That's what I was referring to when I said "co-processor" - similar(ish) to how the iPads (and iPhones?) now have several low-power/high efficiency cores, and a couple of high-power/high performance cores.
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.
ARM based MacBook Pro discussion [merged]
- Thread starter Omega Mac
- Start date
-
- Tags
- arm intel macbook pro macos sierra
- Sort by reaction score
Would work great with cross platform apps and would help to use it for idling and other features within MacOS
I honestly don't know if that scenario is even practical though. It almost certainly would mean two physical chips, but that might also mean things like the Mac Pro could have a socketed off-the-shelf Xeon that's upgradeable while getting the benefit of the efficient Arm cores for the lightweight stuff.
Maybe this is all the stuff of fantasies.
I'm sure I can't be the only nut on MacRumors to imagine what my next laptop may be. I'll likely be ready for an upgrade in two years, and I imagine I'll buy one of two laptops - either the redesigned 2020/2021 13" MacBook Pro with slimmer bezels and the new scissor switch keyboard or the 20XX MacBook/Air/Pro with an ARM processor (but likely the 2nd or 3rd generation after all of the kinks have been worked out).
Regarding an ARM Mac, do you foresee it to have enough power akin to the current 2019 13" MacBook Pro, and able to handle medium-to-heavyish tasks? Or do you imagine Apple's ARM laptop will be more suited for light tasks (email and web surfing)? How much power are you expecting from an ARM MacBook?
Regarding an ARM Mac, do you foresee it to have enough power akin to the current 2019 13" MacBook Pro, and able to handle medium-to-heavyish tasks? Or do you imagine Apple's ARM laptop will be more suited for light tasks (email and web surfing)? How much power are you expecting from an ARM MacBook?
Last edited:
Hah, man, it's been a long day. Updated my post. Thanks, Howard2k!Are you mixing ARM and AMD?
I think it will have the horsepower. Look at the A series. All the same, application compatibility would be my concern. I don’t think ARM would debut in the Pro though, so by the time it makes it to the Pro I anticipate the bugs will be mostly worked out.
it probably will Happen, not the full conversion but perhaps a hybrid integration (like the T2 Chip) they may use an arm chip in order to maximize let's say light work for optimized power savings, or maybe introduce it on base models and keep intel (or Ryzen) for high-end.
I was surprised when they discontinued the MacBook. I thought that was a sure fire guarantee for ARM. Perhaps they’ll reintroduce it when they’re ready.
I was surprised as well, but honestly the 2016-2018 MacBook line ups was tad confusing for $ >performance ratio, kinda made sense to knock it off in order to reinstate the MacBook Air as a proper product in their lineup.
if they do bring ARM chip into their Macs, this will make iPads look like off-spring child awkwardly ?
Probably MBA, because they need that one device that can allow easy transition gap of iPad > MacBook line ups.
the way I see it, Apple can't just provide arm on a high end MacBook Pro since ARM architecture conjunctions of processing utilization is completely different because they aren't at the same level in development.
better way to output this out in context; the A12X chip (iPad Pro) performance equivalencies would match a barely a Y dual core intel processor and a S series at this point of time. to translate intel's catalog:
C – Desktop processor high performance graphics
H – High performance graphics
K – Unlocked
M – Mobile
Q – Quad-core
R – Desktop processor base
S – Performance-optimized lifestyle
T – Power-optimized lifestyle
U – Ultra-low power
X – Extreme edition
Y – Extremely low power
for the proper performance growth, the arm based need to catch intel in certain categories to perform the following combination of intel's chips just to match in benchmark:
Y, U, S and T
(tbh that's a lot of catch up lol just to represent in theory; one matching intel' family tree as stated above list.) because right now that a12x is doing half of all these together.
In order to justify the ARM architecture on a high end computing device such as MacBook Pro without cannibalizing mobile devices such as iPads or sabotaging foundations and investments of corporate software (graphic design) and of course the sub market system of MacOS and iOS, they need to execute this accordingly with the proper development of maturity of the technology.
