Apple's follow-up to M1 chip goes into mass production for Mac
asia.nikkei.com
asia.nikkei.com
Apple just showed to the world how powerful and efficient processors can be. All that with good design.
Customers are going to demand more from Intel and the likes.
Just imagine Apple releasing the Mx chips for server infrastructure. Intel ought to be sweating bullets now.
edit: a word.
It's been available for a year: https://aws.amazon.com/about-aws/whats-new/2020/08/amazon-ek...
In short, it's for media production, not computing. You put it in a media rack.
Right now I've got IDE to SATA adapters in mine, which leaves just enough space to also squeeze 2.5" sata ssd into 3.5" drive caddies.
Remember when Apple made the Power Mac QUAD, to indicate that the system had FOUR processors, because "men should not live of single thread performance alone"?
> "With quad-core processing, a new PCI Express architecture and the fastest workstation card from Nvidia, the new Power Mac G5 Quad is the most powerful system we've ever made," Philip Schiller, Apple's senior vice president of worldwide product marketing, said in a statement.
Of course they can't advertise the multi core performances, because they are lower compared to similar systems, just like with the Power Mac they could not advertise that the system was greener, because it wasn't.
AMD is the one left without an "efficiency core"… but I wouldn't be that surprised if they somehow manage to squeeze almost the same efficiency out of just low power modes on the same Zen core, which would leave them with better multi-core performance as usual.
> M1 and follow-ups are about single thread performance, the only actual performance most consumers are likely to notice
I beg to differ, multi core performances will be very much noticeable as well.
Especially in this heavy media era
video encoding, image editing, music production etc. are things that nowadays every kid does
single core performances are good only for benchmarks and marketing
The performance boost there is extremely noticeable on new M1 chips.
The instant turn-on. No fans. Safari alone feels like a completely different beast and a friend and I spent about two hours running through some of the most demanding web sites, side by side against my 2018 15” MBP.
Holy crap. Just...all the subtleties.
I’d actually say it’s easier to notice the immense general speed-up than I notice the difference between a 4min and 2min render time in Final Cut.
Dragging 4K clips around in FCP was infinitely smooth; and it made you immediately be able tell the UI/UX itself was dropping frames and needing to catch up to itself sometimes.
These are things you don’t notice until you’ve tried the M1 for the first time.
It truly is an undeniably insane piece of engineering. They killed it.
2 - 90% of the laptops sold Worldwide aren't Apple laptops, most of them are sub $500 laptops. Amazon best selling is ASUS Laptop L210 Ultra Thin Laptop, 11.6” HD Display, Intel Celeron N4020 Processor, 4GB RAM, 64GB Storage priced at $199
Basically only Mac users who upgraded to M1 are going to notice the difference
Apple’s share of the total PC shipments is closer to 15% now and their share of the laptop market is even higher.
This is perhaps the wrong metric to look at. What % of total profit is Apple capturing? Several years ago the accepted wisdom in phones was that Apple was getting ~30% of the unit volume and 90% of the profit.
Worldwide 9.6%.
I rounded it at 10%.
> Apple achieved a global desktop/laptop market share of 9.6% and ranked fourth behind Lenovo, HP, and Dell. In the United States, Apple accounts for 12.7% of the PC market
This kind of competitive pressure will inspire a response from Intel and other firms.
The result I would predict and hope for in a few years would be better chips from everyone in the market.
Intel has certainly dominated consumer computer sales over the past decade, and until 4 years ago they were largely still selling the best chips for most consumer use cases (outside of mobile.) Intel had several missteps, but I don't think their dominant position was the only source of their problems, or simply that they thought they didn't have to try. They legitimately had some bad decisions and engineering problems. While the management is now changing, and that might get their engineering ducks in a row, the replacement of Intel with Apple Silicon in Apple's products is not likely to be some kind of "waking up" moment for Intel, in my opinion. Either they'll figure out their problems and somehow get back on an even keel with other chip designers and fabrication, or they won't.
Meanwhile other competitors in x86 and ARM have also have a short-term history of success and failure, again regardless of what Apple is doing. And the timelines for these plans of execution are often measured in the scale of two to three years, and I'm not seeing how Apple successfully designing CPUs would change these roadmaps for competitors.
For everyone involved, particularly those utilizing TSMC, there are benefits over time as processes improve and enable increases in performance and efficiency due to process rather than design, and the increased density will benefit any chip designers that can afford to get on newer processes.
I guess if I'd attempt to summarize, it's not clear who is motivated and able to compete against Apple in ARM design. In other words, is there a clear ARM market outside of iOS/macOS and outside of Android (where chip designers already compete)? And in the Linux/Windows consumer computing space, there's going to be a divide. Those that can accept a transition to macOS and value the incredible efficiency of Apple Silicon will do so. Those that continue buying within their previous ecosystems will continuing comparing the options they have (Intel/AMD), where currently chips are getting better. AMD has been executing very well over four years now, and Intel's latest chips are bringing solid gains in IPC and integrated GPU performance, though they still have process issues to overcome if they wish to catch back up in efficiency, and they may also need to resolve process issues to regain a foothold in HEDT. But even there, where AMD seems pretty clearly superior on most metrics, the shift in market share is slow, and momentum plus capacity give Intel a lot of runway.
The only other consideration is for Windows to transition to ARM, but there's still a bit of a chicken and egg problem there. Will an ARM chip come out with Apple Silicon like performance, despite poor x86 emulation software in Windows when run on ARM? Or will Microsoft create a Rosetta-like translation software that eases the transition? I'm not clear on what will drive either of those to happen.
trading speed of development with 'hardware optimization'.
Mac being a small market, some users are still opiniated/loud enough to reward nicer/native UI vs cross-ish platform lowest common denominator UI. Maybe there is hope Arm mac stays fast.
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-checked-app-si...
>The request to http://crl.apple.com/root.crl is simply checking the revocation status of Apple’s own Developer ID Certification Authority intermediate signing certificate. If you examine the cert in the System Roots keychain, you can see that URL under CRL Distribution Points. This request contains no information specific to third-party developers or apps. In contrast, there’s no CRL for third-party Developer ID leaf certs signed by Apple’s intermediate cert. Their status is only available via OCSP.
Notably, this:
>This request contains no information specific to third-party developers or apps.
https://eclecticlight.co/2020/11/25/macos-has-checked-app-si...
The parent comment is incorrectly stating that each request to Apple is checking a binary, and it's not. This has been well documented both here on HN and across the web.
Most javascript developers are already using babel-et-al pipelines to build electron apps, which are already transpiling between major variants of javascript, and I wouldn't at all surprised to see a thing where it gets compiled into WebAssembly rather than interpreting javascript. I also think there's a thing, right now, where it's possible to build electron apps with Rust+WebASM; I'm not sure, but I think the main thrust here is it definitely would eliminate a huge chunk of the slowdown.
I guess the main takeaway is just that the development revolution that's happened recently is mainly about how insanely good browser dev tools have become, and not about javascript - javascript was just along for the ride. As an aside - I recently saw a video of someone demoing one of the old Symbolics LISP workstations, and I was shocked to realize how much they had in common with a modern browser console - specifically of being able to inspect all of your gui components, live, and look at all the properties set on them. It's provided a hell of a way for me to explain what the deal was with all the old gurus in the 80s "AI Winter" diaspora who were pretty butthurt about having to move from programming in that environment, to having to write C on DOS (or whatever else paid the bills at the time).
tl;dr it has really sick floating point double performance which directly translates to JS performance
all other numbers are floats
Probably depends on the IDE and the language. My impression is that newer computers (especially in the "consumer" / "general public" market) have especially improved efficiency. They're not much faster, but they last longer on smaller batteries.
My late 2013 15" (2.3 GHz / 16 GB RAM) works great with Pycharm for moderately sized projects. It's even usable for personal Rust projects (with the intellij-rust plugin).
For Rust it's somewhat slower than my i5-8500 desktop but not by much. For incremental compiles, I don't feel like I'm waiting around more. The i5 mainly wins when it can use all six cores.
It's however quite a bit faster than my work HP ProBook with an i5-8250U which is a much newer machine (2018 I'd say).
Aside from battery life, which is somewhat lesser, all in all the mac is a much, much better machine than the HP, especially for "creature comforts": gorgeous screen, no misaligned panels rubbing against my wrists, great trackpad, no random coil whine, inaudible fans unless it starts compiling for a while, no background noise on the headphone out, integrated optical output when I want to use my home stereo.
