Mac With Apple-Designed Arm Processor Coming in First Half of 2021

[Mago]

No single word in Kuo or Blomberg article point to ARM (or Any) Specific Platform, For Apple should be wise to Adopt AMD Zen (x86 compatible) given AMD also licenses it the same as ARM does, comes in sense as also AMD uses the same process as Apple to manufactrure its CPUs.

 

[Zdigital2015]

All of which are cisc processors. There's a lot of software that does directly interact with hardware, and relies on those robust instruction sets. sure, for some apps it's a port. But for a lot of others, it's a rewrite. And in either case, it's not likely to be cost effective to maintain two fundamentally different code paths. This isn't Rosetta....it's a different world now...a lot of people rely on intel.
[automerge]1582570259[/automerge]
There's a massive difference between virtualizing a cpu and emulating one.

SPARC and PowerPC were both RISC, not CISC.

Intel doesn’t get to dictate the rules of computing...I’m sick of their milk things to last drop and trickle out tech as they see fit or their marketing department deems necessary to try an c*ck block AMD. If only MS had had more balls in the 1990’s maybe Intel wouldn’t seem to be the only game in town...tired of Intel and while I appreciate their contributions, they need to be humbled and laid low.

 

[oneMadRssn]

Yes! The zillions of macrumors fanboys who screamed for a G5 PowerBook are finally going to get their wish. (except with a RISC ISA and implementation that’s more than 10 times faster, and no water cooling needed!). With probably even more apps available for it.

Or do all of you really want that water cooling?

Honestly, if Apple continued naming their in-house laptop / desktop chips as the G-series - G5, G6, etc. - that would probably sway a lot of the old school techies for the nostalgia reason alone.

 

[bluecoast]

I’m really surprised that this is so far out and that at least one or two of the entry level consumer lines aren’t coming this year.

Surely it would be possible for an entry level consumer MacBook to go ARM this year?

If not I’d guess it’s one or all of the below:

- multithreaded multitasking pc/Mac performance is just not there on the A class chips yet & that the 2020 chips will give them the boost here that’s needed.

- MacOS and the various frameworks just aren’t there yet. One could expect a closer merge in behaviours to the iPad with a Mac ARM (sleep states, the system expecting to work with a modem just as much as with WiFi or wired networking, presumably a UI makeover etc) and this will all take time to do right. Also see: Catalina.

- it’s going to take devs time to then get their software working right.

- Apple wants to get a 5g modem out of the door and ensure that all new Arm macs have the ability to add this as a order option.

 

[MRrainer]

I will miss VMWare Fusion.

It will be interesting to see what the performance is for other "popular" apps.
Like Visual Studio Code, Firefox, Sublime, OmniGraffle, LibreOffice, MS RDP Client...

 

[firewood]

Judging of how bad first generation products are compared to later iterations in the Apple lineup, then 2024 is the year to buy my first ARM Mac

Unfortunately, I'm a software developer. So just like I had to be an early adopter and get a day 1 PowerMac 7100, a day 1 Intel PowerBook, a day 1 iPhone 5s (arm64), etc., I'll probably have to get in line to get whatever new-ISA Mac (or iOS device) they first make available to non-NDA developers.

 

[mrkite77]

The "I don't care about performance, just compatibility" market is small. Just ask Sun, SGI, etc.

That makes no sense.. both Sun and SGI were all about performance on proprietary architectures that weren't compatible with the rest of the market. In fact, as they were losing market share, both companies switched to x86 in a last ditch effort to remain relevant.

You kind of proved the opposite of your point with those two examples.

 

[jdiamond]

Everyone seems to keep forgetting that Microsoft is also pushing to get Windows on ARM. So you'll be likely able to run NATIVE Windows on ARM (ADM bootcamp). The bigger concern is all these Linux utilities that are only available in binary. That being said, you don't need native speed emulation EVERYWHERE - if your native OS can do the high performance things, you only need x86 emulation for the legacy stuff.

 

[panjandrum]

Gotta say, I don't think Apple really understands just how deep the need for Bootcamp runs. Maybe I'm wrong and it's just the power-users and Gamers that Apple will end up losing, but in the end, they shouldn't even feel safe losing those sector. Hopefully they will pull something out of their bums and make this all work, but I've got a pretty bad feeling that switching to ARM will wind up being a mistake. Now, if they offered it as an alternative, then that's a different story. "Do you want your 13" MacBook Pro with Intel, AMD or ARM" and then the end-user gets to choose based on the pros and cons of each. But moving to ARM? That would immediately eliminate 1, 2, 3, 4, 5, 6, 7, 8 people I can count right off the top of my head just from my closest inner-ring of friends and family that absolutely could NOT buy another new Mac, because as much as they wish it was otherwise, they need to run Windows well.

