And iOS active installed base is equal to Windows, times they are a changing. With the Mac only making up 10% of Apple profits and slow growth it seems like the obvious candidate for a complete overhaul.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Plans to Ditch Intel and Use Custom Mac Chips Starting in 2020
- Thread starter MacRumors
- Start date
- Sort by reaction score
And iOS active installed base is equal to Windows, times they are a changing. With the Mac only making up 10% of Apple profits and slow growth it seems like the obvious candidate for a complete overhaul.
Thinking back to the PPC to intel transition, who wants to bet on an A chip development Mac Mini box being able to purchase after WWDC running a dev build of the latest macOS beta?
And then the A chip Mac Mini being released some time in the autumn - maybe even with the new MacBook. If not, I suspect that the MacBook is coming in 2019.
By the way, I suspect that Pro desktop Macs will be the last machines to move away from intel - if ever.
I really think that this will be about the Mac mini -> MacBook -> iMac -> MBP.
And then the A chip Mac Mini being released some time in the autumn - maybe even with the new MacBook. If not, I suspect that the MacBook is coming in 2019.
By the way, I suspect that Pro desktop Macs will be the last machines to move away from intel - if ever.
I really think that this will be about the Mac mini -> MacBook -> iMac -> MBP.
Last edited:
Yes I recall the challenge of running Windows XP in a virtual environment using Qemu on my PowerMac G5 it was flaky at best and maxed out the resource hungry G5's.
Totally. Imagine:
Macs with a combined App Store (with ‘marzipan’ apps)
Neural engine AI and machine learning advances in iOS also coming to the Mac
iPad components also coming to the Mac ie 5G
And Macs that are cheaper (look up the price lists for Intel processors, even Apple must still be paying quite a bit)
Finally someone who remembers the pain of switching from PowerPC to Intel. But for most of us it was worth the pain because switching to Intel enabled Apple to create more modern computers that are much more powerful than PowerPC, more efficient, and cheaper to buy.
With this, I don't see why Apple would make the hassle to switch to ARM. We already have really modern computers with Intel chips (12-inch MacBook), and it would create a developer nightmare to switch. Is it worth to create this situation when people are buying more iOS devices than Macs?
But that's not Apple's problem - that's your problem. Don't want to carry two laptops? Don't use Windows.
I still don't see why Apple wouldn't use their own chips.
It's a good idea for the consumer line up. The A series is clock for clock a lot more efficient than Intel/AMD chips and the GPU is extremely good too. Power consumption very low.
How does it benefit the consumer who has spent a significant amount of money on software designed to run on Macs with Intel architecture?
This. Can you imagine the power of a 16-core or 32-core A12X chip and far less expensive then an Intel Xeon CPU and arguably significantly faster. Not only that if that's in a laptop-like form-factor the battery life could be a week. Yum.
Some of us don't have the option to not use Windows in our professional lives so it would mean not using MacOS if I only wanted to carry one laptop around
You're the one who's saying people should buy multiple machines and as I've said before, I would sooner drop OSX than drop Windows based on my needs.
The biggest issue I can see with any arch transition will be ensuring that common software is there (e.g. Office, Adobe, etc. (and I don't mean any cut down version)) and that enough people don't care about software that is dropped.
If enough people do care, then it could become Apple's problem.
Apple forced to wait on Intel.... I almost laughed to death when reading that one. Apple sticks to old CPUs even when new ones are on the market for ages. Just look at mac Mini, Mac Pro, and usually also the rest of the lineup.
With its own chips, Apple would not be forced to wait on new Intel chips before being able to release updated Macs, and the company could integrate new features on a faster schedule.
TLDR; A possible move to ARM would not be motivated by tech. The challenges would also not primarily be technical. The main drivers would be to benefit Apple itself rather than something the customer wants or needs.
Let's be clear about one thing here; if/when Apple indeed moves to ARM processors for Mac, the motivation and main challenges will not be technological in nature.