I use mainly VSCode these days though and Firefox, but even Unity game engine works just fine on this machine.
Any IDE I also throw at this machine works just fine. Only things that feel slow, are things that need a beefy GPU.
My work 2019 16-inch MacBook Pro actually doesn't feel that fast considering how fucking expensive it is.
I've also had issues with the laptop waking from sleep, where it'd either hang for up to minutes, or just outright kernel panic. Big Sur seems to have fixed the kernel panics but waking from sleep still feels really slow.
Quiet and cool with about 11 hours of battery life when writing, reading or some simple productivity stuff.
Fast but loud and hot when doing anything else with about 3 hours of battery life. Even the Touch Bar (which I don’t hate, but I also don’t love) is uncomfortable to the touch.
It seems there is no in between. I’m generally happy with the machine, but I’m very interested in what’s next for the big MBP.
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.
Except for the keyboard. Been replaced three times. It's defective by design. Such a letdown.
We’re in a golden age of cpu improvements. A dual core 2016 mbp isn’t good enough for me any more.
Just wondering, don't those cores still have HyperThreading or such, resulting in actually more cores than just 2 physical ones ?
We have taken careful care in order to write fast and memory friendly javascript if possible, avoiding using slow features of the language (the list is long, but things like forEach loops, ineffecient data structures etc) and taking care to profile and optimize.
Result is an application that feels almost as fast as native and doesn't really consume so much memory even, although we are doing canvas based graphics realtime.
My suspection is that many web-developers (and thus, qualified to developer for Electron) just don't have the tenacity or background to write efficient code.
You can write slow-ass molasses javascript code very easily. Just take somebody who has done webdev for maybe like 2 - 3 years and doesn't have any other deeper CS background. Watch what kind of memory inefficient and slow code structures, especially with javascript where you don't really understand how heavy an operation like map() can be in worse cases, or where you are creating copies of your data multiple times for example, and voila, you have a slow-ass memory-hogging electron application.
Maybe I should do a blog post about how our Electron -application is performing just to show people that you can write fast code using javascript also. But it takes skill and time, and maybe in this current day what matters is just cranking out builds that work somehow.
We have been developing this application since 2011 already, and I also develop in C++ and 3D graphics, so performance has been something I've had to think.
But yeah, there are many guides on what to avoid in JS already, how to optimize for speed and memory. But it is not definitely an easy thing to realize, as it's not inherently visible what can be slow and what not. Most blog posts at this time talk about how to optimize for the V8, and I'm not really interested in that so much, as it can be taunting to try to understand how V8 works for example.
One of my former employers had the great idea of hiring hundreds of "senior" JS devs and redoing the entire frontend of the website. When it launched to production the whole system scaled at approximately 1 enterprise-grade server to 1 concurrent user.
While I applaud your efforts to teach people how to write code faster, the majority of JS devs I have found just want to hit feature-complete and go home.
And when nearly every Stack Overflow post asking about performance implications is answered with a litany of “YAGNI” and “premature optimization”, its not hard to see why.
The current comp sci culture seems to discourage this in all but the lower level languages like C.
You can make an Electron app not slow/a battery hog, but it requires working backwards from a state of "what on earth is going on under the hood here". This is the inverse of building a native application, where that complexity is often introduced by you. I personally find this significant - you can teach someone how to avoid the latter, but it can require crazy amounts of insight to learn to debug the former.
It is 100% possible to build better Electron apps. I trust almost nobody to actually commit to doing it.
Also, particular yet tangential to this thread: apps like Signal or Discord which still make their Electron apps run under Rosetta 2 on an M1 are a nuisance. I just run Discord in a browser tab at this point.
Yeah many people do not just have the time to look into this, and most companies don't really care, as we have so much RAM and CPU power these days, so I can see why for example many messaging apps just choose to write their app in Electron, instead of trying to figure out how to do it natively cross-platform, which can be a more PITA especially when you also have web as a platform.
This feels like... a trap.
I think there's a complexity in doing that which could be summarized as follows: it's possible, but the more complex your codebase, the more difficult it is to reason about how to get to or maintain that 80%.
Even if we assume that 80% is the maximum, if most products (or the notable ones that we all complain about, I suppose) are hitting 50%, then... well, that's a problem.
For full disclosure: I've defended Electron on the merits of "there is sadly no better way to ship cross-platform UI-based software". I criticize it with this in mind.
Finally sold the 2011 MacBook Air a few months ago for $300. I got so much value out of that laptop.
PS: apparently the quote is from 1997, go figure.
c - the speed of light, independent of the speed of the observer.
w - the speed of Windows, independent of the speed of the hardware.
1. Garbage collection and minimum runtime
Garbage collection decreases memory efficiency, browsers need a certain minimum amount of resources to render a page, no matter how complicated leading to 500MB RAM text editors like Atom and worse once you open files. Similar problems plague Eclipse and IntelliJ which both often consume 1GB of RAM. The JVM often needs 150MB for an empty or perfectly optimized app.
2. Everything is an object with pointers
This is especially bad in Javascript where every object is basically a hashmap. This causes performance slowdowns because even something as simple as looking up a field is a pointer chase through several layers now. Raw numerical data may consume a lot of memory if you are not using typed arrays. Especially bad with ArrayList<Integer> in Java.
3. JIT
JIT compilers can only spend a certain amount of time on optimizations, which means JIT languages tend to either suffer from slow start up times or faster start up but less optimizations.
4. GUI complexity
Things like having animations and constantly recomputing layouts.
If you designed your processors for these things and made them fast at this, the only further source of slowdown is a lack of caring because you have already exhausted the unavoidable technical reasons. E.g. your processor is so fast, you write a naive algorithm that takes 5 seconds on the fastest computer available but then takes 20 seconds on an old computer.
Look at multi-platform products like ms office and how slow they run on a Mac. I suspect because there is a translation layer to bridge win32 calls to Mac equivalents. And that seems like it would be point 5 on your list.
What does that look like?
There is a reason Apple doesn’t do #1 and #3 and is moving away from #2 in their own code.
They are just inefficient ways of doing things. Designing a processor to support an inefficient mechanism will still lose out to a processor which doesn’t have to.
Diffing a VDOM might fit here, but that's not really GUI-specific - just a combination of your earlier points.
The fact that Apple writes much of their software for iOS and MacOS at the same time means much of it is designed to run on fairly light hardware.
I know we’re all stuck with bloated stuff like Slack and some of us with MS Office, but just go native where you can it reap the benefits of the platform.
I don't know about that, LLVM is getting slower with each new release too [1]. Same for Xcode: startup is very noticeably slower than a few years ago on the same machine. Starting into a debugging session for the first time is much slower (it used to be instant, now it's multiple seconds). The new build system in Xcode is most definitely slower than the old one, at least for projects with more than a handful compile targets. Etc etc etc... new software is only optimized to a point where performance doesn't hurt too much on the developer's machine.
[1] https://www.npopov.com/2020/05/10/Make-LLVM-fast-again.html
One nice thing is Swift is slow to build but quick at runtime.
But, I do agree that there is a lot of jank in software that we write these days. Over the weekend I started writing an application that uses glfx + dear imgui, running on a 360Hz monitor. The lack of latency was incredible. I think it's something that people don't even know is possible, so they don't even attempt to get it. But once you know, everything else feels kind of shitty. This comment box in a web browser just feels ... off.
(Except for lsp/company which occasionally just messes up everything).
Agree, and I would add that devs on some platforms have not been as thorough about using (or abusing?) all that new power. I've been able to continue use a Sandy Bridge laptop from 2011 - it'll turn 10 years old in a couple of months - because most apps on Linux still have so little complexity that you could run them on a computer from 2006. Electron-based stuff is not the norm. Adding an SSD a few years ago has been my only significant hardware upgrade.
Meanwhile, Macbook Pros that came out two years after my laptop was built can't even be updated to the latest version of macOS, before you even get to the issue of software jank.
So I actually think that Mac software will hold the performance edge for a long, long time.
It's fine to have an ARM version of Windows but there's no equivalently high-performance ARM chip to run it on. Unless you're talking about Bootcamp on Macs.
I have a vague tickle in my memory that both Intel and AMD have dabbled in ARM before.
Even if they don't assign large production volumes to it now, they could still have a design team working on it for a contingency plan.
https://en.wikipedia.org/wiki/AMD_K12
https://en.wikipedia.org/wiki/Zen_(first_generation_microarc...
