ARM based MacBook Pro discussion [merged]

[Falhófnir]

RIP Bootcamp gaming

Think Apple are more interested in putting rocket boosters under native gaming - they might just be able to do it if a developer can target the combined Mac and iPad markets with a single port. Apple arcade, controller support, further unifying the platforms, and providing more power - the pieces are there for Apple to make a grab for a slice of the gaming market. The Nintendo Switch shows Arm ports of principally Windows/ x86 console games can be done where there's a will to do it/ market to buy it.

 

[happy-raccoon]

on one side, the switch to ARM is a quite interesting rumor. But it gets me thinking now. I finally have enough money for purchasing a 16" MBP. It's the machine of my dreams that I've been waiting for since the keyboard disaster, fulfills all my needs, now also with a sweet 16" screen it has everything I want. But I am heavily questioning this purchase right now, not because I want ARM but because

a) How long will Apple deliver new macOS Versions for Intel Macs?
b) What is going to happen to the resale Value of Intel Macs?

I did some research and found that 2007 with Leopard was the last major release for PowerPC Macs after the switch from 2005.

I just don't want to spend a huge amount of money that will get software support dropped in 3 years, this machine should last me 5+ years.

 

[MakeAppleAwesomeAgain]

I did some research and found that 2007 with Leopard was the last major release for PowerPC Macs after the switch from 2005.

I just don't want to spend a huge amount of money that will get software support dropped in 3 years, this machine should last me 5+ years.

Leopard was updated to 10.5.8 in August 2009 and received security updates until 2012, I think they did a great job with it and the cleanup in SL was necessary. History repeats itself, Apple will be transitioning once again and macOS is in dire need of massive 'under-the-hood' cleanup.

 

[maflynn]

History repeats itself, Apple will be transitioning once again and macOS is in dire need of massive 'under-the-hood' cleanup.

I'm not denying that but just looking at the feature list of OS from the prior versions, there hasn't been a major OS update for some time. APFS is probably the biggest OS specific feature in a long time. The rest seem to fall in Safari, Photos, and its music app (dropping iTunes)

 

[D.T.]

I tend to favour that hypothesis, too. Prosser says pro apps support may be coming to the iPad.
Looks like "laptopisation of iPads" is more likely than "iPadisation" of the MB lineup.

https://twitter.com/i/web/status/1252187152831692800

If [when?] XCode makes it to the iPad, iPadOS needs to be ready to provide native external monitor resolution (or at least native aspect ratio even if it's scaled). XCode has a really information dense UI, and I don't know that it can be effectively iOS'ed (like MS Word for example). On top of that, many of us (devs) are also working against an API, and need a browser/command-line/other tools to debug that end of the product - FYI, I'm not suggesting a full on windowing system, etc., it's just a complex design consideration due to the nature of the work.

 

[Nugget]

On top of that, many of us (devs) are also working against an API, and need a browser/command-line/other tools to debug that end of the product - FYI, I'm not suggesting a full on windowing system, etc., it's just a complex design consideration due to the nature of the work.

You know that wonderful feeling when you finally figure out how to solve a pernicious bug in your code and you get to go close the 30 or so browser tabs that you opened up while you were figuring it all out. It's hard to imagine how that would work in iOS.

 

[thedocbwarren]

iPad Pros and ARM in general are a lot more capable than I think is beleived. Yes it's currently designed for power management like all low power processers are, but i can see using an iPad to do iOS dev.

You know that wonderful feeling when you finally figure out how to solve a pernicious bug in your code and you get to go close the 30 or so browser tabs that you opened up while you were figuring it all out. It's hard to imagine how that would work in iOS.

 

[Nugget]

iPad Pros and ARM in general are a lot more capable than I think is beleived. Yes it's currently designed for power management like all low power processers are, but i can see using an iPad to do iOS dev.

The discussion was focused more on the UI capabilities offered in iOS and not so much the processing power in the hardware. The iPad can do a lot, but I always feel like I'm wearing mittens when I'm using it. It just struggles with complicated workflows and, as @D.T. points out, the solution can't just be to slap a regular windowed interface onto iOS. It's a tricky problem.

 

[cullenl87]

Don’t they have to show this off at WWDC? Don’t they have to get the ball rolling and developers excited for this move? Or are they just going to put this under the umbrella of Catalyst to not openly knock Intel and AMD?

 

[BootLoxes]

Don’t they have to show this off at WWDC? Don’t they have to get the ball rolling and developers excited for this move? Or are they just going to put this under the umbrella of Catalyst to not openly knock Intel and AMD?

I am sure many big developers are already well aware of the transition

 

[Birkan]

I think transition to ARM based Macs will be much smoother than we actually imagine. Some people keep saying stuff about x86-64 apps being a huge problem but some of them are already or very close to being AArch64 compatible.

