has a new architecture even.
What an end to an era. It's crazy to think she started this journey at 18 and now finished 5 years later. Not many people believed they would be able to make the GPU work in Asahi linux. Kinda curious what her "Onto the next challenge!" link means. Is she working for Intel Xe-HPG next?
Yes I think so. Her resume says she started working for Intel on open source graphics driver this month.
Wish her the best with this. Intel staying competitive in GPUs can only benefit the consumer. Those who want a mid-tier graphic card, without paying to compete with AI use cases, may not a huge group, but we do exist! Those who use desktop Linux may be a small group among that small group, but we do exist!
Too bad it was not Apple who hired her for M4, but in business leaders are always the most closed ones.
Apple is too much about beeing closed and creating barriers not sure that would have been a good fit. Plus that's a good way to flee a country quickly degrading.
Good luck to her. That's one of the pieces of Intel I think will survive its slow motion implosion.
The author basically speedran modern graphics APIs on 'impossible' hardware and then just... walks away. Total mic drop.
Pretty cool. She’s achieved more at 23 than I have after over a decade in the industry. What a talented engineer.
No clue who you are, but real talk she’s achieved more than I will in my entire life. I’ve been in the industry for decades.
1. student at uoft 2. a lead in a job at Collabora 3. very succesful and ambitious hobby project
how tf does she juggle and managed to do all this? I can barely do one of the above properly.
One of the few people who are actually competent.
Although most likely she’s well compensated, and doesn’t have to waste time on useless efforts at work, this level of discipline and striving towards a goal is just very rare in general.
Possibly also no family, limited social life and no other hobbies.
For myself, when I lived on a different continent to my family, had limited social life and job with strictly set hours, it was much easier to have the time needed to make significant progress on a hobby.
However, discipline is an enormous factor too, actually using that extra available time on something “productive” is no easy feat.
Now I have kids and live in the same area as my parents and siblings again, entirely happy, but less free time.
One of the unspoken benefits of being young, you're unlikely to have grown into a management position and can focus on not-management stuff.
This is what managers tell themselves to feel better about their idleness, but in the end it’s just another excuse.
Every person is different of course, there might be this one brilliant engineer forced to manage against his will somewhere.
2021 and 2022 was also when many places were only just coming out of COVID lockdowns. I remember how much dead time I had back then. I used it to watch lots of series and youtube videos. I wish I had the discipline and motivation to work like she did during that area with all that free time.
Not much to say beyond a hearty “well done, you!” That, and looking forward to see what’s next.
May I ask something, I want an apple silicone MacBook Air and I am probably just be running Linux on it, what are pros and cons of getting an m1 vs m2? Except for more ram or so. Thx
Kind of amazing Alyssa didn’t end up working at Apple (instead of Intel).
They seem closely aligned with the Free Software Foundation (FSF), so I could very well imagine that being a major ideological reason not to want to work with Apple. Yes, Apple sometimes upstream patches and they do contribute to open source here and there, but they certainly are no FSF poster child. Intel on the other hand are about as open as it gets when it comes to their track record in the graphics space. I personally have nothing but admiration for Rosenzweig's work and I hope they will continue to find environments where they can flourish and do great things in the years to come.
Maybe she didn’t pass a leetcode interview. :p
You do have to wonder how that kind of interview would go. Hopefully it would be actual engineers that created what she reverse engineered instead of some gatekeeper trying to one up her somehow.
Maybe she doesn't want to.
Inspiring stuff! I didn't even expect basic Linux support on M1 to be so good in such a short time-span, leaving graphics aside. I was very pleased when I tried booting up Asahi on M1 a couple months back and went on to get work done in it and even enjoy some games.
Thanks for all your amazing contributions Alyssa and all the best for the road ahead!
Sorry to hijack, but since the topic is related: is the development of Asahi Linux still actively ongoing, or has slowed down a lot? The progress for M1 and M2 was steady and now almost everything is done, but the M3+ work still seems to not have started. And with major contributors leaving the project I'm kind of worried for the future of Asahi (on newer Apple hardware).
The new leadership team set a short term goal of getting their existing work upstreamed, which seems to be going well.
> Our priority is kernel upstreaming. Our downstream Linux tree contains over 1000 patches required for Apple Silicon that are not yet in upstream Linux. The upstream kernel moves fast, requiring us to constantly rebase our changes on top of upstream while battling merge conflicts and regressions. Janne, Neal, and marcan have rebased our tree for years, but it is laborious with so many patches. Before adding more, we need to reduce our patch stack to remain sustainable long-term.
https://asahilinux.org/2025/02/passing-the-torch/
> With Linux 6.16, we also hit a pretty cool milestone. In our first progress report, we mentioned that we were carrying over 1200 patches downstream. After doing a little housekeeping on our branch and upstreaming what we have so far, that number is now below 1000 for the first time in many years, meaning we have managed to upstream a little over 20% of our entire patch set in just under five months. If we discount the DCP and GPU/Rust patches from both figures, that proportion jumps to just under half!