Once Apple released M1, rumors of AMD ARM development started again https://www.techspot.com/news/87851-amd-rumored-working-arm-...
Also, Apple’s buggy releases (which is tradition now[0]) didn’t really help.
0. Apple essentially stopped fixing bugs and they just release another buggy OS in a year or so with a new name and changes that break everything.
Azure and Google need to really step up and get some competitive (or even any) ARM instances--fast.
What's even more crazy is that you cannot buy them. Such price/perf is literally only available to rent from AWS.
I hope Oxide does something about this soon; it is a sort of dangerous situation to have such a competitive advantage only be available via rental from a single (large and anticompetitive) company.
There's also nothing wrong with the $999 M1 Air. Isn't the Surface also ARM? And most Chromebooks?
After you've already got the code, though, it's Amazon's way or the highway.
Apparently Oracle Cloud is going to offer both VMs and bare metal on Ampere Altras.
Intel? Well, maybe 12 months after they catch up to AMD in x86.
AMD's chips I've seen are decently competitive with M1. If ARM is a not-too-bad microcode adaptation and ditching a whole lot of x86 silicon real estate, AMD might be able to come to market quickly.
Intel isn't just getting pantsed in x86 by AMD and process by TSMC, they are seeing their entire ISA getting challenged. Wow are they in trouble.
For my Java services so far, no change. Aurora was a button click. ECS is the next thing to tackle.
For M1, it obviously outperformed competitive.
https://www.anandtech.com/show/15578/cloud-clash-amazon-grav...
Bonus! This makes performance review time super easy; you've saved the company thousands of dollars this year, how about a raise?!
In a server high load environment, is this still the case? Continuous load, consistent jobs -- is the M1 still that much better or does its advantages decrease?
The problem is that you effectively have to run macOS to take advantage of it and that's a no-go for a wide variety of people. I don't even care if I can ever run Windows or Linux on an M1 because Apple will make it a pain in the ass to do so. They don't even have a fucking bios UI... Imagine enterprise IT shops rolling Windows out on that garbage? It'll never happen.
And I don't want ARM at all unless the systems built around it are as modular and open as the x86 ecosystem.
As a developer, I'm sweating bullets. Imagine a future where Apple is the only serious platform left on mobile, server and desktop, then developers can forget about starting that next disruptive startup and instead they will become just second-class Apple "employees" with contracts similar to Uber drivers.
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 ARM space; in fact I think we will see the emergence of new players as ARM isn't as finicky about licensing their tech as Intel is. The only reason Intel even had a competitor (AMD) in the past 20 years is because they licensed x86 to them before it started becoming the de facto choice.
To their credit, Apple takes the time and invests the money to build the stronger house.
In 1968 there weren't that many computing devices around. When video games and home computing came around, there were massive opportunities and the entrenched players missed some of them.
Same with mobile phones, same with the internet, same with smartphones.
Still ibm then have no consumer whilst apple now has no business and cloud business domination ... still lots of space to play.
But if windows moved to arm with the same result I cannot say intel can survive though.
The dual wake-up call from AMD and from Apple (ARM), combined with the money Intel has in its pocket will have a serious influence on the cpu market. Unsure if they'll come out ahead, but it will get interesting and probably cheaper, not only for consumers.
To my understanding, this is quite unlike AMD64 (ie, Intel's current 64 bit ISA licensed from AMD) - which extended x86.
Does anyone know who(mses) designed the instruction set?
RISC-V likely will dominate in microcontrollers and specialized/embedded devices. But in general purpose computers? Highly doubt it. Arm has a giant first mover advantage there now. Great momentum, excellent ecosystem. AWS Graviton2 basically sealed the deal.
Example:
Perhaps, but I guess it would take on the order of a decade for a competitor to build their technology including an ecosystem with customers.
And who says they won't copy Apple's business practices?
> but everyone else is just sitting on their asses while Apple's chip team puts in the work.
That's because the complexity of the semiconductor industry is daunting. You need far more than a team of IC designers, especially if you want your product to become successful in a world where there is one main player.
Sign me up for more of those kinds of "business practices", thank you very much.
Ad targeting can be based entirely on the content of the page where the ad appears.
Competition has brought us such great processors. We should be thankful for it.
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. Also Google and Microsoft have complementary businesses but they barely coordinate on this. The closest thing I have seen is Android helping Microsoft with Dual screen Android devices. In most other areas they are setup to compete (chromeos vs. windows, azure vs. gcp, stadia vs. xbox cloud, gsuite vs office, bing vs google search etc).
2) Samsung is the most equipped to lead this.
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)
Mac mini uses like 25-27w at the wall. 15-25w TDP for the 5800u doesn't include RAM (it's LPDDR4 at 1.1v will use a lot more power than M1 LPDDR4x at 0.6v). 5800u also has a chipset that consumes a few watts.
5800u has some features that the M1 does not (these would likely increase M1 power usage), but still doesn't seem close in power consumption.
4800u isn’t that different as far as power goes (mostly different in performance). A mini-PC with it uses almost 70 watts under load.
https://www.anandtech.com/show/16236/asrock-4x4-box4800u-ren...
I see glimpses of this with some of the cheap commodity hardware available, but transportation and design costs are still too high.
Ability to break in would, I suppose, depend on that cost of design and adoption. The ability of large companies to efficiently acquire, though—-not sure what would change that situation.
retina imacs and macbook pros beg to differ
8,9 years later still no good alternative
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...
(ryzen APUs are due for a facelift in the graphics department)
This is so deep and powerful if you really think about it. This is what we were promised. While not perfect, I applaud Apple for at least trying.
They didn't even do anything special. People have been telling multiple ARM IP holders to build a wide, powerful chip for ages. Especially for the server space.
Apple is just the first one to actually do it. AMD and Intel have been doing it all along, which is why it's so impressive that Apple's chip is finally in spitting distance of them despite being a fairly normal design with some specific optimizations for use cases.
That was 12 years ago. I guess software licenses go up and I am certain that current processes have chip prototypes to cost $500K-$1.5M per evaluation.
Yes, you can say everyone sit on their hands and/or asses.
I say that having money in bank for several dozens of thousands of chip prototypes certainly helps Apple in the design.
On what grounds do you suggest you know anything about how apple accomplished M1?
Or are you just intuiting that it was some sort of "outside the box" thinking?
Maybe it was raw engineering and management mastery combined with elite head hunting and millions of dollars poured into research and design over the course of a decade?
Nah. Apple's too cool for that. Probably some eureka moment during meditation class, right?
The other day there was an article here about Swift on the server and it was full of replies from people who would love to use it more but lack of investment from Apple makes it often a poor choice. It's doubtful even Apple is using it much server-side in house.
I'm happy to be wrong, but I think you're buying too much into the SV/media fearmongering on a handful of companies whose main business is actually just attention economics.
I was using admittedly vague language for rhetorical effect. I only technically meant the past 21 years.
https://www.statista.com/statistics/576473/united-states-qua...
Success is not a crime.
I'm pretty sure it's device count. Which means we are comparing $90 smart phones to $1200 iphones that will result in thousands of dollars spent through the apple ecosystem.
In 2019, Apple captured 32% of the global revenue for the mobile phone industry and 66% of all profit in Q3 of 2019 [1]. Asymco used to report on this frequently but I'm having trouble finding recent examples [2].
[1] https://www.forbes.com/sites/johnkoetsier/2019/12/22/global-... [2] http://www.asymco.com/2011/08/02/apple-share-of-phone-revenu...
Nothing gets me thinking about disruption more, than new, faster, cheaper platforms. We’ve had to accept 10% better per generation for way too long.
I’ll know that Apple is a serious contender for datacenters when PHP is updated to use native macOS APIs when running on macOS, because that’s a great litmus test for whether open source is ready to compromise on “generic computing is king” in favor of “high-performance OS APIs for specific hardware platforms”. I’m not holding my breath, and I don’t think most open source projects are willing or able to either decide to take these steps or actually deliver on any such intent. Apple seems content to offer rackmount macOS and that’s more than sufficient for serious businesses that need serious amounts of Apple-accelerated hardware. But I don’t consider most “Internet cloud server” businesses to need that, and I don’t see the developers of our clouds as being willing to give up their dylib freedoms in exchange for higher performance per watt.
