Earlier quoted context omitted.
[flagged]
Please don't link to Asahi Linux on HN - Asahi doesn't like it.
Apple unveils M3, M3 Pro, and M3 Max
971–980 of 1001 posts
Re: Apple unveils M3, M3 Pro, and M3 Max
#972Earlier quoted context omitted.
I'm wondering now if Apple tracks this sort of stuff in their released machines. I know my iPad always asks me if I want to send stats to Apple (and I always say no). So let's say that enough people do, then do they have a good idea of how often all the performance cores are used? Max memory B/W consumed? Stuff like that. Back when I was at Intel there were always interesting tradeoffs between available silicon/therm…
Apple shares anonymous usage data by default on all their operating systems and users are asked again on every major update. Given that there has never been any public incidents about it and what we know about similar defaults I would be surprised if Apple is getting less than 95% opt-in rate. But I suspect at high-end they only really care about the performance of a few dozen professional apps e.g. Logic or Final Cu…
It's not a default because users must choose yes or no. So there basically is no default.
Re: Apple unveils M3, M3 Pro, and M3 Max
#973What should I buy? Honestly it’s all confusing as hell. 8gb ram seems like too little but increasing it is $$$$.
Re: Apple unveils M3, M3 Pro, and M3 Max
#974Earlier quoted context omitted.
My Windows desktop has generated significantly more e-waste per year than the Macs I have owned. Full stop. I've upgraded SSDs from 500gb to 1TB to now 2TB, upgraded graphics cards, and replaced a massive broken aluminum and copper heatsink. All in the last two years. I've had the same MacBook, in a similar timeframe. It gets just as much use. Its still rock-solid. Your concern is a hypothetical one. The inability to…
Sounds like your own problem. Apparently you like to replace and tweak things. Most people buy a good computer and are done with it, and if they ever upgrades something, the frequency is much lower than you. Don't blame your problem on others.
But, that's my point? If most people just buy a good computer and are done with it, which I agree to be a correct statement, then why does upgradeability matter from the perspective of global e-waste generation? Outsizing the impact of upgradeability, by literal orders of magnitude, is Reliability and Longevity; buying less in the first place.
The problem that I have is: memory modules and SSDs aren't exactly reliable, and need to be replaced. One of those SSD upgrades was a Want; the other was a Need. Hypothetically, Apple should be subject to the same problem; and they are, to some degree, but what we're talking about are degrees. If their machines need to be repaired X% less often, but those repairs do require the replacement of an entire logic board; there's a value of X where they are generating less e-waste per user than modular solutions.
Sure, that's a lot of `if`s; but everything I've seen suggests they're on the right side of that equation. Especially considering their repair loop is closed, and they're obviously very proud of their considerable recycling capabilities.
Re: Apple unveils M3, M3 Pro, and M3 Max
#975Earlier quoted context omitted.
I'm also still running an i9. I can't justify buying a new one since my 2019 model is still running strong!
I can't relate. I would probably have upgraded my i9 as soon as the M1 Pros were announced if it hadn't been my employer's property. I had constant thermal throttling from the instant I booted the damn thing. Worst laptop I ever had.
Mine runs warm a lot. I was looking through M1s on eBay today, I might pull the trigger. I just have a hard time spending 4 grand on a new laptop when the one I have does everything I need it to.
I had an M1 at work for a while, it was amazing and I loved it.
Re: Apple unveils M3, M3 Pro, and M3 Max
#976Earlier quoted context omitted.
No. It's nowhere near as powerful as desktop Nvidia cards. It's about as powerful as a PS5.
It could be theoretically as powerful as a ps5 but not nearly as performant. Game studios are not spending any time optimising for apple silicon and dumping tons of dev hours to make sure it runs smooth on the console. Sony uses its custom graphics API (GNM and GNMX) and custom shader language and x86 platform. Xbox uses DirectX. Apple went in on its own route with Metal. That widens the gap.
Game studios release games for the iPhone. Which means they do spend time optimizing for Apple Silicon and dump tons of dev hours to make sure things run smoothly.
Re: Apple unveils M3, M3 Pro, and M3 Max
#977Earlier quoted context omitted.
