Intel Expecting Apple to Transition to Custom ARM-Based Chips Starting in 2020

[jdiamond]

Well if this is true then I dearly hope Apple can pull it off.

I am old and ugly enough to remember the transition to Intel from PPC; now that was a great move. No question.
But, Apple's base was far smaller… and even then it took a lot of heartbreak.

Not really looking forward to the "new" Rosetta as companies suck teeth while deciding if they are porting it all over.

Of course there was BootCamp — which was great — but this time round?

Anyway… deep breaths and let's see if Apple come up with some magic.

Kind'uv' like Apple disavowing 32-bit apps on the *same* hardware means I lose ~99% of my favorite Mac Software? Apple has found far easier ways to kill entire software bases than merely switching the processor. One related wildcard is how quickly Microsoft will switch to ARM - then the only issue is how fast will ARM-based Linux evolve into maturity? My feeling is for *old* legacy x86 apps, they can probably be emulated fast enough, while recent x86 apps may largely get ported. When Microsoft switched to Vista and made a lot of Windows software obsolete, I was sure there'd be companies dedicated to hosting cloud based Windows XP apps for customers - but that didn't really happen. So it seems customers just suck it up and take it when they lose their software forever.

 

[maflynn]

I remember the transition from PowerPC to Intel. I think it required a rewrite of major programs to be compatible.

Will a transition to ARM require new software?

Yes on both counts but your latter question may be less of a hassle. Apple is moving to letting iOS apps run on macOS, so in theory once that is set up, you'll just be running existing iOS apps.

 

[Reindeer_Games]

Depends on the definition of everything ...

It was statement of hyperbole based on the users reply and assumption that and ARM based MacBook-will "destroy" their current iPad Pro-so therefor it is Intel's fault.

The full OS of macOS is most likely why they own it-not how powerful it may be.

 

[Defthand]

It doesn't make sense that Apple would do this to accommodate iOS apps on the desktop—a platform that consumers have abandoned, including Apple in spirit. I suspect the opposite is true. Apple is steering MacOS devs to develop laptop–like apps for the iPad. Indeed, I would not be surprised if Apple replaces the OS on their laptops with iOS eventually.

The iMac will be abandoned. The only desktop Mac will be the modular model for occupational needs. I'm fine with this possibility as long as Apple doesn't compromise MacOS by making it fit into an iOS–verse.

 

[Scooz]

Most of the best engineers I worked with, from AMD, DEC, etc., are at Apple. They have the folks to do it. And Intel hasn’t shown that they can advance things anymore. If Intel today were Intel of a decade ago, I’d say Apple shouldn’t bother. Apple has already gotten past most of the technical hurdles to do it. What remains is pretty easy.

I would second that conclusion, technologically. From a market perspective one has to ask if that‘s what the customer wants. Intel
compatibility is a major sales argument, even for AMD.

And even if the usage number might not show everybody using Windows or Linux on a Mac, the idea of being able to do
so might be an important feature.

I am not sure that I would want a machine that would not allow me to start a proven VM for development. Though a good Intel emulation or native ARM VMs and compiles might also have their value, if anybody will port the whole VM engine to ARM, that is. Which is my main concern here. Many things just run on my Mac right now because they don‘t need to be recompiled or emulated.

Tricky to judge without knowing more about their actual reasoning.

 

[nt5672]

What better way to get everyone to buy a new computer than switch supported architectures? Imagine 2 years after full rollout there will be no updates to Mac OS on intel. None. This will supposedly be a big boost the bottom line. Perfect Cook strategy.

Now to make macOS run architecture independent (both intel and ARM) that would be brilliant and future proof, but not possible with Apple's current management. It's not in the best interested of cash generation for this quarter.

 

[ObsidianKt]

What does this mean for boot camp? Does this mean no more windows on Mac? If so, seems bad for any Mac gamers.

 

[maflynn]

What does this mean for boot camp? Does this mean no more windows on Mac? If so, seems bad for any Mac gamers.

Yes, basically. Windows runs on Intel only and while MS has a varient for ARM, its not for Apple's AX chip and MS may not be willing to port it to Ax either.

 

[timber]

My PhD dissertation was on CPU caches. I was a designer on the original Opteron, UltraSparc V, Exponential x704, etc. Designing A12 is not that different from designing a multi core x86.

I owned several Opterons, still have a working 144 at home now at just its standard clock. What a great work.

 

[Idec50]

This sounds like an unmitigated disaster. There is a ton of software that will never be ported to ARM. Will it also use a browser that can’t run Javascript or just not have a browser? This rumor has me questioning whether I am wasting my time reading this blog, and I just created an account to post here. As others stated, why would a developer purchase a Mac if this rumor turns out to be true? To develop for Apple products on a almost completely closed system. This would likely kill macOS for all other professional use, including mine. If that is accurate, it will be a death spiral for macOS users who depend on any third-party tools, particularly those that modify functionality in any way.