EDIT: Think of M1 as a hardware platform dedicated to code that's been ported to Apple’s microkernel, if that helps any. Yeah, it’s not “micro”, in theory; except they run a stripped-down iOS on their HDMI dongle (yes, seriously), so clearly they’re quite capable of “micro” kernel. More on kernels today: https://news.ycombinator.com/item?id=26954547
I think that dual 4K is the old way, single 6K/8K ultra wides are the future.
That said I am rocking the dual 4Ks on the Mac Mini M1 and it works great: https://twitter.com/BenHouston3D/status/1384693982249340935
No idea if it's the monitor of the laptop, but it's the exact same problem and hopefully hardware doesn't glitch in such a consistent way and it can be fixed in software.
Neither monitor supports Thunderbolt. But I run off of a USB-C to Display Port cable and one is using the HDMI port. Both run at 60Hz as it is HDMI 2.
The amount of available screen space at 1:1 is refreshing (not as crazy as imagined though), but it makes me realize how grainy "normal" monitors are. In some years 8K TVs might be cheap enough for an upgrade, but on that time scale I can see VR being a serious contender as well (it's really good already).
The last time I tried a large format display, I couldn't move back far enough to make it work but I really loved working in 4k without scaling.
It's so big that I actually don't sit in the middle but to the left of the display. I use the right of the display as a second monitor.
As an aside, I have the Philips Momentum (model 436M6VBPAB) which is sold as a monitor but can double as a TV. It has HDR, which is very nice in a light office. I suspect it is no longer produced in any significant quantities. The price has been going up.
> they are grainy up close
Hmm, maybe it's its "TV-ness", but the individual pixels look pretty sharp to me at a close look (no quantum dots or weird pentile arrangement, that should definitely be looked out for), but I have no other low DPI display to compare it to. My reasoning was always "it's just an LCD panel, why would it be different" and so far I feel proven right.
> I also wonder if it's good for your eyes to focus on something that's constantly far away
My appeal-to-nature-argument would be that the natural human mix would probably have been much heavier towards looking into the distance than it is now (see the massive rise of near-sightedness in cities), so it can only be better than what we're currently doing.
There are few differences between a 4K 42" screen and 4 1080P 21" screens: the smaller screens can be angled, they have bezels, and are probably more expensive in aggregate.
I run it at 3840 x 2160 and use a trackball to get around. I like it a lot. LCD TV screens do not work well extremely close up but an OLED is fine even at 12" away. Text in the corner of an LCD TV is dim and poorly defined but on an OLED it is just like it would be on a 55" sheet of paper.
Of course I was concerned that the OLED screen would burn in, but it was an older TV so I took a chance.
Most days the computer and TV are on for about 12 hours. I have run them like this for 14 months now. There is no burn in at all. The overall brightness is slowly declining, but it has been doing that since I got it in 2016. (LG OLED55B6P)
Most TV's have a "gaming" setting that doesn't try to "improve" the video feed and just passes it through.
I can slide the desk out of the way when I want things back to normal in my living room.
Sure but the Air and MBP already support 6k external displays.
They don’t support 2x4K.
I’ve been using an ultrawide since 2015 but have almost always had a second side monitor with it. The period where I didn’t have a second monitor was short as having a separate monitor has always come in handy. All of the extra space on an ultrawide is great but when you’re doing something that takes the whole screen, it still leaves you in the same position you would be with a single 16:9 display.
TBF my old setup was a decade old so it was merely a 24" (1920x1200) and a 19" (1280x1024 in portrait).
Though my new display has a much larger logical size, the inability to put something fullscreen on one of the monitors (whether the small one or the large one, depending on the content) is definitely one of the drawbacks I had a hard time getting used to, not only can it be inconvenient, it actually wastes space as I can't get rid of window decorations, which putting e.g. a video player or browser window in fullscreen allowed.
that said, i have no issue with your choice. for some applications (eg video editing is the default choice for ads of UW monitors) it is really better.
i do have issue with you denigrating 2 monitors. it is not at all "an old hack".
I personally prefer the visual break as I find it useful for creating fixed 'work areas': terminals/xterms, browser, mail, etc.
I'd strongly consider getting a 6k monitor, or maybe a 5k2k ultrawide if I didn't already have two my two Dell 27inch 4ks.
In the near future 16:9 8ks are going to be standard (too bad it won't be 16:10) but right now you can get a decent triple 4k monitor setup for half of what an 8k monitor costs.
IMO ultra-wide is great for productivity in a situational sense but it sucks for gaming and media consumption. Also getting proper support for ultra-wide to be standard is probably going to take another decade.
I think there's something to be said for each monitor being a separate entity. I think some people will still prefer multiple monitors over ultra-wide even when support for ultra-wide is a solved problem.
My personal setup now is a 28" 4k@60hz, a 32" 1440P@240hz as a primary and a 24" 1440p@144hz. I have my work computer, my desktop, a mac-mini, a switch, and a PS4 running off of them so having separate monitors is ideal for me.
Ultra-wides and super-ultrawides are cool, but IMO they aren't as practical yet.
There's support, and then there's that support being so flawless that its the same quality of experience you would get with a standard aspect ratio. Each of us has our own standard of what's good enough. Its working out for you but the people I game with that were using ultra-wide monitors switched back due to the number of issues they were having, as recently as last year. I did some research myself when I was upgrading my monitor and some of the games I played would have had issues so or me personally it wasn't good enough.
Another thing to consider is a lot of game content is designed with standard aspect ratios in mind, so whether expanding the viewpoint makes it better is going to be a personal standard. It will be interesting to see if UW monitors do become a standard in a couple of decades whether game developers start specifically making content that utilizes the extra space to offer something that isn't possible with existing games.
Example: https://i.pcmag.com/imagery/articles/07toBDd6lpyucCyM0xWrcQv...
Also, with two screens, I can adjust the angle between them to be more or less severe.
Also, 2 monitors means more choice in layout; for example, my 2nd monitor is in portrait orientation.
I work on mac all day and on my 1440p widescreen I am constantly trying to put stuff so I can use everything I want to at the same time. `[Code|App|Docs]` is so common we ought to be able to split a wide screen 3 ways the way we do a normal 2.
I was using Spectacle before, but I think the project died, and Rectangle is what I use now.
[1] https://www.apple.com/macbook-pro-16/specs/
edit: oh right, that's not M1 brainfart
I have an old Macbook from 2013 that runs Linux now with an Intel integrated GPU and it supports dual 4K monitors at 30Hz. I use the Thunderbolt ports as Mini DisplayPort outputs and connect to two 4K monitors with DisplayPort and it works fine.
The mini can support two external displays. One via thunderbolt at that same 6k and 60Hz and the other a 4k at 60Hz through the HDMI port.
I use the Dell D6000 and I run three (1080p) external monitors in addition to the built in monitor.
I have two Thunderbolt/USB-C monitors that I was hoping to daisy chain with one wire from my Mac. Alas it's not possible.
My hope is power into a dock. Thunderbolt from dock to laptop to power laptop. Thunderbolt/USB-C from dock into first monitor. Second Thunderbolt/USB-C from dock using the DisplayLink tech to second monitor.
(But seriously though... what is 16gb these days? Looking at process monitor, I think my firefox takes over 5gb now since I have over a thousand tabs in treestyletab, clion + pycharm take another 5gb, parallels vm for some work stuff is another 10gb; if doing any local data science work that's usually at least a dozen gb or a few dozen or more)
A 128GB of a ram in a consumer PC is about $750 dollars (desktop anyway, laptop may be more?). That's less than a single high end consumer gaming GPU. Or a fraction of a Quadro GPU.
So to the extent that developers ever run things on their local hardware (CPU, GPU, whatever) 128GB of RAM is not much of a leap. Or 256GB for Threadripper. It's in the ballpark of having a high-end consumer GPU.
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.
I now use an M1 mac mini with two 4k monitors (not from one cable though)
1 monitor (Dell P2715Q) is hooked up using DisplayPort. The other (LG 27” IPS 4K) is using HDMI 2.0. Both are running 4K retina at 60Hz without any issues.
I had so much trouble finding this out before I bought it which is why I am vocal about the fact that it works.
This year we should see a lot of CPU manufacturers change to (LP)DDR5 for memory - high end Samsung and Qualcomm SoCs are already using LPDDR5. It's a safe bet Apple is also switching to LPDDR5 which would put the upper limit on 64GB for the M2. This is notably still lower than the 128GB you can put in the current 27" iMac (for an eye watering $2600 extra), but is an amount far more pro users can live with.