A MacBook Air uses a Y processor, a dual core performing a very low Watts at idle and high peak. at this point, in benchmark an iPad Pro may show higher number but this obviously means at the arms architecture benchmark and should not be compared fully to the intel's benchmark. if you add 2x A13 in a MacBook Air, it will actually perform similar if not same has the current Y and almost a U processor on the MacBook Air/pro
remember, we talking about Octa core (8 cores) that needs to be doubled in order to execute the same order of execution as an U processor (dual core for the sake of example). so this should give you an idea how behind the arms are, in theory it sounds amazing and the potential future replacement for mobile laptops chips but not any time soon. intel have a long based dominance in chip development for years and years, in order to have something that can replace it, will and should take the same amount of time; with less expense though
Last edited:
well, the 12” MB and MBA would be the obvious first candidates - and they wouldn’t be intended for medium-to-heavy is how tasks.
Looking at the way iPadOS is being pitched, I think possibility is receding and they’re pushing iPad Pro as the solution for the 12” MacBook market.
Overall, though, if it happens, it will probably depend on your application.
Unless Apple just dump MacOS on an iPad Pro (might make sense as a prototype for developers) An ARM Mac will probably come with a previously unseen A series processor, so nobody knows the details. To speculate:
ARM isn’t night and day faster than Intel, but it’s more power efficient and compact, so the likely advantage will be cramming more cores, specialised hardware accelerators, bigger iGPUs etc. into the same space and thermal profile. Apps that exploit such things should benefit from that, ones that don’t might be a bit disappointing.
Modern Apps written for MacOS that use the official MacOS frameworks for graphics, acceleration etc. are likely to perform well, and will also be quickly released in ARM native form. Apps that need x86 emulation/translation will be slower, but not so much if they still do their heavy lifting via Apple frameworks.
The concern would be if you use large, cross platform apps, esp. with loads of third party plugins.
That said, there’s going to be a big cull of “abandonware” over the next year, anyway, as Catalina drops 32 bit support. In many cases, porting to ARM will be less work than supporting 64 bit.
If you really need to run VMs of x86 machines (Windows, MacOS for existing Macs,specific x86 Linux) then ARM is likely to be a nonstarter. Maybe you won’t need to do that a few years down the line: ARM Linux is already pretty credible - and if you’re doing web development or suchlike the CPU architecture is largely irrelevant. Maybe ARM Windows will take off, but the sort of legacy apps that some people use Windows VMs for are probably the ones that won’t run on ARM...
Very, very helpful. From all of this, given my computing needs and my distain for being hurt as an early adopter, sounds like I should plan for a second generation 14” MBP with the smaller bezels and the new keyboard (after the kinks are worked out in the first year, and we vet the new keyboard design).
The only thing Windows' new ARM version seems to be lacking is support for 64 bit apps, emulation will cover any 32 bit windows application already - I am imagining a lot of the legacy Windows software is probably 32 bit by nature, being older/ no longer supported. I would think computers like the SP X will be more aimed at the office productivity type use case though, so browser and word processor will be 95% of what's needed there. Now it's on an attractive new flagship design rather than a lower cost 'gimped' looking version I think that might make people more open to trying it, particularly if they hear that it can emulate at least some x86 Windows software, with 64 bit support intended to follow eventually. Maybe if it takes off it will reach critical mass this time.
Apple high-end ARM chip
Slightly bigger than the 8CX cores
A large in chip AI accelerator
8 Cores with no SMT
Comparable to i7/R7
First product: MacBook Air
Under consideration: MacBook Pro and iMac
Coming to the market in 2020.
Well, I'm not sure if this info is legit since Apple never leaked their own chips before. But it's a fact that they are preparing to ditch Intel after 2020. This ARM-based Mac rumored since 2011 btw.
Microsoft made ARM-based computers for a long time and yet they are totally failures due to lack of compatibility and performance eve in 2019. So I wonder if Apple can manage to make a proper ARM-based Mac in 2020?
ARM CPU on either MPB or iMac is very interesting if it's true. But yeah, the performance won't be that good initially.
Any thoughts?
That sort of chip spec sounds superficially plausible - and the sort of thing that Apple would need to replace Intel in the MacBook Pro (you wouldn't stick an A13 in a higher-end MacBook Pro although you might get away with it in an Air). Sounds mainly speculation though - just riffing in the Intel leak from months ago that they expected Apple to start moving to ARM in 2020, which could just have been Intel worst-casing as part of their "due dilligence".
Totally disagree on this "mobile/laptop bridge" thing being behind the shift to ARM, though. You do *not* need to shift to ARM to interoperate with mobile, because mobile applications - and other non-legacy apps - don't generally give a wet slap about what CPU architecture they're running on.