Chromium powered browsers (Chrome, Edge) are already AArch64 compatible. Chromium powered Electron also released an AArch64 compatible version on Windows on Arm (WoA) so apps like Spotify, Discord, Slack or VS Code can and will be AArch64 compatible.

Bunch of smaller but widely used software such as git, homebrew, docker, clang, llvm, gcc, nodejs, python3 are also have AArch64 compatible versions thanks to various efforts from ARM or Raspberry Pi 3/4 community.

They are already deprecating kernel extensions and moving some of the functionality into userspace so I don't expect any legacy software using them to survive even on x86-64 in upcoming macOS versions. Probably no OpenGL or OpenCL as well since they are also deprecated. Userland graphics drivers from AMD or Nvidia(???) will be interesting to see.

Also, don't forget that many if not all Catalyst apps will be running natively on AArch64 since they target system libraries/frameworks instead of an architecture so with a recompilation on newer versions of Xcode they should be good to go.

Of course, there will be bunch of software such as Adobe products or Microsoft Office which do not have official AArch64 versions yet but given Adobe's relationship with Apple and Microsoft's commitment to WoA, we should expect them to get ready quite soon.

Bootcamp with WoA might be possible if Apple decides it won't make ARM powered Macs like iPads. Functionality of WoA would be a big question as well since that will probably mean no traditional games or only emulated 32-bit apps since AArch64 compatible software is hard to come-by.

And honestly, I hope they do not provide any x86-64 emulation on AArch64 so people actually start working on AArch64 versions instead of relying on translation layer. Let's hope at least for some software news during WWDC 2020!

 

[dudeslife]

I look forward to jailbreaking macOS on ARM notebooks in order to install what I want on there..

 

[09872738]

I look forward to jailbreaking macOS on ARM notebooks in order to install what I want on there..

No jailbreak required. Its not like ARM systems are locked down by default. Its the OS, not the hardware

 

[Stephen.R]

Of course, there will be bunch of software such as Adobe products or Microsoft Office which do not have official AArch64 versions yet but given Adobe's relationship with Apple and Microsoft's commitment to WoA, we should expect them to get ready quite soon.

Microsoft is so committed to Windows on Arm, Office for Windows on Arm runs under the x86 32bit emulation, a mere <checks notes> 7 1/2 years after the release of their first Windows on ARM, and 9 1/2 years after the announcement of said Windows on Arm.

If they weren't going to release a desktop Office using native arm on Windows, what exactly makes you think they'll release one for macOS?

 

[imrazor]

Think Apple are more interested in putting rocket boosters under native gaming - they might just be able to do it if a developer can target the combined Mac and iPad markets with a single port. Apple arcade, controller support, further unifying the platforms, and providing more power - the pieces are there for Apple to make a grab for a slice of the gaming market. The Nintendo Switch shows Arm ports of principally Windows/ x86 console games can be done where there's a will to do it/ market to buy it.

Yay, now I can play Candy Crush Saga on my ARM Mac! Who needs Civilization, Tomb Raider, Fallout or Mass Effect?

On a more serious note, if Apple decides to go this route, they will likely have mobile (ie, ARM) developers fill the initial demand for gaming. I've not taken a serious look at any iPad or iPhone games in a while, but most of the ones I have seen lack depth or a decent narrative. Plus keyboard and mouse control may be sacrificed for touch or controllers. Ugh. And then there are the micro-transactions (though to be fair, an increasing number of PC games have this 'feature'.)

 

[Jamjer]

on one side, the switch to ARM is a quite interesting rumor. But it gets me thinking now. I finally have enough money for purchasing a 16" MBP. It's the machine of my dreams that I've been waiting for since the keyboard disaster, fulfills all my needs, now also with a sweet 16" screen it has everything I want. But I am heavily questioning this purchase right now, not because I want ARM but because

a) How long will Apple deliver new macOS Versions for Intel Macs?
b) What is going to happen to the resale Value of Intel Macs?

I did some research and found that 2007 with Leopard was the last major release for PowerPC Macs after the switch from 2005.

I just don't want to spend a huge amount of money that will get software support dropped in 3 years, this machine should last me 5+ years.

Leopard was updated to 10.5.8 in August 2009 and received security updates until 2012, I think they did a great job with it and the cleanup in SL was necessary. History repeats itself, Apple will be transitioning once again and macOS is in dire need of massive 'under-the-hood' cleanup.

Remember the G4? Third party devs like adobe only support for 3 years after. 16" Macbook released in 2019, ARM based in 2021 and looking the timeline calculation that latest G4 did, the 16" 2019 model will only have 5 years or less of life rather than 8+ years like older macs did. Of course apple will still support it while in transition time, but only apple itself not 3rd party devs.