While we still have quite a way to go, this progress has already made rebases significantly less hassle and given us some room to breathe.
https://asahilinux.org/2025/08/progress-report-6-16/
Found out from some Reddit discussions that the developers aim to first upstream everything for M1/M2 to the kernel, and as of https://asahilinux.org/2025/08/progress-report-6-16/:
> With Linux 6.16, we also hit a pretty cool milestone. In our first progress report, we mentioned that we were carrying over 1200 patches downstream. After doing a little housekeeping on our branch and upstreaming what we have so far, that number is now below 1000 for the first time in many years, meaning we have managed to upstream a little over 20% of our entire patch set in just under five months. If we discount the DCP and GPU/Rust patches from both figures, that proportion jumps to just under half!
So if the discussions are true, it can take years for the developers to finish M1/M2 upstreaming with all the Linux kernel bureaucracy. That is, unless they decide to start working on M3 before finishing the upstreaming
Makes sense, every patch they upstream is less maintenance and forward-porting work that they have to do. Keeping a downstream kernel up to date is very painful, even one that's "near mainline" as with Asahi's.
i hope some day a used M1/M2 macbook air will be the greatest linux laptop around
I would hope not. That would mean that no other vendor has shipped working ARM hardware support for Linux or has upstream support in the kernel. Forget the hostile nature Apple has proven to possess when consumers dare treat their hardware as if paying for it makes it their own.
Qualcomm has been beating the marketing drum on this instead of delivering. Ampere has delivered excellent hardware but does not seem interested in the desktop segment. The "greatest Linux laptop around" can not be some unmaintained relic from a hostile hardware company.
As somebody that has worked in a company that did Qualcomm devices in the past - Qualcomm just cares about money grabbing, and is not any less hostile to developers than Apple.
If you want to do a device, and your only chip option is Qualcomm I'd recommend not doing a device at all.
FLOSS stacks for Qualcomm devices are actually a lot more feature complete than some other brands like MediaTek or Exynos. Still nowhere near any kind of "daily driver" status but at least getting somewhere, whilst others haven't even gottten started.
I'd pay easily let's say $100-200 a year to have linux running on modern apple laptops with full features. I'm sure I'm not alone. Their hardware, "our" OS would be perfect. Well, except notch and lack of OLED - but, reportedly that's in the works too.
https://opencollective.com/asahilinux
the great thing is, you can!
"with full features"
the macbook pro uses MiniLed, in term of contrast, it's quite good, much better than ips.
Macbook pro display is one of the best laptop display.
And somehow it has way less blooming than other mini-led displays I have used. Not clear how they pull that feat off exactly
Most likely, they have more mini-leds and/or more ability to independently control them. Of course the localized "blooming" of mini-leds is a lot better on the eyes regardless than the all-around bloom of a backlit display.
(Better for the battery, too, if you can keep most of the screen dark.)
The M3+ GPU is also very different. So while it may be true that the driver development for M1/M2 is now more or less complete as OP says, future work along the same lines will very much be needed.
You set an ambitious goal and executed beautifully despite a very busy schedule.
Well done.
She just started working at Intel in August and has already accomplished more than most would in a year[1]. Incredible!
[1] https://rosenzweig.io/resume-en.pdf
Sad to see Asahi linux self-implode but great to see his work on the gpu side for years.
Huge respect for her. I consider myself a talented software engineer. I can’t get anywhere close to this accomplishment.
I never understood this project. Maybe I'm missing something, but the timescale is such that by the time they're done the product isn't even being sold anymore
At least with Panfrost it made more sense bc it still being used
M1 chip laptops can only be bought second hand at this point
I believe Walmart has a deal[1] with Apple to sell[2] M1 MacBook Airs. This has been the case for a year or so, so I don't think it's old stock. They have been in stock for since that date, and slowly getting cheaper.
[1]https://9to5mac.com/2024/03/16/walmart-m1-macbook-air-launch...
[2]https://www.walmart.com/ip/Apple-MacBook-Air-13-3-inch-Lapto...
Why would the product have to be available new for the project to be worth it? There are still many M1 chips out there, and this helps prolong the usefulness of those chips.