The real issue is not shipping the M1 with LPDDR5 support. We already have tons of Qualcomm phones and even Intel processors shipping with since last year. If Apple had done the same, we wouldn’t be talking about this today.
I brought this up before but it was pointed out to me that the only manufacturer of LPDDR5 last year was Samsung, which was producing not-completely-standardized LPDDR5 at the time and probably didn't have enough spare volume for Apple anyway. Having 12GB LPDDR4X modules from one vendor (24GB total) probably is not enough reason for Apple to switch up their supply chain either, not for the M1 at least.
And, to be fair, I think Apple did get away with shipping low memory computers by targeting only their cheapest computers.
I think we as programmers have lost sight of how a single gif in memory is about the same size as a whole docker container.
Go check out the resource usage. Yes, it involves virtualization, but the stripped down virtualized kernel uses very little resources. Getting that information is a bit complicated by the fact that it's difficult to see the difference between the virtualized kernel and it's applications in a way that's consistent with the host OS, but it's really in the dozens of megabytes range of overhead for something like docker on mac.
Just to check, I opened simultaneous projects in Logic Pro and Ableton Live, I have a Rust project open in VSCode (Electron), compiling Erlang again, I have Slack, Discord, Signal, and Hey running (all Electron apps), and I have Firefox open with Twitter, Hacker News, Facebook (a dozen tabs, just for good measure), Instagram, FBM, Techmeme, NYTimes, CNN, and MSNBC. This is a 15" 2018 MBP with a 6-core i7 and 16GB of RAM, and the CPU is working pretty hard but RAM usage is currently at just over 12GB.
Safari shows a memory warning when a tab uses ~2GB memory, and can be too aggressive about it.
Yes, she was getting both. To the point that I asked her to keep activity monitor running in the background so we could see who the problem applications were.
I'm not sure why this is in question.
If "compressed" or "swap space" approach 32GB+/64GB+ (depends on amount of physical memory installed) that can do it, but running out of disk space is a more common problem.
Of course, quitting the top app in the memory list probably is going to fix the issue.
I had to have her leave activity monitor open in the corner because it was running so slow it wouldn't open.
Everyone raves about how 8GB machines feel, but the massive SSD usage and pitiful lifespan shows that there’s not a free lunch here. 16GB (or a 24GB option) goes a long way toward preventing early death. I actually suspect they’ll be launching a free SSD replacement program in the next couple years.
I’ve also heard claims from a couple devs that side-by-side comparisons with their x86 macs show a massive increase in RAM usage for the same app for whatever reason. I’d guess that’s solvable, but could contribute even more to the SSD problem. On the bright side, all this seems to indicate a great pre-fetch and caching algorithm.
Not even the most alarmist estimates suggest that the SSDs on the 8GB machines will start dying after two years!
At the moment the SSD lifetime concerns are little more than FUD. Everything we know is perfectly consistent with these machines having SSD lifetimes of a decade, even with heavy usage.
https://techreport.com/review/27909/the-ssd-endurance-experi...
700TB was the minimum that any drive in the above link managed. If you used 100gigs of swap per day it would take you two decades to hit that level.
Partly because the GPU works differently, partly because some inaccurate methods of counting memory (Activity Monitor) have cosmetic issues displaying Rosetta processes.
If the nvme drive is fast enough, the filesystem cache is not as noticeably effective, and if you have aggressive “suspend” of processes (like iOS) then lack of RAM capacity does not really impact performance at all. But memory latency, speed and the CPU do.
for me video cache would be really helpful, but after effects and especially premiere right now seem absurdly slow.
premiere especially i can hit play on a non-edited clip in preview and it takes multiple seconds to start... maybe i have a wrong setting somewhere
Here's an incomplete and continuously updated list of monitors (not) working with Apple hardware: https://tonsky.me/blog/monitors-mac/
I think it's quite likely that M2 will have a 128-bit lpddr5 interface. Using widely available modules, this would allow them up to 64GB ram.
Returning it still hurts. I was really hoping for something that would take more than 32GB of RAM but am not surprised that it's still going to be later this year. The guts in the new iMac would be perfect for my Mom if they just offered a 27" replacement too.
Oh well. A few more months won't kill me.
Could you explain what you mean by this? I don't think I understand.
A far cry from the kind of tolerances and sustained high throughput of PCIe.
A 1 Tb Mini with all the SSD write lines benchmarked at
Write: 2,983 MB/Sec Read: 2,710 MB/Sec
That's as fast or faster than Intel Macs, and way higher than most PC SSDs.
That said, Intel mobile CPUs don't have fantastic I/O either, this will matter more when Apple has to scale up to iMac Pro/Mac Pro territory.
"SSD performance" is a very high-level view of I/O, and depends more on the size and configuration of flash chips used than the CPU design.
Even their Mac mini doesn't offer a 32 gigs of ram. One of the reasons I had to go with Intel version instead of M1 because at least I could upgrade to 32gb ram myself (plus was cheaper). There is no reason why they can't offer 32 gigs ram on non-laptop lines.
I think most observers have been expecting an “M1X” for Apple's first pro-oriented ARM Macs, so an M2 already would be a surprise.
I'm pretty sure that's what this is, rather than a next-generation ("M2") chip. It will likely have the same cpu core designs as the M1, just more of them. And possibly paired with the new, long rumored "desktop class" Apple GPU.
As to what Apple's choice of marketing name, that's entirely arbitrary. (For what it's worth, my guess is they're ditching "X" suffixes and will designate higher spec variants with other prefix letters e.g. the "P1" chip.)
The M1 was available on Nov. 17th, 2020. The article states that the chip is entering mass production, and due for release sometime in 2H. This could easily be released a year after the M1, if not 13 months later.
But now we know that it was always intended to be a real player (it's in the iMac and iPad Pro).
So this news is interesting to me because now it seems to cut back the other way that maybe the M1 was designed for a short shelf life.
In a world where Apple is controlling so much of its silo, "normal" design and building deadlines will be upended.
Which is a weird take when you consider the thermal issues that Intel macs were plagued with. It's almost like the chassis was designed with 10w of dissipation in mind which Intel couldn't operate within, but the M1 could easily.
I had assumed that Apple designed for the M1 and then fit Intel chips into those designs.
At the time of the Intel transition, Apple had already gone through the process once before with 68k to PPC. It had to be clear to the long-game thinkers at Apple that this cycle would keep repeating itself until Apple found a way to bring that critical part of its platform under its own control. Intel was riding high in 2006, but so had IBM in 1994.
Within two years of the Intel transition, Apple acquired P.A. Semi. The iPhone had barely been out for a year at that point, and still represented a fraction of the company's Mac revenue – and while it looked to us outsiders like the acquisition was all about the iPhone and iPad, in retrospect, a long-term replacement for Intel was almost certainly the endgame all along.
but with the 2016 line of macs, it was obvious that apple was expecting faster, smaller, cooler, more power efficient 10nm chips from intel, and intel fell flat on their face delivering. it's not clear how far before that that apple knew intel was flubbing, but 2014 seems a reasonable assumption given product development timelines. as intel's downward trajectory became clearer over the following months, along with the robustly upward trajectory of apple silicon, the transition became realizable, and eventually inevitable.
as an aside, i'm using a beat up 2015 macbook pro and eagerly awaiting the m2 version as its replacement, seeking to skip this whole intel misstep entirely.
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 bed of empty promises.
Apple has so much experience with processor transitions they don't have to stay on ARM either. And they have the capital to move somewhere else if it makes enough sense to them. I find it highly unlikely - but if it made sense it would be highly probable :)
I don't mind upgrading every other year. I just want the upgrades to be meaningful.
M2 is probably based on the Arm v9 ISA and has been in design for years.
Well, if that is indeed a stop gap then I can't wait to see their first "real" chip :)
Having said that, Apple does have something of a history of pushing out v1 of a product that sets a high bar for everyone else to try and catch up with, then immediately pushing out a v2 that raises the bar well above where everyone else was aiming.
Overall though, it's awesome that the Macs now get to benefit from the vast investment going each year into making faster/better CPUs every year, for hundreds of millions of new iPhones.
An M2 name implies some architectural differences like extra cores or more external bandwidth. I'd be totally happy with an M1x with some tweaks like more external connectivity and more memory.
Which, for me, would be quite perfect. The only reason I'm holding back this purchase is the 16 GB memory limit.