 

[Falhófnir]

Yay, now I can play Candy Crush Saga on my ARM Mac! Who needs Civilization, Tomb Raider, Fallout or Mass Effect?

On a more serious note, if Apple decides to go this route, they will likely have mobile (ie, ARM) developers fill the initial demand for gaming. I've not taken a serious look at any iPad or iPhone games in a while, but most of the ones I have seen lack depth or a decent narrative. Plus keyboard and mouse control may be sacrificed for touch or controllers. Ugh. And then there are the micro-transactions (though to be fair, an increasing number of PC games have this 'feature'.)

This is my point - Civ VI is available on iPad (and Arm based Nintendo Switch - as are several sizeable Windows first titles such as Fortnite, Witcher 3, Diablo 3, Skyrim to name a few). That shows there's no technological constraint by going to Arm, it's an economic decision by the studio as to whether it's worth the cost to do a Mac port. Combining the Mac and iPad markets might tip the balance in favour of doing it for more big name games.

iPad has something like a 300M installed base, Macs something like 100M, but even more than that, at the moment only iMacs and 15/16" MacBook Pros really have the graphics grunt to run anything like the titles you mention satisfactorily, whereas the iPad Pro (from Apple's own mouth) has Xbox One S level graphics performance. Currently you could say the bulk of iPads will be less than that, the Air/ mini 5 only has half that performance, and the iPad with A10 has about 1/3 - but going forward the installed base of ever more powerful iPads is only going to grow. The A13 has about 2/3 of the A12X graphics performance, and the A14 is likely to get even closer to parity. Add in equally or more powerful MacBook Airs and you might just have a compelling target market to attract the big game studios.

 

[imrazor]

Add in equally or more powerful MacBook Airs and you might just have a compelling target market to attract the big game studios.

From a technical perspective, you're right. The problem is the word might in your quote. I'm glad that a few major PC games have been ported to the Switch (and I may get one of those yet), but there certainly hasn't been a mass migration of AAA titles to the Nintendo. The advantage of the current x86 architecture is that one can simply reboot your Mac and presto!, you're ready to play an honest to goodness AAA game. As well as run any Windows software you like. That makes switching easier and more palatable for plain old Windows users, as well as 'serious gamers.'

EDIT: Maybe Windows users are less attractive to Tim Cook than getting Android users to switch to the Apple walled garden? Perhaps Apple executives are looking forward to a post-Windows world. If they are, I hope we're not in for a sea of lousy mobile-like games.

 

[Falhófnir]

From a technical perspective, you're right. The problem is the word might in your quote. I'm glad that a few major PC games have been ported to the Switch (and I may get one of those yet), but there certainly hasn't been a mass migration of AAA titles to the Nintendo. The advantage of the current x86 architecture is that one can simply reboot your Mac and presto!, you're ready to play an honest to goodness AAA game. As well as run any Windows software you like. That makes switching easier and more palatable for plain old Windows users, as well as 'serious gamers.'

... as long as you're one of the minority of Mac users that have a 16" MacBook Pro, or an iMac/Pro or Mac Pro with dedicated graphics, sure. But that's little good to Apple. I'm talking about Apple actually building a compelling platform of Arm Macs and iPads to attract the big game developers. To make it worth their while to do an Apple port of their game. Macs alone, with their tiny market share, will never manage it. Especially if people are loading up Windows to game. Leveraging in the much larger iOS market and further supplementing it with relatively powerful consumer/ mass market Macs like the Air gives Apple the opportunity to get a slice of the ($250BN industry) pie, and given services and sales are the core of their strategy now this must be quite an appealing prospect.

 

[Macbookprodude]

PowerBooks running x86 emulation - enough heat to fry eggs.

If you are referring to the 1.67 DLSD model powerbook, Mine isn’t too warm.

 

[subjonas]

Speaking purely from the perspective of a user who depends on my Mac for work, I don’t care what chip is powering it. Ultimately, I don’t even care what OS it’s running. Bottom line is, as long as I can do what I need to do with the least amount of trouble, what else matters?

But I’ll say this- if Apple’s goal is to phase out x86 and/or macOS, and replace it with Arm and/or a more robust ipadOS, they will have to be aggressive about it, at least in the industrial market. If they simply offer two product lines, I don’t see why businesses and their software developers would jump over, unless Apple forces a deadline for developers to transition their software, or proves that there is major financial motivation for the transition. Because businesses aren’t tech enthusiasts. They just want things to work so that they can do what they need to do to make money. And x86/macOS works.

