Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

An Update on Apple M1/M2 GPU Drivers

lwn.net

Post reply on HN
Print view View original

251–260 of 291 posts

Re: An Update on Apple M1/M2 GPU Drivers

#251

Post by outworlder »

Cthulhu_ wrote: post #142
whitehexagon wrote: post #128

What I really respect is the dedication to completing support for the M1/M2. Too many projects get dropped the moment the next shiny ray-traced toy comes along. I know this type of work can be challenging to say the least. My own dabble with Zig and Pinephone hardware drivers reminded me of some of the pain of poorly documented hardware, but what a reward when it works. My own M1 was only purchased because of this pr…

One thing I noticed in the M4 macbook announcement comments was how many people were happy with their M1 laptop, and second, how many people kept their Macbooks for nearly a decade; these devices are built to last, and I applaud long-term support from Apple itself and the Linux community. Second, since it's open source, Apple themselves are probably paying attention; I didn't read the whole thing because it's going o…

> how many people kept their Macbooks for nearly a decade; these devices are built to last, and I applaud long-term support from Apple itself and the Linux community.

Anecdotal but I have a White Macbook from 2010. It's sitting on a shelf not because it doesn't work (minus the battery), but because it's too outdated to be of much use. And there's a small crack in the case.

I have a Macbook Pro from 2016. I still whip it out from time to time when I don't want to use my latest one for whatever reason (say, network troubleshooting). It works fine. Even the battery still holds charge. If those two had USB-C (and charging over that port) I'd probably use them more often. Their keyboards is also pleasant (since they are before the scissor key nonsense).

My company has thousands of Macbooks. It's rare that I see anyone at all with issues. They aren't perfect, but the build quality is far better than most PC laptops and so is the attention to detail. The price premium kinda blows, but the M line made them way better.

Re: An Update on Apple M1/M2 GPU Drivers

#252

Post by ho_schi »

bluedino wrote: post #235
ho_schi wrote: post #159

Earlier quoted context omitted.

ThinkPad X220 here. From late 2012 until late 2023 in service with Linux. It is still usable but finally replaced it by an X13 Gen3 AMD with Linux. The magnesium body is a bless and feels a lot better to the skin than aluminium. The HiDPI display is the biggest upgrade. The six row keyboard is sturdy but a downgrade from the seven row keyboard. I miss the notch to open the lid. It got worse with Gen4/5 which now have…

I never kept my ThinkPads around long because of the screens (which they have fixed, somewhere around the T450 timeframe. The X220/230 had that godawful 1366x768 panel. A shame when the 13" Air at the time had a 1440x900 panel, which while it wasn't amazing with the viewing angles and colors, it was light years ahead of the screen in something like a T430.

ThinkPads come usually with two panel options: The cheap one and the good one. Often even a third or fourth.

Lenovo should only ship the good ones! If I buy a ThinkPad I want the good one.

I struggled to get the HiDPI panel for the X13 and ordered it from the replacement parts. Same for the X220, I replaced the TN panel by drumroll IPS panel.

Apple takes ridiculous amounts for memory or disk but you always get the good stuff with the base model. This makes it simple and reliable for customers.

Except keyboards

The ThinkPads always win, convex key caps, better switches and pressure/release point, travel way, better layout and a TrackPoint. In some models you can still replace it within 30 seconds (T14). With the X13 or MacBook it is a horror, requires removal of the mainboard. Not mentioning Apples failure with the TouchBar, which “feels” like the tried to save money on expensive switches by replacing them with a cheap touchscreen and selling this horrible downgrade as feature. And the infamous butterfly switches are the reason to avoid any used older MacBook (often defective).

Re: An Update on Apple M1/M2 GPU Drivers

#253

Post by _zoltan_ »

saagarjha wrote: post #129
jjmarr wrote: post #104

Earlier quoted context omitted.

I didn't even know this until now. I looked at her resume and I can't grasp how she's been doing an undergraduate degree from 2019-2023, meaning that writing an entire GPU driver is the side project she made during her degree.

There’s a lot of free time when you’re in college.

So much this. I wish I had a fraction of the free time I had when I was doing my degrees...

Re: An Update on Apple M1/M2 GPU Drivers

#254

Post by lowbloodsugar »

Der_Einzige wrote: post #221
bzzzt wrote: post #189

Earlier quoted context omitted.

Only if you insist on running software that needs more RAM, in which case you shouldn't have bought it.