And yet, with all this said, I'm not particularly thrilled about a possible transition to ARM. To me this doesn't seem to be a transition that would be driven by customer needs or wants. Similar to how the touchbar wasn't really something that anyone was asking for. Similar to how USB-C only was not really something that anyone was asking for. Similar to how removing the headphone jack was not really something that anyone was asking for.
If Macs were to move to ARM processors, I can't see that being motivated by any particular technological advances. The motivation really must lie in the domain of money and control. Things that benefit Apple itself, but do not really transfer clearly to customers. Me, being a customer, that makes me skeptical towards such a transition. It would take away a major reason for why I choose to use a Mac over a Windows or a Linux box.
And reading through these forums, what are customers really asking for?
Let's be clear about one thing here; if/when Apple indeed moves to ARM processors for Mac, the motivation and main challenges will not be technological in nature.
- Moving macOS to a new CPU -- all else being equal -- is not the biggest challenge. You can be pretty darn sure that they've had macOS running on ARM in labs for several years already.
- I would imagine chipset stuff being the bigger challenge. Things like memory controller, PCI, USB, Ethernet, Wifi, Bluetooth, disk controller, audio etc etc. Much of this is currently integrated in or near CPU:s. Apple would have much of this already, though not necessarily in a form suitable for Mac. Not a major problem though.
- "Porting" apps to a new CPU is mostly trivial these days. Load your project into XCode, click the build button and a few minutes later you're done. That's of course a massive simplification, it won't hold true for really large projects like Adobe products etc, but in theory, if the toolchain stays the same, it won't be a major issue to transfer apps. I have a 30 kloc hobby project myself of high performance low level code that currently builds on macOS, Linux, Windows, X86 and ARM. I'm pretty sure it would transfer to an ARM mac in about 10 minutes, most of that time would just be waiting for it to compile.
- What's more of a concern is that you would need an ARM Mac to build the project. No issue for large developers, but for smaller developers expect that this may not be their highest priority. There could be waiting time involved, and there are lots of semi-abandoned apps that are still functional and useful, but are not getting updated and probably wouldn't be brought over to ARM.
- The CPU used for macOS really has nothing to do with how easy or difficult it is to port Windows apps to macOS. Neither would macOS and iOS being on the same CPU really make it easier to transfer apps from one to the other. There are other initiatives that would though, like the rumored ability for 10.14 to run iOS apps.
- For running Windows apps as they are however, the CPU does matter. Parallels, VMWare, VirtualBox, BootCamp, ..., they all depend on having an X86 CPU. AMD processors would work just as well, but ARM wouldn't. An ARM Mac would lose the ability to run Windows apps the way it's done now.
- Sure, there is Windows for ARM. And yes, two years from now there could be more apps available than there are now. But no matter how they do it, there won't be even close to the amount of apps that are available for X86. But having the Windows OS available on ARM is not really the issue. That's the "easy" part.
- Sure, you could imagine software X86 emulation, some helper chip, or even a full X86 coprocessor. I'd question whether emulation would be fast enough, but it has come a long way in the past decade and I'll allow for th possibility that performance would be acceptable.
- The CISC vs RISC thing, or X86 having "baggage" is mostly a pointless argument. Modern X86 processors are effectively RISC chips with an X86 front end. They do things so advanced that it's hard even for experts to fully grasp what's going on. I certainly wouldn't presume to understand half of it myself, and I still love to get pretty close to the metal when I can.
- As many have noted, Apple has been through transformations before, and I'm pretty sure that if any business knows how to do that well, it's Apple. There would be issues for sure, but for the average user, if you're just using a web brower, plays some videos/music, and maybe use the Mac for music or video editing -- I could imagine a transition to be very smooth. Almost unnoticeable in terms of hiccups and things.
- Modern CPU performance is to a large extent a power challenge. Apple got this over a decade ago. Others get this too. This is true for the low end, but it's the same for the high end. This is the reason behind going multicore instead of running CPU:s at 20GHz. It's a physical limitation, and Apple will be equally unable to break the laws of physics as Intel or AMD or whoever. What they could do, of course, is take development in new directions. Innovate. Possible, but I'd imagine that Intel has a pretty massive lead.