So Apple will just have to wait to get my money until fall (or whenever they announce the successor to the 27in iMac).
Unless they are going to use this same CPU for the Mac Pro, this is right on schedule.
If they would have released the new iMac with a bigger panel so there were options for the 27" as well as the former 21" then my mother would be rocking a new iMac next month.
I know they said up to two years for the transition but I want to transition now :)
Main reason being the M1 was a more proof of concept and rushed out (despite being as good as it turned out to be). The M2 will be a more refined M1 but with notable improvements such as LPDDR5 support - akin to AMD's Zen1 and Zen+ releases.
On the other hand, there could be a M1X being readied for release in the upcoming June WWDC. It may be architecturally older than the M2 but still superior performance on a big cores differential, e.g. M1 only has 4 big cores and 4 small cores, the M1X just needs more big cores to be notably more performant.
All highly speculative of course, will have to find out in about a month.
Support for one of the recent Radeons was recently added to macOS, so it's a possibility. No reason the M1 can't do PCIe, as far as I know the only thing keeping eGPUs from working on the M1 right now is software support. It could also be that the driver was added because of the extensibility of the Pro, though.
My expectation is that they'll keep the GPU on the same level, which is "good enough for most users", and focus on hardware acceleration for tasks like video and audio encoding and decoding instead. With an ML chip and fast audiovisual processing, most consumers don't need a beefy GPU at all, as long as you stick to Apple's proprietary standards. Seems like a win-win for Apple if they don't add in an external GPU.
Still I can’t see Apple only developing one integrated GPU per year unless they somehow figure out how to magically make them somewhat approach Nvidia and AMDs modern chips. What would the ARM Mac Pro use?
It seems that Apple has put in a lot of development resources into getting Octane (and maybe Redshift and other GPU accelerated 3D renderers) to support Metal (to the point where it sounds like there may have been Apple Metal engineers basically working at Otoy to help develop Octane for Metal) and I can’t just imagine that happening just to support the the Apple Silicon GPUs. I wouldn’t be surprised if we see eGPU support announced for ARM Macs at WWDC (and maybe even the iPad Pros that support Thunderbolt. Yeah the idea of plugging your iPad into an eGPU enclosure is funny, but if it’s not to hard to implement, why not?)
What do mac users need a beefy gpu for?
AFAICT apple just need a GPU that's good enough for most users not to complain, integrated Intel-GPU style.
Most people I know doing 3D rendering nowadays just connect their UI to a remote rendering farm. So a Macbook Air would be more than enough for that.
For video processing you don't need 3D rendering capabilities, just hardware acceleration for the video formats you are using.
At the start there will still be a lot more Mac Pros running AMD hardware that must be supported.
It may not be obvious, but Apple has repair work to do in the pro community. Four years ago this month, Apple unusually disclosed that it was "completely rethinking the Mac Pro." [1]
This new Mac Pro design wasn't announced until June of 2019 and didn't hit the market until December 10th of 2019. That's just _six months_ prior to the Apple Silicon announcement.
So, unless Apple simultaneously was trying to honor pro users while also laying plans to abandon them, it is hard to imagine that Apple spent 2017-2019 designing a Mac Pro that they would not carry forward with Apple Silicon hardware. Keep in mind, the company had just gotten through a major failure with the Gen 2 cylindrical Mac Pro design.
The current, Gen 3 2019 Mac Pro design has the Mac Pro Expansion Module (MPX). This is intended to be a plug-and-play system for graphics and storage upgrades. [2]
While the Apple Silicon SoC can run with some GPU tasks, it does seem it does not make sense for the type of work that big discrete cards have generally been deployed for.
There is already a living example of a custom Apple-designed external graphics card. Apple designed and released Afterburner, a custom "accelerator" card targeted at video editing with the gen 3 Mac Pro in 2019.
Afterburner has attributes of the new Apple Silicon design in that it is proprietary to Apple and fanless. [3]
It seems implausible Apple created the Afterburner product for a single release without plans to continue to upgrade and extend the product concept using Apple Silicon.
So, I think the question isn't if discrete Apple Silicon GPUs will be supported but how many types and in and what configurations.
I think the Mac Mini will remain its shape and size, and that alongside internal discrete GPUs for the Pro, Apple may release something akin to the Blackmagic eGPU products they collaborated on for the RX580 and Vega 56.
While possibly not big sellers, Apple Silicon eGPUs would serve generations of new AS notebooks and minis. This creates a whole additional use case. The biggest problem I see with this being a cohesive ecosystem is the lack of a mid-market Apple display. [4]
[1] https://daringfireball.net/2017/04/the_mac_pro_lives
[2] https://www.apple.com/newsroom/2019/06/apple-unveils-powerfu...
[3] https://www.youtube.com/watch?v=33ywFqY5o1E
[4] https://forums.macrumors.com/threads/wishful-thinking-wwdc-d...
Apple designed and released custom hardware that used a new slot to accelerate compute. My point is that this illustrates Afterburner as a product shows clear direction for Apple to put Apple Silicon into discrete graphics or other acceleration compute in the Mac Pro.
You say that like Apple doesn’t do stuff like that all the time.
eGPUs can be a nice addition, but I doubt Apple will release an official eGPU system. You're already limited to AMD GPUs after the clusterfuck of a fight Apple and Nvidia had, and I doubt Intel's Xe lineup will receive much love for Apple right after the Intel CPUs have been cut from Apple's products.
Honestly, for the kind of work that does need an arbitrary amount of GPU horsepower, you're barking at the wrong tree if you buy Apple. Get yourself a Macbook and a console or game streaming service if you want to play video games, and get yourself a workstation if you want to do CAD work.
I don't think the work Apple would need to put into a GPU solution would be worth it, financially speaking.
They already have one[1], and you can even buy eGPUs from the Apple Store[2].
1. https://support.apple.com/en-us/HT208544
2. https://www.apple.com/sg/shop/product/HM8Y2B/A/blackmagic-eg...
Option 2 - no, but does it matter? It's not like the previous gen Macs had great GPUs and no one is gaming on a Mac anyway.
Option 2.5 - bring back eGPU
True, but previous macs were never really competitive with PC alternatives on the hardware side, since they all used the same chips just with a higher price tag. With M1, that's starting to change, and Apple has the opportunity to attract a much large customer base for Mac than it ever has.
And of course, they're much more interested in gaming nowadays thanks to iOS. Maybe not interested enough to suck up their pride and apologize to Nvidia for giving them the finger, but probably enough to at least stick a beefier GPU into macs.
Apple's modus operandi of quickly and frequently deprecating old architectures and requiring app developers to constantly keep up goes down very badly with the traditional video game development model - of spending time and budget finishing up one game, releasing a few patches, then moving on to the next game with little further upkeep or maintenance of the now-done game. (Yes, games as a service is more common nowadays, but a huge number of games still go by this old model.) This model relies on long-term compatibility of old binaries on the platform being pretty stable, which is fairly true for consoles and Windows, but Apple platforms are anything but.
There are massive piles upon piles of only slightly old games that are not just unsupported but simply refuse to run on both the iOS App Store and Steam for Mac (including Valve's own back catalog!), due to the abandonment of 32-bit binary support a few versions back. And even if the developer is willing to do bare minimum upkeep work to recompile an old game and make it run on current hardware, chances are that between the time of release and now, lots of new mandatory hoops (eg. natively support a certain screen size) have been added to the app store checklist so that passing store certification requires tons more work than a simple recompile, further deterring the dev.
Perhaps you could chalk it up to the dev being lazy for not doing regular maintenance of the game, but the rest of the game industry doesn't force you to do that, while only Apple does.
A Surface Book 3 with an intel processor and an outdated Nvidia 1650 TI laps around M1 in games. Almost 2x performance. I'm not even going to compare it to laptops with modern GPUs.
https://www.anandtech.com/show/16252/mac-mini-apple-m1-teste...
The M1 chip is amazing but I'm a tiling window manager man.
Doesn't matter if Apple does mobile and low power stuff exclusively, but if they can scale this design into a higher TDP/core count it's going to get interesting.
Agreed. I mostly just needed the keyboard driven window snapping I was used to on Gnome and rectangle has filled than need 100%.
- https://www.reddit.com/r/unixporn/comments/jupmda/aquayabai_...
- https://www.reddit.com/r/unixporn/comments/mvuplf/yabaimacos...
Its called Yabai (+ skhd): https://github.com/koekeishiya/yabai
That is, you can have a tilin WM today with all the advantages of running MacOS.