Apple should not have sold devices with 8GB or less back in 2018. Them doing it in 2024 is a sign that they think their users are idiots.

I just bought the original M1 Air 8GB for my spouse from walmart for $650, to replace an aging intel Air 8GB. The difference is stark. The M1 Air is a pleasure to use, even with the inevitable library's worth of browser tabs.

Plenty of people have no need for more than 8GB. My spouse. My mom. My english-major son. Meanwhile, my M2 is saying I have 72GB in use, with IntelliJ #1 at 6.05GB. I would not be happy with 8GB. Sounds like you wouldnt be either. So don't buy one.

Re: An Update on Apple M1/M2 GPU Drivers

#255

Post by talldayo »

flohofwoe wrote: post #230
talldayo wrote: post #225

Earlier quoted context omitted.

Because a lot of the "cheap hacks" aren't even cheap. Famously, using real-time SSR to simulate ray-traced reflections can end up slower than ray tracing them from the jump, the same goes for high-res shadowmaps. Ambient occlusion is a lighting pass you can skip entirely if you render shadows the right way with RT from the start. If you keep stacking these hacks until you have every feature that a RT scene does, you'…

Yes, for some specific problems, raytracing definitely makes a lot of sense, but rasterization and 'geometry compression' via triangles also makes a lot of sense for other situations. I think the best approach will always be a hybrid approach instead of a 100% pure raytracing pipeline. The other question is how much of the raytracing features in 3D APIs need to be implemented in fixed-function hardware units, or whet…

From everything I've seen, hardware acceleration seems to be the only option. Software-accelerated RT existed for years and was mainly the reason why it was disregarded as a technique in the first place. Here's the OG M1 Pro, manufactured on TSMC 5nm, getting blown the fuck out by a 3070 Ti on Samsung 8nm: https://youtu.be/NgDTCNPm0vo

The architecture will probably improve on both Nvidia and Apple's side going forward, but in theory both GPUs should be easier to support for longer since they're not focused on accelerating obsolete junk like MSAA or HBAO. It enables a better convergence of features, it makes porting to-and-from Mac easier, and with any luck Apple might even quit being so sheepish about working with Khronos since they're not actively neglecting the Vulkan featureset anymore.

Re: An Update on Apple M1/M2 GPU Drivers

#256

Post by gigatexal »

dtquad wrote: post #39
gigatexal wrote: post #30

Is anyone else astonished at how much is missing in the hardware and how much is emulated?

The graphics pipeline in modern GPUs are mostly a thin low-level Vulkan/Metal-like layer on top of a massively parallel CUDA-like compute architecture. It's basically all emulated. One of the reasons GPU manufacturers are unwilling to open source their drivers is because a lot of their secret sauce actually happens in software in the drivers on top of the massively parallel CUDA-like compute architecture.

You know it’s not like Apple has a lot to lose in terms of market share if they opened even the specs of their hardware up to make folks like the Asahi team’s lives’ easier. What does Apple have in terms of the desktop market? 5%?

If anything working, fully baked desktop Linux working on ARM Mac hardware would drive sales and Apple would still get the profits. Besides their services are still mostly available too. Tux loving folks can subscribe to AppleTV, and music etc.

Re: An Update on Apple M1/M2 GPU Drivers

#257

Post by jrochkind1 »

TylerE wrote: post #246
jrochkind1 wrote: post #209

Earlier quoted context omitted.

Agreed, hard to repair (which is a problem), but my experience is owning one for 8 years, and another for 4 years now (with hopes to keep it another 6), which never once needed a repair.

The only Mac hardware failure I’ve experienced in over 25 years of ownership (My first Mac was black and had a G3 in it and ran OS 8), I’ve experienced a grand total of one hardware failure, and that was 15+ years ago.

Oh you know what, I realized I wans't accurate. The 8-year MacBook I had _did_ have a failure at one point. They replaced the entire keyboard/trackpad for me for free when it did.

Re: An Update on Apple M1/M2 GPU Drivers

#258

Post by School-Cotton »

hurutparittya wrote: post #178
Cthulhu_ wrote: post #142

Earlier quoted context omitted.