- Performance on the high end. There's little doubt that they could make a 12" MacBook based on ARM and have that work out fairly well. Or a Mac Mini device. However, making something like the 2020 equivalent of the 8700K for a mainstream iMac, or something equivalent to Intel high-end mobile processors, or anything worthy putting into an iMac Pro or a Mac Pro -- well....... it's not theoretically impossible, but I'd hold it for very unlikely that Apple can pull that off by 2020. We have not really seen any evidence of this so far, not really heard any rumors that indicate that they have that level of performance. And for all of you who are looking at past Ax chip improvements and extrapolating linearly -- it doesn't work that way, sorry.
- Windows app interoperability. I wouldn't presume that anyone here prefers Windows over macOS, and we may not even depend on Windows apps for the majority of our work, but for a minority of apps I think lots of Mac users find that having easy access to Windows apps is extremely valuable. If nothing else, then for games. I don't think that should be underestimated. I wouldn't expect that most of us here get their main income from playing games. It's probably not something we spend a significant part of our time on either. But I would imagine many of us do enjoy playing a game every once in a while, and booting into Windows is really how you have to do that currently.
And yet, with all this said, I'm not particularly thrilled about a possible transition to ARM. To me this doesn't seem to be a transition that would be driven by customer needs or wants. Similar to how the touchbar wasn't really something that anyone was asking for. Similar to how USB-C only was not really something that anyone was asking for. Similar to how removing the headphone jack was not really something that anyone was asking for.
If Macs were to move to ARM processors, I can't see that being motivated by any particular technological advances. The motivation really must lie in the domain of money and control. Things that benefit Apple itself, but do not really transfer clearly to customers. Me, being a customer, that makes me skeptical towards such a transition. It would take away a major reason for why I choose to use a Mac over a Windows or a Linux box.
And reading through these forums, what are customers really asking for?
- Easier to use commodity hardware like graphics cards
- Easier to use Windows apps
- More raw performance in terms of CPU and memory
- Easier to upgrade systems
Last edited:
It is important to remember that AIM failed because IBM could not produce new PPC chips reliably or within the thermal envelopes needed for Apple's mobile Macs. The design stalled and no one was willing to fund the project further. I don't see a new Apple CPU design ending any differently than AIM.
An interesting thought. Intel has stated that they would do custom fab work on ARM and this rumor would line up with that fact. Then if Apple were to eventually tire of allocating resources for ARM CPUs, Intel would still be there to bring them back into the fold.
Sorry, but I won't be joining the party bus for that ride.
You can achieve pretty much everything using Windows which ultimately could be the deal breaker should Apple ditch Intel and catapult the Mac back to the dark ages. Upon which there would most likely be wholesale adoption of Windows 10.
1. Apple won’t have to wait for Intel for the chips to upgrade their computers.
No; they’ll have to wait on Apple to upgrade their computers.
2. Apple can have custom features built in to the chips.
Again, why is this needed on a CPU? Apple can foundry the chips and work with TSMC or other chip manufacturers to achieve their goals.
3. Other companies will have to re-write their code for these new chips.
Yeah, they will, or Apple can make a translation layer, and that’s a bit of a pain and it slows down the execution of the code. THat should soften the blow..
No; they’ll have to wait on Apple to upgrade their computers.
2. Apple can have custom features built in to the chips.
Again, why is this needed on a CPU? Apple can foundry the chips and work with TSMC or other chip manufacturers to achieve their goals.
3. Other companies will have to re-write their code for these new chips.
Yeah, they will, or Apple can make a translation layer, and that’s a bit of a pain and it slows down the execution of the code. THat should soften the blow..
Why do people trust gossips so easily?? The transition that Apple is trying to achieve makes perfectly sense in perspective of marketing. Designing it's own chips makes Apple less dependent on companies like Intel. But marketing also includes the amount of software that will support the Mac hardware that runs mainly on Apple's own chips rather then intel. Meaning, it will takes yeaaaaaaars, probably way beyond the 2020 date that Bloomberg names as the earliest launch window for a first Intel-free Mac...