 

[CarlJ]

... and this is one of the areas I think a custom Apple chip would really shine. The A#X chips in the iPads are pretty good at sustained performance without fans or throttling...

Yep, and the chips we've seen from Apple thus far have been optimized for use in a 1/4" thick sealed box with very tight restrictions on the thermal and power budget. I think there is plenty of room for them to make chips optimized for use in a fanless (or nearly so) tiny laptop (Air or "MacBook nothing" form), as well as more powerful chips for a MacBook Pro class machine (with substantial fans and bigger batteries allowing for more elaborate chip designs, more cores, and higher clock speeds), and desktop systems that are much more "sky is the limit" as far as power and thermal budget.

Apple, with Motorola, and then IBM, and then Intel, has always been a passenger on the bus - they can get on and off, and they can make suggestions to the driver, but they don't really have much control over where it's going. With a switch to their own ARM-based chips, suddenly Apple is driving - they can decide where to go, and how fast, and when to change direction, and when to add more resources to get what they want when they want it (rather than asking Intel, "pretty please?").

 

[rp2011]

If that great software is compiled with Xcode, as nearly all Mac software is, then recompiling it for Apple's ARM-Mac chips is not going to be crazily difficult.

Yep.
And there’s people complaining about losing 32 bit support. Legacy apps and consumers will need to decide if they want to stay in the past or keep moving forward.
Eventually we have to discard the legacy baggage that bogs everything down. That’s software, hardware and consumers.

 

[techwhiz]

I believe it when one ships. I will also not be buying anymore Macs. VM support to run Linux X86 is mandatory for me to work.

 

[realtuner]

Only on MR could you find people arguing with an actual processor/chip designer (@cmaier) about what is/isn’t possible when switching to another architecture.

 

[cmaier]

I think there will be a major rewrite to a new Mac OS based system. Darwin, regardless of how robust and stable it is and what it gave us ... we may see something more aligned to a particular build or part of Linux.

That said the real gains we'll see is to customization of the instruction ... such as what RISC-V promises. who knows.

What does RISC-V have to do with it? If Apple wants to customize their ARM-based instruction set, they are free to do it.
[automerge]1582573218[/automerge]

Only on MR could you find people arguing with an actual processor/chip designer (@cmaier) about what is/isn’t possible when switching to another architecture.

No, also in my house. My little girl is very argumentative. I get no respect

 

[Rafterman]

I just bought a maxed out 16". I'm hoping that by the time it gets "old" (3-5 years), all the third party software will have caught up by then.

 

[cmaier]

Gotta say, I don't think Apple really understands just how deep the need for Bootcamp runs. Maybe I'm wrong and it's just the power-users and Gamers that Apple will end up losing, but in the end, they shouldn't even feel safe losing those sector. Hopefully they will pull something out of their bums and make this all work, but I've got a pretty bad feeling that switching to ARM will wind up being a mistake. Now, if they offered it as an alternative, then that's a different story. "Do you want your 13" MacBook Pro with Intel, AMD or ARM" and then the end-user gets to choose based on the pros and cons of each. But moving to ARM? That would immediately eliminate 1, 2, 3, 4, 5, 6, 7, 8 people I can count right off the top of my head just from my closest inner-ring of friends and family that absolutely could NOT buy another new Mac, because as much as they wish it was otherwise, they need to run Windows well.

If every single current Mac owner never bought another Mac, but 10% of non-Mac owners who own iPhones or iPads buy the new ARM macs, it will be a huge gain for Apple.

 

[Janichsan]

Still don't think it will be ARM-based, rather, a full-custom Apple design. Stay tuned!

Doubtful. Creating a capable, completely new CPU architecture from scratch is hard. Apple has experience with ARM. If they really want to cripple the Mac, that's the route they will go.

- No more dual boot. Looks like we're confined to virtual machines.

Not even virtual machines. Emulators.

 

[cmaier]

All of which are cisc processors.

I assure you - PowerPC and SPARC were most definitely *not* CISC processors.

I know this for two reasons.