If Apple does force a move to ipadOS though, which is not an industry battle-tested platform, with businesses’ livelihood on the line, I can’t help but think they run the risk of losing some of their market to Windows.

The only possible issue I see with the switch to MacOS on arm is that the transition could take some time. And while businesses are switching to new hardware, which they will surely be slow to do, developers, including Apple, would need to support their software on both chips and also have full file inter-compatibility.

 

[Birkan]

Microsoft is so committed to Windows on Arm, Office for Windows on Arm runs under the x86 32bit emulation, a mere <checks notes> 7 1/2 years after the release of their first Windows on ARM, and 9 1/2 years after the announcement of said Windows on Arm.

If they weren't going to release a desktop Office using native arm on Windows, what exactly makes you think they'll release one for macOS?

You are right on the emulated 32bit office on WoA but probably it wasn't that much important for them to have native Office on Windows RT or later versions as much as it matters now. Their recent push on services with Office, Azure, Windows 10X or on devices with Surface Pro X or upcoming dual screen devices makes me think they will release a native AArch64 version of Office sooner or later. That would be even more true especially if they think Apple might release an ARM powered Mac device, maybe kickstarting a bigger transition to ARM from x86-64 than just themselves. Also, don't forget that they already ported some parts of Office to iPad so it will be even easier now compared to 2012 or 2015. So let's say I'm hopeful but we'll see how it goes.

 

[Woochoo]

So let's say I'm hopeful but we'll see how it goes.

Wait, is optimism that even allowed here? I thought this thread was only about how good x86 is and how bad ARM is

Now for real, people loves complaining and seeing the negative things (even when those things may not end happening!) or exagerating about them, and ignoring all the good things ARM has to offer. I'm gonna try to cheer people up with the good stuff (and the stuff that it's not as bad as you tink):

1. Apple ARM chips are freaking beasts. They are already on par or even beating Intel laptop chips not only in benchmarks but also in "real world" usage like video exporting. And that was with a 8 core iPad Pro (4 high perf cores, 4 efficient cores), the upcomming Macs are supposed to have 8 performance + 4 efficient cores (the double for heavy tasks).
2. A chips have specific processing units (and more may come) like the NPU. Did you see Nvidia RTX Voice? It's freaking amazing, and that's the kind of stuff you can do with AI specialized cores (like NPUs, tensor cores...), not only scientific computation. You can't do this kind of stuff in any x86 chip unless the same laptop already includes an RTX Nvidia.
3. Extremely efficient: I'm talking about a chip with about 5-10W consumption doing the same job in the same task as a 45W (at least) consuming chip, so battery lifetime will skyroket. And that also means no more overheated and throttled CPUs that can't hold sustained loads at max frequency. That probably will also mean lighter laptops as thermal disipation required will be smaller.
4. Project Catalyst will be like Rossetta to ease the transition from x86 to ARM, and if the rumours are right the first macOS with ARM should be able to run natively x86 and ARM instructions. Even if the first are translated at low level having a bit of overheat, the powerful SoC should make it hard to notice for most users.
5. You will probably be able to execute iOS and iPadOS apps on your mac, that's great for both consumers and developers
6. No more situations like the 2013 15" MBP being "refreshed" with the same CPU because of Intel delays on their roadmaps. No more
7. Bootcamp will disappear? Maybe, maybe not since Microsoft already has ARM Windows. If Apple succeeds with this, Microsoft will put way more efforts in ARM native software (not emulating x86 with extra layers) and developers aswell.
8. USB 4 will include Thunderbolt 3, that means it will probably end up in the ARM Macs aswell
9. For those who don't want to become early adopters, the 16" (and probably a new 14") x86 Macs still exist being great machines that will last for years performing well enough.
10. Apple already got big guys like Adobe into developing for iPad Pro which is way more niche than Macs as a whole. Aside from all the professional software developed for iPad Pro which should be easily portable to ARM Macs, Apple can get more big suites into new Macs without much problems.

So all in all it seems that the only real problem will be the inital lack of software the first year until devs get their hands on it, from then things will get easier and easier for everyone.

 

[subjonas]

Actually, I’m curious, how did the transition go in the enterprise market the previous times that Apple did a major chip architecture change? Are there any possible key technical differences between those times and this time? And was there a big difference in the Mac’s ecosystem and reach/market share between now and then, and would that significantly affect the transition?

 

[Nugget]

Actually, I’m curious, how did the transition go in the enterprise market the previous times that Apple did a major chip architecture change? Are there any possible key technical differences between those times and this time?

There was one huge technical difference. The last time Apple attempted an architecture change they were moving away from an abandoned, boutique, languishing CPU architecture and moving to the near-universal, mainstream, compatible architecture that the rest of the universe was already using.