System Integrity Protection needs to be (partially) disabled
for yabai to inject a scripting addition into Dock.app for
controlling windows with functions that require elevated
privileges. This enables control of the window server, which
is the sole owner of all window connections, and enables
additional features of yabai.
This holds for anything in the Apple ecosystem, up to Fortnite.
Yabai has been going for long, and every issue could be worked around relatively painlessly.
Holds for anything in almost any ecosystem. With that said, Apple has stated over and over the Mac will remain developer friendly. Being able to disable SIP, IMO is part of that.
I have some predefined layouts[1] for my most common usage. So, one keyboard shortcut arranges the screen how I want, and I have other keyboard shortcuts[2] (along with using Karabiner Elements for a 'hyper' key) to open or switch to common apps.
[1] https://github.com/kbd/setup/blob/1a05e5df545db0133cf7b6f1bc...
[2] https://github.com/kbd/setup/blob/1a05e5df545db0133cf7b6f1bc...
This is of course not so on Apple mobile devices: it's dox yourself to the App Store or GTFO over there.
Irrelevant from a mass market perspective. That chip is still out of stock at Best Buy and the Amazon third party version is marked up 54% versus MSRP.
Incidentally the 11900k is also out of stock at many retailers, but it's so much cheaper. You can still buy a pre-binned version that clocks 5.1GHz; even with markup that costs 30% less than the aforementioned third party 5950x.
Availability and price matter. My take on AMD's heavy focus on the enterprise segment right now is that they have no choice. If your enterprise partners get faced with scarcity issues, they will lose faith in your supply chain. You can tell a retail customer to wait patiently for 6 months, and even an OEM that makes retail devices (eg Lenovo) may forgive some shortfalls as long as there's a substitute available, but Microsoft and Google aren't going to wait around in line like that.
Honestly the price/performance ratio there is pretty nice in the eyes of a Mac user like me, but I don't know what office is buying Alienware, and a bulk order would no doubt take longer to deliver. Those are the only machines popping up when you filter for AMD 5000 series on dell.com.
Considering that 5950x is AM4 compatible, folks who had bought a pre-built machine and want to upgrade are also part of the mass market. And I think you can't discredit the homebuilt PC crowd for a high-end desktop chip. The people who care enough to want this chip can probably figure out how to clip it into a motherboard and tighten a few screws and connectors here and there.
- iPhone 2 had 3G and was so much better than 1 - iPad 2 was about 3 X slimmer/lighter than 1
Lots of other examples if I though about it.
I haven’t read many negative reviews.
It’s coming:
https://arstechnica.com/gadgets/2021/04/apple-m1-hardware-su...
(Virtualized Linux is already well-supported on M1 Macs)
I disagree with OP though, not second, but third generations have been the one to get for me: iPod 3rd gen, iPhone 4, Apple Watch Series 3. OK, iPad 3 was a bit of a failure, but still, first retina.
A lot of users would like to run multiple external monitors and have > 16GB of RAM. I know I'm in that group.
The main things I would hope for though are more RAM, and probably a much beefier GPU
I'm just hoping there are some more surprises in store with that slightly bigger battery.
* Only about 50 a 55% of all the software is M1 ready ( looking at some tracking sites ). Technically this is not a M1 flaw but you need something to push the market/iron out the bugs. While when the M2 gets introduced, you may be at 60 or 70%. As in less likely to run into issues as the M1 users are the real beta testers. Even Parallels only recently got good M1 support ( with massive speed increases ).
As of this moment, buying a M1 laptop is a less beta tester feature then it was several months ago. If you ended up buying in Nov/Dec, you spend a lot of time under Rosetta 2 or dealing with issues.
> I haven’t read many negative reviews.
The above was kind of always overglanced by a lot of the reviews, as the youtubers mostly looked at their own us for video editing etc and there most software was on point very early in the release cycle.
You see a lot of reviews in Jan/Feb from people going back to Windows laptops after a month, not because of the CPU being bad but because they ran into software issues.
In the mean time the software situation has evolved a lot more but the progresses with software being made M1 ready has also slowed down a lot.
As a PC user i want a M1 like laptop, that has long battery life, is silent and still powerful ( unlike a lot of Windows laptops where its always the old saying: Pick 2, you can never have all 3 ).
But i prefer one with 8 performance cores, double the iGPU cores ( with preferably DDR5 ) for light gaming and standard 16GB. So some macBook 16 Pro or whatever, if the price is not insane. We shall see what Apple introduces...
So far the new offerings from AMD and Intel are fast but still power hungry and heat generating ( aka fan noise! ). AMD is only going little.big in 3nm.
Intel's alder lake may be a M1 competitor ( for battery life under light loads ) but its again first generation product so expect to be a beta tester until Windows gets fine tuned for a long time to properly use the little cores! For heavy loads, ... well, 10nm is 10nm, no matter how many +++ you add.
You can also look at it this way: The M1 is not the first version. It is (IIRC) an enhanced version of the A14 SoC family, and a clear evolution of the work Apple has been doing on its silicon for the past decade.
- released with the (32-bit) Core Duo processor in May 2006 - replaced with (64-bit) Core 2 Duo model in November 2006
It only was supported through to 10.6 Snow Leopard as 10.7 Lion went 64-bit only
AirPods VS AirPods Pro is another I just remembered
I think watch v2 was a big improvement too.
I was so burned by that experience that I only bought an iPad again a month ago.
Imagine what would happen if the compute/$ M1 laptops would perform in some respect better with Linux than macOS. Things may get out of hand, when huge chunks of the Linux community gets involved.
This is just pedantry and needless nitpicking. Replace "work flawlessly" with "work well" in my previous comment.
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 problem. A14 and M1 are now selling in the millions, so they get a huge benefit. More importantly, they are in devices that get replaced a bit more frequently.
M1x (8-16 cores) will be in bigger iMac and laptops. They don't sell nearly as many of these. In addition, yields on the larger chip will be lower. Finally, people with these devices tend to keep their machines longer than phones, tablets, or cheaper laptops.
The 16-64-core chips are a major potential headache. If they go chiplet, then no big problem (actually, the M1x as two M1 chiplets seems like a desirable direction to head). If it is monolithic, the very limited sales and production will drive prices much higher. A logical way to offset this would be bigger orders with the excess being sold to Amazon (or others) to replace their mini cloud, but that's not been mentioned publicly.
Though the fact that they've just labeled the chip in the latest iPad Pro as such does add a bit of confusion to that.
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 machines just because of arcane numbering. And there’s precedent, Intel Core Duo Macs were released before Intel Core “Solo” Macs.
But if they’re actually ready for WWDC, then no, it’ll just be called M1x.
As for the Mac Pro… we’ll it’s definitely going to be weird. I think the existence of the Afterburner card proves that Apple sees a future in highly specialized modular add-ons providing most of the differentiation, but buyers would still want a bigger chip than in the 16” laptops, so who knows… of course nobody even knows how an M1 would perform with proper cooling and a higher clock!
[edit] also making M2x ahead of M2 will give them the benefits of “binning” some chips
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.
Hoping my next laptop will be a M2-powered MBP, assuming they can increase the maximum RAM to at least 32GB.
Of course, the Apple rumour mill; grain of salt, etc - but I wouldn’t be surprised if we saw an M1X that could, for instance - support 32GB of RAM by the end of the year - (which is the only blocker from me buying an M1) - and pop out the M2 next year maybe with an Apple-CPU-powered Mac Pro?
Food for thought. :)
Since they are non-upgradeable, I will certainly be waiting until a unit with at least 32GB RAM (ideally 64GB) before I'd upgrade at all and consider it future-proofed, but this is great to know!
So 6 months from now you'll probably have the choice between an M2-based MacBook Air with better single core performance or an M1X-based MacBook Pro with more cores and more max. ram - unless they decide to do the reasonable thing and shift the regular and X versions to the same release cycle.
I keep my personal laptops for a long time, and I don't want to be stuck with 16GB when it's getting close to not cutting it now based on what I do. I can't imagine being stuck with 16GB in 8-10 years.
My current personal laptop, a 2015 13" MBP, has 16GB of RAM. I can't imagine NOT upgrading.
I keep my personal laptops for a long time, and I don't want to be stuck with 16GB when it's getting close to not cutting it now based on what I do. I can't imagine being stuck with 16GB in 8-10 years.