First - it’s widely known.
Second - I designed PowerPC’s at Exponential Technology (the x704 processor) and SPARC at Sun (UltraSparc V). I’m pretty sure I didn’t put any microcode ROMs in there, I didn’t add support for any variable length instructions, and the only way to access data memory was load/store instructions. So, yeah, I’m going with “those are RISC processors.”
[automerge]1582573513[/automerge]

No single word in Kuo or Blomberg article point to ARM (or Any) Specific Platform, For Apple should be wise to Adopt AMD Zen (x86 compatible) given AMD also licenses it the same as ARM does, comes in sense as also AMD uses the same process as Apple to manufactrure its CPUs.

AMD does not license it “same as ARM does.”
[automerge]1582573576[/automerge]

Only on MR could you find people arguing with an actual processor/chip designer (@cmaier) about what is/isn’t possible when switching to another architecture.

Was away and came back and started paging backward and saw someone ”correcting” me by claiming PowerPC and SPARC are CISC.

So it’s going to be one of those days.

 

[Falhófnir]

Yep, and the chips we've seen from Apple thus far have been optimized for use in a 1/4" thick sealed box with very tight restrictions on the thermal and power budget. I think there is plenty of room for them to make chips optimized for use in a fanless (or nearly so) tiny laptop (Air or "MacBook nothing" form), as well as more powerful chips for a MacBook Pro class machine (with substantial fans and bigger batteries allowing for more elaborate chip designs, more cores, and higher clock speeds), and desktop systems that are much more "sky is the limit" as far as power and thermal budget.

Apple, with Motorola, and then IBM, and then Intel, has always been a passenger on the bus - they can get on and off, and they can make suggestions to the driver, but they don't really have much control over where it's going. With a switch to their own ARM-based chips, suddenly Apple is driving - they can decide where to go, and how fast, and when to change direction, and when to add more resources to get what they want when they want it (rather than asking Intel, "pretty please?").

Yep sometimes it feels like when people see 'switching to Arm' they think of it as switching to just another chip vendor, not what it actually is, Apple taking almost complete control of the chip design. When you think of it this way, it seems pretty logical you're going to get better results with a custom designed solution wherein Apple can target all the performance bottlenecks as opposed to an off the shelf part that has to have good general performance but will ultimately be designed in line with priorities which won't perfectly match Apple's.

 

[smulji]

If every single current Mac owner never bought another Mac, but 10% of non-Mac owners who own iPhones or iPads buy the new ARM macs, it will be a huge gain for Apple.

Even if 10% of only iPhone users bought the new ARM Mac, that would be 90 million new Mac users right there => close to the same size of the x86 Mac user base

 

[cmaier]

Even if 10% of only iPhone users bought the new ARM Mac, that would be 90 million new Mac users right there => close to the same size of the x86 Mac user base

And a much more dynamic user base, at that. (And fully bought-in to the entirety of Apple’s ecosystem. Gotta be at least 10% of people who use macs with android phones, install linux on the mac, use the mac just for windows, etc.)

 

[Zdigital2015]

Only on MR could you find people arguing with an actual processor/chip designer (@cmaier) about what is/isn’t possible when switching to another architecture.

All day long and twice on Sundays.

 

[oosamon]

12” MacBook making a return?

I hope so!

 

[Labeno]

Here's my two cents...

• Pros want highly tuned software: I see Apple helping companies port Pro software, like Adobe
• Developers want XCode: need to make those iOS apps
• Casual users want a web browser (for emails, browsing, cloud software like Office360, etc) and iMessage for chatting

So, in essence, there's not much software to port. If you want Microsoft software (I don't mean cloud software), then you'll probably need a separate Windows PC. Apple isn't in the business to create Microsoft PCs, they're in the business to get you hooked on great experiences exclusive to them. I like my privacy and mostly bug free ecosystem.

 

[MallardDuck]

Gotta say, I don't think Apple really understands just how deep the need for Bootcamp runs. Maybe I'm wrong and it's just the power-users and Gamers that Apple will end up losing, but in the end, they shouldn't even feel safe losing those sector. Hopefully they will pull something out of their bums and make this all work, but I've got a pretty bad feeling that switching to ARM will wind up being a mistake. Now, if they offered it as an alternative, then that's a different story. "Do you want your 13" MacBook Pro with Intel, AMD or ARM" and then the end-user gets to choose based on the pros and cons of each. But moving to ARM? That would immediately eliminate 1, 2, 3, 4, 5, 6, 7, 8 people I can count right off the top of my head just from my closest inner-ring of friends and family that absolutely could NOT buy another new Mac, because as much as they wish it was otherwise, they need to run Windows well.

Yep. One example: anyone who runs a small business and uses TurboTax for Business to do their taxes. That's windows only. And if they haven't already ported to Mac on Intel, there's no way they'll do it on ARM.