Live data from Hacker News

Apple's follow-up to M1 chip goes into mass production for Mac

asia.nikkei.com

531–540 of 661 posts

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#531
post #17

Earlier quoted context omitted.

M2 seems very unlikely to me because that will create product misinformation easily. Imagine the following M1 -- 4 A14 cores M2 -- 8 A14 cores M3 -- 4 A15 cores That "third generation" sounds better, but is actually inferior to the second generation. X1 or M1x seem much more likely. It's the same reason why they made an A12x instead of calling it A13. They probably need 3 designs, but economies of scale begin to be a…

Another possibility is that they skip M1x altogether; if the Pro machines are coming in the autumn and not for WWDC, then the frame will be closer to the iPhones and it would make sense for them to use that later tech. M1 (winter 2020) — 4x A14 cores M2x (autumn 2021) — 8x A15 cores M2 (winter/spring 2022) — 4x A15 cores etc. There’s really no reason for the naming and timing of Pro machines to lag behind consumer ma…

There’s a couple good reasons for a M1x using A14 cores.

With new nodes, it takes time to get things right. A smaller chip is generally preferred as easier to work with and test. It also gives the node more time to stabilize so your big chips get better yields.

Addressable market is another factor. There are over a billion iPhone users replacing devices every 1-3 years on average. There’s around 100 million Mac users as of 2018 replacing their computers every 5-8 years and that being split between 2 and 3 major market segments.

That means making over 250 million A14 chips, but only around 15 million M1 chips (aside from iPads which are there to improve economies of scale by reducing two designs to one). The M1x probably addresses around 4.5 million computers and the M1p or whatever they call it probably addresses under a half million machines.

A lot of people wanting the performance boost or fearing outdated, unsupported x86 machines will probably cause an increase in M1 demand, but that will be followed by a decreased demand for M2 and maybe even as far as M3 or M4 while demand catches up with the altered purchase cycle. Apple is no doubt well aware of this.

In any case, it’s best to try a new node with a smaller part where you know you need a million wafers anyway. Once you’ve optimized and refined a couple steppings, you move to the larger chip with smaller demand where yields are worse in general.

Personally, I think they would greatly benefit from a unified chiplet design starting with something close to the A14 then scaling up.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#532
post #429

Earlier quoted context omitted.

Honestly wouldn't be surprised if it came out that they started working (at least theoretically) on Apple Silicon when they transitioned to Intel in the first place and it just took this many years for all the pieces to be ready and lined up.

I think Apple would have been perfectly happy buying CPUs from Intel as long as Intel kept their end of the bargain up. After the PowerPC fiasco and IBM leaving Apple high and dry, I have zero doubt that there was a contingency plan always under way before the ink even dried on the PA Semi acquisition, but it wasn't probably a concrete strategy until about the third time in a row Intel left Apple high and dry on a be…

I actually wonder if the end-game isn't going to be a Intel/M1 hybrid on the high-end ... a system with multiple chips and architectures.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#533

Earlier quoted context omitted.

I think it’s mainly for Mac / iOS development CI systems like for CircleCI or Mac Stadium. The hardware just seems too expensive for it to replace more generic arm services that AWS is offering.

On-site music and video production. You can rack it in a portable rack with your other racks of music equipment and it fits in the back of a van.

Well, if your portable rack happens to be way deeper than most audio racks…

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#534
post #145

Earlier quoted context omitted.

In a laptop?

Maybe the commenter has a very interesting use for it, but why would you buy a 256 GB RAM machine (which isn't a lot of memory for this arena either) to develop the model, instead of using something smaller to work on it and leasing a big cluster for the minutes to hours of the day that you'll need to train it on the actual dataset?

I have 128GB in my desktop, which is the same amount of ram that our cluster compute nodes have. I use it for Python/pandas forecasting work.

The extra ram in my machine means I can work on the same datasets we use in prod locally, which is a huge productivity boost. Most of our developer time is actually spent building datasets. loading a random sample doesn’t really work when doing time series or spatial transforms and using a time or space limited subset makes description (graphs, maps) a chore. More ram is a huge productivity enabler when working with in memory tools like pandas.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#535

Earlier quoted context omitted.

Hmm, but wouldn't features like wifi, all the cameras, the touchscreen, Cellular data, NFC, touch ID, speakers, bluetooth, and so on all come in via IO as well? You're right that they only have one port, but they still connect to a lot of different components.

The problem with this analogy is that NFC, touch ID, speakers, and bluetooth are all in the MB/s or less range. On the desktop side, you have to deal with 10gb/s ethernet and 50-100gb/s for ram. It's just a whole different ballgame.