> M1 chip laptops can only be bought second hand at this point
New M1 Macbook Airs are still available at Walmart (maybe elsewhere). But even if not, who cares? People are still writing code for computers that haven't been sold since the 1980s.
Hum.. I can still buy a 100% hardware compatible NEW C64...
Lucky you, Intel.
What a legend in the making!
Im glad she stepped away from Asahi linux. Its absolutely great from a techincal perspective and the progress that team has made, but talented people like her shouldn't be trying to reverse engineer software/hardware from shitty anti-consumer company that can make the entire project work in a heartbeat by publishing documentation, in lieu of building better stuff from the ground up.
Reverse engineering requires a different mindset and somewhat different skill set than “forward” engineering. I’ve met people who were happy to only do reverse engineering (to figure out what make things “tick”) without building anything new.
If it was up to me, 2 years of successful reverse engineering (of a variety of projects/products) would be a requirement to be called an engineer. You learn a lot from working things that you can’t learn from a book (and without having to do the mistakes yourself first…)
Just to make it clear: I am not implying anything about Alyssa - just stating an observation based on my own experience.
> in lieu of building better stuff from the ground up
To be fair, even if you have the best CPU and GPU designers, it's not as if you can call up TSMC and have them do a run of your shiny new processor on their latest (or even older) process. You can't fab them at home either.
Fortunately for her, Intel - her new employer has "fabs at home". Though on older nodes, TBF.
Intel's GPUs are manufactured on TSMC though.
Even with proper documentation, there still would have been loads of work to get M1/M2 GPUs working on Asahi Linux. Writing GPU drivers worth a damn is about as difficult as targeting a compiler to a new CPU architecture. It would not be "in a heartbeat".
Honestly kind of heartbreaking to see her leave asahi Linux. She has done insane work building the vulkan driver from scratch. I wish her well working at Intel. If I ever buy an Intel GPU I can rest much easier it will work well on Linux. If she is working on the Linux driver stack that is.
There isn't really anything left to do for her - everything missing (including work on the newer graphics chips) can be somewhat easily done by less talented people, building on her work.
She did the challenging stuff she cares about. One aspect of nerd brain often is that you can hyperfocus on challenging stuff, but can't get the motivation to work on stuff you don't care about - and even what would be a 20 minute task can end up taking days because of that. It's great that she has the self awareness to set goals, and step away once they're done.
I didn't have that in that age - and still sometimes struggle. I was lucky enough that my employer back then recognized my issues, and paired other people with me for doing the stuff I was not interested in, and now usually manage to load those issues onto other co-workers by myself.
This is of course great as long as you can find enough "challenging" work to perform, but any successful project is going to involve a whole lot of seemingly "boring" work. A big part of true maturity and professionalism is being able to find the interesting challenge even in these more run-of-the-mill tasks and successfully engage with them.
(Mind you, I'm not talking about a matter of inborn temperament or character, much less a moral flaw! Rather, finding the compelling challenge even in "boring" tasks is a valuable skill and situational tactic that anyone should explicitly learn about and aim to acquire as part of becoming a mature professional, not a matter of morality or somehow being dismissed as "lazy"!)
Sounds like ADD to me. Easy to be labeled as "Lazy" etc.
Yeah, no way there’s any evolutionary fitness to having a brain that only works on problems it finds worthwhile. /s
M3, M4 and soon to be M5 are ready to be cracked open :). From what I understand they are actually different somewhat hardware wise. So it's really not like there would not be a continuations of this work. But of course it's natural to want something else after years of working on the project.
Take into account that she's focusing on the 3D stack, not the overall hardware. Even with hardware differences there's a good chance it's not different enough to make it an interesting new challenge.
Given the features that have been advertised for the M3+ graphics and compute stack, there's rather a good chance that it is different enough to create big, new challenges for third-party support.
Yes I meant the GPU specifically not the hardware in general. An example is the support for hardware ray tracing in M3 and beyond. In some now deleted french fediverse post Alyssa indicated M3 has a new architecture even.
Great work
Great stuff
Does Xe HPG compete with NVIDIA?
Xe HPG is also known as A750 and B580 so yes, it competes with the 3060/4060/5060.
Congrats on a job well done.
[flagged]
> I wish i had half of the energy of this guy
Trolling will get you banned here, so please don't.
i swear i'm not trolling, i had no idea the author is woman.
Sorry for misinterpreting you! All we can do is pattern-match, and sometimes the pattern doesn't match.
not your fault, she is trans so sensitive subject. She generally hides references to it, but her picture is on her site linked by others.
[flagged]
Moderation is guesswork.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
i think it's a woman?
https://rosenzweig.io/