Android apps are predominantly shipped as bytecode for the Dalvik virtual machine, "modern" MS Windows apps are increasingly compiled to bytecode for the MS Common Language Runtime (which has been a thing since .Net came out) and while I believe iOS Apps are still shipped as ARM binaries, for anything capable of getting accepted into the App store that is a tick-box in Xcode - in fact, iOS apps are tested in Xcode by compiling them to x86 and running them in a "iOS x86" sandbox. That's not counting the increasing amount of stuff written in "script" languages like Javascript or Python. When the source code is identical, maintaining multiple binaries is no biggie, especially when everything has to be distributed by an online App store.
So, at least for "modern" apps, the days of letting your applications choose your processor architecture are coming to an end and the only people who need to worry about whether the target has an ARM, x86 or RISC-V (esp. when everything is 64-bit and little-endian) are the systems programmers who write compilers, VMs, drivers and runtime libraries.
The issue keeping x86 alive is established "desktop" (as opposed to mobile) apps that still contain a lot of legacy code or - for some other reason - implement their own drivers and/or eschew the official OS frameworks. (...and for some users that includes x86 Windows virtualisation - which will probably be the biggest issue for ARM Macs).
Microsoft has a huge albatross around its neck in the form of legacy applications, including a vast amount of ancient in-house written software in the corporate market - if MacOS is nicer than Windows that's a big part of the reason. They've faced a massive uphill struggle to get rid of Windows XP - and many of the faults in that (e.g. everything running with Admin rights) were because it had to run apps written for DOS and Win3.1 and Win9x ...and really old x86 code, even in C, can be really non-portable, since early DOS/Windows wasn't properly 32 bit and those OSs didn't have such extensive hardware-independnet frameworks.
Apple don't really have the corporate drag-anchor and can be much more flexible - has already completely switched OS once (Classic MacOS to son-of-NextStep-aka-OS X) and processor twice (68k to PPC, PPC to Intel - thrice if you want to count 6502 to 68k). Over the last year, they've thrown 32-bit support under a bus and declared OpenGL obsolescent (I'm sure MS dream of being able to do that). Dropping 32 bit in itself has probably cleared out a lot of the "dead wood" that would hamper an ARM switch.
As long as its running "modern" apps which don't need x86 emulation - a MacBook Air with that chip in is likely to smoke the x86 Air. Even a Rosetta-like 'just in time' x86-to-ARM translator might be pretty slick with software that makes good use of Metal, Core Audio, Accelerator kit etc. rather than hard-coded Intel SIMD etc. calls. Its going to be waiting for things like Adobe CS and all your third-party graphics/video/audio plug-ins that will be an issue (but that's not really what people buy an Air for...)
Totally disagree on this "mobile/laptop bridge" thing being behind the shift to ARM, though. You do *not* need to shift to ARM to interoperate with mobile, because mobile applications - and other non-legacy apps - don't generally give a wet slap about what CPU architecture they're running on.
Android apps are predominantly shipped as bytecode for the Dalvik virtual machine, "modern" MS Windows apps are increasingly compiled to bytecode for the MS Common Language Runtime (which has been a thing since .Net came out) and while I believe iOS Apps are still shipped as ARM binaries, for anything capable of getting accepted into the App store that is a tick-box in Xcode - in fact, iOS apps are tested in Xcode by compiling them to x86 and running them in a "iOS x86" sandbox. That's not counting the increasing amount of stuff written in "script" languages like Javascript or Python. When the source code is identical, maintaining multiple binaries is no biggie, especially when everything has to be distributed by an online App store.
So, at least for "modern" apps, the days of letting your applications choose your processor architecture are coming to an end and the only people who need to worry about whether the target has an ARM, x86 or RISC-V (esp. when everything is 64-bit and little-endian) are the systems programmers who write compilers, VMs, drivers and runtime libraries.
The issue keeping x86 alive is established "desktop" (as opposed to mobile) apps that still contain a lot of legacy code or - for some other reason - implement their own drivers and/or eschew the official OS frameworks. (...and for some users that includes x86 Windows virtualisation - which will probably be the biggest issue for ARM Macs).