Ah, that makes sense. Do iPhones connect to their RAM differently? I know it’s physically different than desktop RAM (for obvious reasons), but I don’t really know enough about hardware to know how it connects.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#536

Earlier quoted context omitted.

You know, I believe you on the 2019 - I've heard this from more people than I care to admit, and makes me glad I skipped that year. I think from 2016-2019 was a rough era for the MacBook Pro when you factor in price for performance; still great machines, but they didn't feel as polished as the 2015 MBP or the 2020 M1's. Edit: year banding.

My 2016 MacBook Pro has been an absolute pleasure to use. It's small, light, great battery life, plenty of performance, I consider it the second incarnation of my favorite laptop ever, the Titanium PowerBook G4. Except for the keyboard. Been replaced three times. It's defective by design. Such a letdown.

I have the opposite experience with my 2016 MacBook Pro. It feels really sluggish to me now. 2 cores is nowhere near enough, and the integrated graphics chip doesn’t feel fast enough for everyday tasks if you have external monitors connected. I’m on a video call in discord the machine lags so much it’s unusable. Xcode with swift is basically useless - it lags so much sometimes doing who knows what that it drops keystrokes. I had the monitor stop working because of the design fault right before two back to back conferences. And the keyboard has been replaced once and is now having issues again. Compiling rust on the machine is about 8x slower than my new ryzen based development machine.

We’re in a golden age of cpu improvements. A dual core 2016 mbp isn’t good enough for me any more.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#537

I haven't yet used an M1 mac but based on what I've read about it I have fully bought into the hype train. Hoping my next laptop will be a M2-powered MBP, assuming they can increase the maximum RAM to at least 32GB.

Why do you want 32GB? With my 8GB m1 Mac mini I was surprised to get away with as little memory as I did. I felt that the M1 needs far less memory than x86 to feel snappy.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#538

Earlier quoted context omitted.

In the consumer space there are only 2 paths for a viable competitor to emerge. 1) a coordinated ecosystem attack or 2) an individual company doubles down. There are things that lead me to believe 1) can't happen. Google ultimately acquiring Fitbit instead of ecosystem compatibility and integration. It seems like the giants are needing to acquire the pieces to build their own Apple, rather than partner with them. Als…

How far behind Apple are AMD at the moment? If ryzen chips were optimized for low power and manufactured on TSMC’s 5nm process like the M1, what sort of performance difference would we be seeing?

TSMC advertises 5nm as +15% performance OR -30% power consumption. Either one of these won’t be enough to bridge the gap in mobile. Even throwing in another massive 20% IPC improvement wouldn’t do the trick.

On the flip side, once you go bigger the game changes. If you increase the M1 power consumption to amount for interconnects, more memory lanes, more IO, etc the M1 seems to offer worse power to performance that shouldn’t be too unexpected. ARM just announced they are splitting into V1 (performance above all) and N2 (best power, size, performance combo)

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#539

Earlier quoted context omitted.

+ Mac OS checking every binary you run over the internet.

It's slightly more nuanced than that - you can of course just block it from doing so, and there's certainly an argument for it being updated to not need a network call each time, but phrasing it like this makes it sound worse than it actually is. I'll just quote Jeff Johnson, who's looked into this and written about it - his comment[1] on this is post is quite useful: https://eclecticlight.co/2020/11/25/macos-has-che…

You added detail, yet didn't refute the statement. That it isn't easy to simply and specifically turn off is not acceptable.

Re: Apple's follow-up to M1 chip goes into mass production for Mac

#540

Earlier quoted context omitted.

Apple isn't harnessing magic here. The opportunity for competition exists, but everyone else is just sitting on their asses while Apple's chip team puts in the work. If you want to beat Mx, you have to think outside the box, just like how Apple thought outside of x86. That being said, I do enjoy the shift to ARM as the dominant arch. x86 has far too many decades of cruft behind it. I don't think Apple will take over…

"The opportunity for competition exists" is the most true statement and response to everyone grumbling about Apple dominance. Somewhere, somehow, the ecosystem of competitors has failed to execute, got lazy, etc and now is looking to regulators to bail them out. It makes me a bit sick. Apple is nothing more than a computing stack wrapped in a portfolio of consumer products. They happen to see this future early, inves…

> the ecosystem of competitors has failed to execute

I see what you did there!

It's probably because the ecosystem is multi-threaded and there's lots of deadlocks and unprotected reads/writes...

Post reply on HN