One thing I noticed in the M4 macbook announcement comments was how many people were happy with their M1 laptop, and second, how many people kept their Macbooks for nearly a decade; these devices are built to last, and I applaud long-term support from Apple itself and the Linux community. Second, since it's open source, Apple themselves are probably paying attention; I didn't read the whole thing because it's going o…

I'm always surprised when people speak highly of Apple devices here. While they do have certain advantages, there are some issues that should be dealbreakers for tech literate people. (in my own, possibly biased opinion at least) In case of Macbooks, it's the fact that they refuse to provide an official GPU driver for Linux and general poor support for things outside the walled garden. The Asahi stuff is cool and all…

> there are some issues that should be dealbreakers for tech literate people. (in my own, possibly biased opinion at least)

I know you admit right after that your opinion is biased, but it's almost ludicrous to assert that all the programmers and engineers using Macs and iPhones by choice must just not be tech literate.

> In case of Macbooks, it's the fact that they refuse to provide an official GPU driver for Linux

MBPs are so much better than any other laptop that, with a few caveats[1], running Linux in a VM on a top-of-the-line MBP is a much nicer experience than using Linux natively on any other laptop. So while it'd be nice if there were more first-party support for Linux, it's certainly not a deal-breaker for "tech-literate" people. (Not to mention the fact that there are "tech-literate" people who use macOS and not Linux, so it wouldn't matter to them at all).

> general poor support for things outside the walled garden

macOS isn't a walled garden, so I don't know what you mean. You can download any software you want from anywhere you want and run it on your laptop, and Apple doesn't do anything to try to prevent this.

> The Asahi stuff is cool and all, but come on, is a 3.4 trillion dollar company really going to just stand there and watch some volunteers struggling to provide support for their undocumented hardware without doing anything substantial to help? That sounds straight up insulting to me, especially for such a premium product.

Now it's unclear whether your point is "I don't understand why people use Macs because there are objective drawbacks" or "I don't think people should use Macs because Apple does stuff that I find annoying". You're blending the two here but they are meaningfully separate points. I've discussed the practical point already above, but as for the stuff you subjectively find annoying: surely the only real answer is that lots of other people just subjectively don't care as much as you.

> For iphones, it's the fact that you are not allowed to run your own code on YOUR OWN DEVICE without paying the Apple troll toll and passing the honestly ridiculous Apple Store requirements.

I don't care about this at all. I've never wanted to run my own code on my own iOS device except when I was working on iOS apps and had an Apple developer account through work. I, like the vast majority of people, use my phone as a browsing/messaging/media consumption device, not as a general-purpose computer.

If Apple tried to prevent me from running my own code on my own MacBook, it would be a deal-breaker, but as I already said above, they don't.

In conclusion I think you've confused "tech-literate person" and "geek who wants to be able to tweak/configure everything". Indeed there is a large overlap between those sets, but they're not quite the same thing.

[1] https://news.ycombinator.com/item?id=41997107

Re: An Update on Apple M1/M2 GPU Drivers

#259

Post by mft_ »

helsinki wrote: post #229

Alyssa's solution to the 4KB vs. 16KB page size discrepancy by running everything in a virtual machine feels like both a clever hack and a potential performance bottleneck. It makes me wonder about the long-term implications of such workarounds. Are we reaching a point where the complexity of bridging these gaps outweighs the benefits, especially when dealing with proprietary hardware designed to be a closed ecosyste…

> Apple's approach to hardware design warrants criticism. By creating GPUs with limitations that hinder standard functionalities like tessellation shaders and using non-standard page sizes, Apple places unnecessary obstacles in the path of developers. This closed ecosystem not only complicates the work for open-source contributors but also stifles innovation that could benefit all users.

Apple designs its hardware to suit its own ends, and its own ends only. It's obvious to everyone here that this supports their closed business model, which actually works for them very well - they make excellent hardware and (some software flakiness more recently notwithstanding) the median user will generally have an excellent time with their hardware + software as a result.

So they're not placing "unnecessary obstacles in the path of developers" at all by designing their hardware as they do - they're just focused on designing hardware to suit their own needs.

(Also note that if their hardware wasn't excellent, there wouldn't be such interest in using it in other, non-Apple-intended ways.)

Re: An Update on Apple M1/M2 GPU Drivers

#260

Post by manmal »

spockz wrote: post #106
manmal wrote: post #63

Earlier quoted context omitted.

Reportedly, the lack of M3 support so far is because there is no M3 Mac Mini which they can use for CI.

I’m sure if that is the issue we can together find a way to sponsor a MacBook with an M3? Or does it specifically have to be a Mac mini?

I can only guess, but it might be a requirement to have the machine hosted by a cloud platform.
Post reply on HN