Why? Simply because Apple will never, and this is 100% true, launch a whole new MacPro that doesn't even support Adobe's desktop programs as long this company, Adobe, didn't had the time to re-write all of its software in time. Seriously, all these emotions about 'future Mac's that doesn't support Windows any longer' are not even close to the reality in the near future. IF Apple would come with a whole new MacPro machine that doesn't support the desktop versions of Adobe or Windows using bootcamp all together then Apple will loos the remaining "pro" users all together in a day. I can't, and will not, believe Apple will go that path.
Why? Simply because Apple will never, and this is 100% true, launch a whole new MacPro that doesn't even support Adobe's desktop programs as long this company, Adobe, didn't had the time to re-write all of its software in time. Seriously, all these emotions about 'future Mac's that doesn't support Windows any longer' are not even close to the reality in the near future. IF Apple would come with a whole new MacPro machine that doesn't support the desktop versions of Adobe or Windows using bootcamp all together then Apple will loos the remaining "pro" users all together in a day. I can't, and will not, believe Apple will go that path.
As someone who works across platforms and for whom the "integration" of the Mac "ecosystem" is irrelevant, I can't see an upside of this for me. Anything that reduces software availability, including of specialist software I use as well as big commercial packages (who will be more likely to port), is a clear negative. So as someone who uses the Mac as a work tool the risks here are much clearer than the benefits.
Will see. If they go somewhere that is unhelpful to me I can go elsewhere.
Will see. If they go somewhere that is unhelpful to me I can go elsewhere.
Well you will still be able to run Virtual Machines, probably with better performance than Intel chips.
Costs will be negligible. The transition will probably take 3 years on the consumer side and 5-6 years on the professional workstation side and most of the software updates will be free just like they were when we moved from PPC to Intel. Software pricing is very different these days too. People who use subscription software won't be paying an extra penny.
How do you come to that conclusion? To run virtual machines on an ARM processor would require emulation rather than virtualisation, which comes with a huge performance overhead so you'd probably need ARM chips 5 - 10 times faster than the Intel ones to achieve faster performance via emulation than native. When Apple made the transition to Intel, Rosetta had around a 30% performance hit over the native PowerPC machines and that was on processors that were far more powerful than the ones they were emulating
Nobody will be really surprised if this happens.
It will be a lot of fun to follow but I will be enjoying from the sidelines. Windows is really a back-seat car, I mostly drive alone but I can't live with a two seater. There isn't really anything to be enjoyed by consumers, just Apple that will be able to increase its margins. When they moved from PPC to intel the deal was, no more dead end slow processors, welcome top of the line industry standard with industry standard software if that is your wish. A pretty good deal if you ask me. And now?
Control? For what? They don't even use all that intel makes available, entire intel generations come an go without Apple using a single chip in some product lines. Control will probably be used to fuel the thinness obsession.
Differentiation? That is MacOS role (and I'm not even counting Apples signature hardware design). It is all the differentiation anyone needs. Stop hackintosh? That is a blip in the screen. Apple was always intelligent about that.
What do consumers (as in who buys, not differentiating use) have to gain? As I said it will be fun just as it is fun watching tornados on TV, not being there.
It will be a lot of fun to follow but I will be enjoying from the sidelines. Windows is really a back-seat car, I mostly drive alone but I can't live with a two seater. There isn't really anything to be enjoyed by consumers, just Apple that will be able to increase its margins. When they moved from PPC to intel the deal was, no more dead end slow processors, welcome top of the line industry standard with industry standard software if that is your wish. A pretty good deal if you ask me. And now?
Control? For what? They don't even use all that intel makes available, entire intel generations come an go without Apple using a single chip in some product lines. Control will probably be used to fuel the thinness obsession.
Differentiation? That is MacOS role (and I'm not even counting Apples signature hardware design). It is all the differentiation anyone needs. Stop hackintosh? That is a blip in the screen. Apple was always intelligent about that.
What do consumers (as in who buys, not differentiating use) have to gain? As I said it will be fun just as it is fun watching tornados on TV, not being there.