Microsoft has a huge albatross around its neck in the form of legacy applications, including a vast amount of ancient in-house written software in the corporate market - if MacOS is nicer than Windows that's a big part of the reason. They've faced a massive uphill struggle to get rid of Windows XP - and many of the faults in that (e.g. everything running with Admin rights) were because it had to run apps written for DOS and Win3.1 and Win9x ...and really old x86 code, even in C, can be really non-portable, since early DOS/Windows wasn't properly 32 bit and those OSs didn't have such extensive hardware-independnet frameworks.
Apple don't really have the corporate drag-anchor and can be much more flexible - has already completely switched OS once (Classic MacOS to son-of-NextStep-aka-OS X) and processor twice (68k to PPC, PPC to Intel - thrice if you want to count 6502 to 68k). Over the last year, they've thrown 32-bit support under a bus and declared OpenGL obsolescent (I'm sure MS dream of being able to do that). Dropping 32 bit in itself has probably cleared out a lot of the "dead wood" that would hamper an ARM switch.
As long as its running "modern" apps which don't need x86 emulation - a MacBook Air with that chip in is likely to smoke the x86 Air. Even a Rosetta-like 'just in time' x86-to-ARM translator might be pretty slick with software that makes good use of Metal, Core Audio, Accelerator kit etc. rather than hard-coded Intel SIMD etc. calls. Its going to be waiting for things like Adobe CS and all your third-party graphics/video/audio plug-ins that will be an issue (but that's not really what people buy an Air for...)
Well, it's a good news since Adobe made the firs ARM-based App for Photoshop which might work on ARM-based Mac!
Really? It’s a fact? Facts are provable with evidence. So, how about some?
Intel Expecting Apple to Transition to Custom ARM-Based Chips Starting in 2020
Apple is planning to ditch Intel and transition to Mac chips starting in 2020, based on multiple rumors we've heard in the past from Bloomberg....
www.macrumors.com
Apple's ARM-based Macs are probably arriving next year | TheINQUIRER
It'll probably cost an ARM and a leg,Hardware ,ARM,Apple,MacBook
www.theinquirer.net
Report: Apple to begin switch from Intel chips in Macs by 2020 - 9to5Mac
Apple has been rumored to be working on creating custom chips for its Macs for quite some time. Now, Bloomberg...
9to5mac.com
Apple said planning move from Intel to Mac chips from 2020
Moving to its own chips inside Macs would let Apple release new models on its own timelines, instead of relying on Intel's processor roadmap
www.livemint.com
Apple is planning to design its own chips for the Mac from 2020
It’s been long rumoured, but now it seems it really could happen: Apple really is planning to dump Intel in favour of its own processors for at least
The catalyst project is already on. iPadOS can be used on macOS
Rumors are not facts. There is nothing in this that we haven't seen/read yet.Intel Expecting Apple to Transition to Custom ARM-Based Chips Starting in 2020
Apple is planning to ditch Intel and transition to Mac chips starting in 2020, based on multiple rumors we've heard in the past from Bloomberg....
Apple's ARM-based Macs are probably arriving next year | TheINQUIRER
It'll probably cost an ARM and a leg,Hardware ,ARM,Apple,MacBook
Report: Apple to begin switch from Intel chips in Macs by 2020 - 9to5Mac
Apple has been rumored to be working on creating custom chips for its Macs for quite some time. Now, Bloomberg...
Apple said planning move from Intel to Mac chips from 2020
Moving to its own chips inside Macs would let Apple release new models on its own timelines, instead of relying on Intel's processor roadmap
Apple is planning to design its own chips for the Mac from 2020
It’s been long rumoured, but now it seems it really could happen: Apple really is planning to dump Intel in favour of its own processors for at leastwww.pocket-lint.com
I never feared the transition to Arm because it will only happen when it comes without a performance drop, and that would in part be possible by creating processors designed for the platform, free of the mobile power restrictions. That is what the video suggests.
The 32-bit clear out has made many more Macs apps incompatible for me than leaving x86 would. I expect a lot to be put in place to make it easier to port iOS apps to Arm macOS to make up the shortfall, although if programmed correctly that should just be a matter of control interaction.
The 32-bit clear out has made many more Macs apps incompatible for me than leaving x86 would. I expect a lot to be put in place to make it easier to port iOS apps to Arm macOS to make up the shortfall, although if programmed correctly that should just be a matter of control interaction.