From what I understand this is about assigning GPU resources (such as register files and other on-core memory) to shaders. Imagine you have a complex shader that has two paths, a frequently taken fast one that needs X bytes of on-chip memory to function and a rarely taken slow path that needs Y bytes (Y > X). Usually this stuff is statically partitioned, so you have to provision Y bytes for the slow path to run this…
in tinygrad, the llama2 model with Metal runtime produces 1k kernels. this means we have to compile them all, leading to both startup and runtime costs from repeated compilations and buffer bindings. someone suggested using one megakernel to call the rest. could dynamic partitioning help here?
Short answer: yes.
Long answer:
If the hypothesis of the parent comment is correct, then the following should be true: joining shader logic paths or parts with very different resource usage, notably register usage, shall no more cause occupancy issues which on traditional architectures could be sometimes resolved by artificially splitting the logic into multiple different kernels with the sole purpose of making it possible for low-resource usage portions of the logic to run with higher occupancy.
If the case, it is of course beneficial not only because such splitting burns development hours (and can be potentially error prone), but also because such splitting introduces overhead of its own, for reasons anywhere from having to repeat some of the calculations in more than one separate kernel (instead of reusing the results) through having to store and subsequently reload intermediate values to communicate between parts, to having to pay the overhead of launching and synchronizing additional kernels.
It stands to reason, however, this is not going to help in those cases when it's not the occupancy that's the problem, but rather sparse SIMD/wavefront utilization of ALU resources is: in cases when the control flow is sub-wavefront divergent, but splitting the code into multiple kernel launches allows for compactification of SIMDs/wavefronts.
Furthermore, the joined shader might still fall a little bit short of expectations anyway because joining code together not always results in the compiler successfully identifying and eliminating redundant or repeated calculations, and/or allocating resources better due to having access to the entire program at once. It's usually the case, but occasionally the opposite happens - sometimes joined code results in compiler's optimizations taking a different path and producing code that is actually worse than it would have been if different parts of the code were compiled separately. The risk of that increases if the frequency with which the shader execution takes specific paths in the program is unusual (statistically not typical where statistics are taken over the whole world of programs across space and time :-) ) and the compiler ends up mispredicting which paths are low- and high-probability.
Re: Apple unveils M3, M3 Pro, and M3 Max
#978Earlier quoted context omitted.
Intel straggler here. And I'll remain one until Linux runs flawlessly on the newer generation. I have an older Thinkpad as my main workhorse and a now quite old Macbook air for on the road, it's battery has been shot for a long time but it still works quite good when plugged in and I'm too much of a miser to have it replaced. But if there is a way to run a full Linux distro on the newer hardware then I'll probably bu…
Replace it yourself using iFixit instructions and parts. Just screws and plugs, easy peasy.
was going to say this as well. The battery replacement process is pretty easy.
Re: Apple unveils M3, M3 Pro, and M3 Max
#979Earlier quoted context omitted.
I think what Apple is pushing for is computing efficiency. It still gets faster but with much less power. Focusing on performance solely would be the wrong way to evaluate these chips. https://www.tomsguide.com/news/apple-m3-chip
This drives me crazy. Apple plays the market like Nintendo. Pick something that no one cares about, do it better than anyone else, and make a big deal about it. I dream of a world where instead of a marketing company becoming a top 3 tech company, a tech company would have. Wonder what they would have done for their laptop... Or maybe this is just an inevitable outcome of capitalism/human biology where a veblen goods…
Considering that Apple has been significantly more innovative (tech wise) than pretty all of their competitors I’m not quite sure what this tells about them.
Re: Apple unveils M3, M3 Pro, and M3 Max
#980Earlier quoted context omitted.
I personally struggle with the 14". It feels too small to be productive on, at least for coding. Anyone else experience this? And yet, the MBA's screen in comparison is serviceable and nice, but nothing outstanding. That's the case for the MBP 14 (when the 16 is just too large and bulky).
The old 15” was like the perfect dimensions. It practically had the footprint of the present 14”, maybe even smaller. Apple made a big deal about how their new chips run so cool, yet they made the pro laptops as fat as they were in 2012 again so clearly thermals were an issue.
Aren't the new 16" laptops the same dimensions as the old 15" ones? I thought the 16" display was simply because they were able to shrink the bezels on the display enough to get an extra inch on the diagonal. Other than the rounding on the edges, my M2 16" Pro feels about the same size as my old Intel 15" one.