I know as good as nothing of this process, but I can’t imagine the latency from silicon wafer to finished product is 3 months. I also can’t imagine some inherent start-up delay for producing the first chip (but as I said: I know as good as nothing of this process), so where do those 3 months go? Is it a matter of “you have to be extremely lucky to hit so few minor issues that it only takes 3 months to have a first working product”?
My mind was blown when I first found that out
More likely I would see Apple making an exclusive server chip licensing arrangement with ARM or something similar.
I don't know how feasible it is to scale that up to a datacenter though, and I expect MacOS licensing costs would wipe away any power efficiency gains. But I do wonder if they could hypothetically scale up and beat the best Intel/AMD have to offer just using Rosetta 2 to translate x86 code.
What's in it for Apple is money and better economies of scale for chips. But I don't really think it fits Apple's MO so I doubt they'll do it.
More specifically, people are running Linux on the CPu cores.
The M1 is a system-on-chip, and according to the floorplan [1], the CPUs are maybe 1/5th of the chip. There are many other features that aren't unlocked, such as GPU (which is a brand new architecture) or power management. The latter is key to exploiting the chip to its full performance envelope.
I don't expect Asahi to get anywhere further than proof-of-concept before it becomes obsolete by the march of the silicon industry.
The GPU stuff is new, but it seems like good progress is being made: https://rosenzweig.io/blog/asahi-gpu-part-3.html
For data centers, it helps that GPU and so on is just less important. It's wasted Silicon, but the CPU is competitive even before considering the added features so that's not the end of the world. There's a good chance that Apple can use that to their advantage too, by using chips with broken GPUs/NNPUs in the DC... or designing a smaller chip for data centers... or one with more cores... or so on.
As for what's in it for Apple, it would be the profit from selling a hopefully large number of chips, but adding official Linux support and also commuting to an entire new market for a minimum of three years is probably far higher a cost on focus than any potential profits.
And then they could resell the box they designed as a server, either with Linux support out of the box (unlikely, but since in this mystical scenario they'd have to write kernel patches for Linux to get it to boot...) or with a build of macOS that could be set up headlessly, in order to recoup the development costs. Apple shops that want machines to run Compressor or Xcode's build server would eat that up.
Then the gains of scale, making a CPU just of the Mac Pro would mean too low production numbers, but with data center usage would drive those up - especially if Apple also sold it to other customers, e.g. bringing the Xserve back. For the OS they could run Linux virtualized or they give the native Linux on Mac developers a hand.
What's interesting to me is to see if they'll use M-series chips in their own datacenters. They already run Linux there apparently.
Or do you mean that they'll start selling parts (SoCs)? Not in a million years :-)
AR/VR/more wearables would be another.
Home appliances/electronics would be another.
Edit: Mistyped 16 as 1, sorry about the confusion
edit: you have to select the SSD size then you can choose the RAM.
I could alter my usage patterns and work with less. But people not used to having loads of ram don't know what they're missing out on.
As a data point: I am running node + iOS Simulator (along with XCode + VSCode + Chrome with 50+ tabs) setup on M1 16GB and it works fine, I also keep them running while I take a break to play a LoL match. Works great for me.
I’m curious to know if it’d work for you, do you have access to an M1 to try out your workflow on? The max ram ‘issues’ seem to be way overblown.
Another [2]: > I really want the next MacBook Pro to support driving two 4K monitors over thunderbolt and have an option to buy 32 gigs of ram.
Meanwhile the last Intel MacBook Pro supports driving four (4!) 4K displays [4]. Apple silicon is far ahead in benchmarks but how does speeds and feeds translate into what customers actually want?
Battery life is impressive but unfortunately not the usual differentiator during a worldwide pandemic. The M1 Macs are quite quiet (the first MacBook Air without a fan—in 2020!) meanwhile the Intel Surface Book was fanless in 2017. We shot the messenger of the recent Intel attack Apple ads [5] but message is still worth reading. I bought an M1 MBA and realized the speed didn’t make a difference as my consumer computer. For the first time in decades I’m not sure if Apple provides the most pleasurable experience.
[1] https://news.ycombinator.com/item?id=26956336
[2] https://news.ycombinator.com/item?id=26955682
[4] https://support.apple.com/en-ca/HT210754
[5] https://www.macrumors.com/2021/03/17/justin-long-get-a-mac-i...
> meanwhile the Intel Surface Book was fanless in 2017
The MacBook was fanless in 2015 and, like many other fanless designs using Intel chips, it was slow.
Well those people must have been bullish on Apple Silicon and 'not just' M1. They think its worth skipping over M1 rather than going all in on the 1st gen product which at the time had primitive support for most mainstream software, especially for developers.
Maybe Apple knew that the M1 could not drive more than 1 monitor on the Macbook Air and in fact left that limitation in with a small disclaimer.
Perhaps they will announce this capability in the M2 Macs.
I agree with your nuance.
> The MacBook was fanless in 2015 and, like many other fanless designs using Intel chips, it was slow.
The Surface Book 2/3 and M1 MacBook Air are not slow (hence the point of my comparison)
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
Was quite a painful wait (as my Macbook 12' got water damage, with the SMC limiting in to 1Ghz).
https://asia.nikkei.com/Business/Tech/Semiconductors/MacBook...
I'm guessing the 2021 MacBook Pro is going to be the fastest laptop ever made.
This could be it tho, but there are probably some desktop TDP chip current gen laptops out there.
https://www.apple.com/newsroom/2021/04/apple-commits-430-bil...
I'm glad we can live that feeling again, even if just for a short while.
I’m responding to a HN commenter who was not just impressed about the power of the M1 but hyperbolically asserts that it is better than everything else yet the next top voted HN comment demonstrates otherwise with a demand for downgraded feature. The tenor of those reactions are opposed and my aim is to reflect the nuance
I love the direction Apple is headed with their hardware.
"tentatively known as the M2"
Blasphemy! Plus then N+1!
It's going to take years to run proper Linux on M1 and even more for the ecosystem to catch up to x64_86.
For me it's reasonable to keep using AMD Ryzen 5000 which is faster than M1 on my multithreaded workloads anyway despite using 7nm. Plus it has better GPU, more memory, more storage, more ports and supports multiple monitors.
Sure it is more expensive, but that's just because my segment is geared to pros with higher requirements. Apple currently has no laptop offering on this tier.
It seems the Mac aficionados (Especially the M1 fanatics) are in denial of the degree of lock-in with the Mac as it gradually descends into become nearly as locked in as an iPhone.
I'd give it 0.1 out of 10 for repairability. At least with the latest Surface line-up the SSD can be upgraded.
I wonder if Apple is familiar with the Osborne effect[1].
They can refresh the core in the M1 of course, and I expect they will do that yearly like the AXX cores, but it would be weird to go even 2 generations of the SoC without addressing the pro cpu.
I meant "soldered ram next to SoC with cpu and gpu cores", i.e. not DIMMs and 200W PCIe GPU. For a pro chip in a desktop form factor I think DIMMs and PCIe graphics are inevitable, and that's the interesting storyline in the evolution of the M1. We know they'll produce better M1's, but the integration with 3rd party graphics, the choice of DIMM type and so on is interesting.
It doesn't hurt to wait a little or perhaps skip the first gen versions due to the extremely early software support at the time, rather than getting last year's model (Nov 2020) and suffering from the workarounds and hacks to get your tools or software working on Apple Silicon.
I won't have a wobbly software transition unlike most M1 early adopters. Afterwards, I'll most certainly skipping M1 to something higher. (Likely M2)
Like I said before, wait for WWDC and don't end up like this guy. [0] Those on Intel (Especially those on 2020), no need to run for the M1, Just skip it and go for M2 or higher.
Downvoters: Here's another foot-gun Apple hid behind the hype squad. [1] For the iPad Pro, if you bought the 2020 version, your keyboard is incompatible with the M1 version meaning you have to fork another $350 for a new one that works.
At the time of the Macbook Air M1 launch (Nov 2020), tons of software issues, even the recovery system fell apart for most people for M1. Even upgrading to this on launch day right here with those issues was an instant no deal.
Once again Intel Mac users, plenty of time to migrate to something even better. (M2 or higher)
[0] https://www.zdnet.com/article/i-sold-my-old-ipad-pro-to-back...
[1] https://arstechnica.com/gadgets/2021/04/new-12-9-inch-ipad-p...
To be honest, it has all been much smoother than I expected, but YMMV.