 

[deputy_doofy]

Whats so impressive about an i7? I bought a brand new iMac last summer with one and it feels more sluggish than my 2013 and has more issues. In the past, a 3+ year gap between computers was a wayyy bigger upgrade

macOS benefits *most* from an SSD. If you have an HDD or even a fusion drive, your Mac will feel sluggish, even with a great CPU and lots of RAM.

 

[cube]

Don't you still remember how painful it was to compete against Intel back then? Moving to Intel was a good move because IBM didn't improve chips fast enough. - At least from a designer perspective, it was great. If Apple can develop faster processors and GPU I am all for it. Right now its painful not to have NVIDA GPU for rendering power.

I don't care. There was also Motorola/Freescale.

And it is good Apple fights for AMD GPUs.

 

[fathergll]

My only concern with this is will this change still allow us to run Boot Camp and Parallels Desktop on ARM based Macs?

Bingo. It won't.

Also say goodbye to the hackintosh.

 

[pmore]

“Even Microsoft..”

The fact that Microsoft failed at something means that Apple must also fail? Is this the “Zune Hypothesis” or the “Windows Mobile Theory” or the “Vista Corollary?”

Macs might succeed since no one takes it seriously anymore. Pros use Windows.

 

[EntropyQ3]

This sounds like an unmitigated disaster. There is a ton of software that will never be ported to ARM. Will it also use a browser that can’t run Javascript or just not have a browser? This rumor has me questioning whether I am wasting my time reading this blog, and I just created an account to post here. As others stated, why would a developer purchase a Mac if this rumor turns out to be true? To develop for Apple products on a almost completely closed system. This would likely kill macOS for all other professional use, including mine. If that is accurate, it will be a death spiral for macOS users who depend on any third-party tools, particularly those that modify functionality in any way.

This is a bizarre post.
Most current apps will not need to do anything to run natively, provided they are distributed as bitcode, which will be translated to the underlying ISA.
I don't see why browsers wouldn't run Javascript, iOS devices do. This question is just weird. You might as well question if ARM-macs will support colour displays.
Regarding developers, it won't change much of anything. They will target whatever API they need to target. There is no difference from today, except that they would run an emulated x86 environment if they want to test run their apps running windows on their Mac. This is no big deal, for most purposes this will be undetectable even to developers.

What professional use do you have now that cannot be run on an ARM Mac? I'm searching my mind, and coming up with precious little where this is even conceivable.

Boot Camp gaming will likely suffer but honestly, little else.

 

[smaffei]

Well if this is true then I dearly hope Apple can pull it off.

I am old and ugly enough to remember the transition to Intel from PPC; now that was a great move. No question.

I am old enough to remember 68K→PPC. That seamless 68K emulator was pretty awesome. Ah, those were the days when Apple pulled off some astounding stuff (and it was so good, no one noticed).

 

[EntropyQ3]

Yes, basically. Windows runs on Intel only and while MS has a varient for ARM, its not for Apple's AX chip and MS may not be willing to port it to Ax either.

Gah, this is just plain wrong, weren't you guys around during the PPC days? Don't you know we ran Windows on those machines as well?
Windows will presumably run in a Parallels like environment, using an emulated x86 CPU. It will be slower doing some things, but many calls will typically be mapped to use ARM natively.
Considering the speed at which the a12x already executes, any future Apple CPUs that would benefit from both better lithographic process and higher power draw allowance, is likely to emulate x86 very nicely indeed.

 

[theluggage]

This will make iOS development more difficult, since we need to support additional hardware configurations.

You're assuming that Apple is going to make Marzipan compulsory for App certification. If they could get away with that, it would certainly boost the number of apps available for MacOS but would, as you say, be a courageous decision.

Otherwise, though, the option of Marzipan makes it far, far more economical to choose to support Mac - c.f. the alternative of maintaining two separate apps. You can lead a horse to water, but you can't make it drink - but it can't drink if you don't lead it to water!

 

[maflynn]

Gah, this is just plain wrong, weren't you guys around during the PPC days? Don't you know we ran Windows on those machines as well?

You mean virtual PC, yes I remember that so well. Performance was horrible, and it was barely useable. Sure an emulation app could be developed but it will be horrible. Don't mix up emulation and virtualization. VirtualPC emulated an intel CPU to run windows, and it was dog slow

 

[smaffei]

Gah, this is just plain wrong, weren't you guys around during the PPC days? Don't you know we ran Windows on those machines as well?
Windows will presumably run in a Parallels like environment, using an emulated x86 CPU. It will be slower doing some things, but many calls will typically be mapped to use ARM natively.
Considering the speed at which the a12x already executes, any future Apple CPUs that would benefit from both better lithographic process and higher power draw allowance, is likely to emulate x86 very nicely indeed.

Are you kidding me? "Virtual PC" was dog on the Mac in those days! Even Solitaire dragged.

 

[deconstruct60]

...
The number of people who need multi-boot is tiny. A fraction of the 5million or so who buy macs in A quarter. Compared with the 100 million iPhone purchasers who apple hopes to mine as future mac purchasers. They just don’t care about you.

it is a bit too myopic just to stop the "its too small" observations at just the multiple boot percentage fraction.

Same point is largely true of macOS on a highly custom ARM for just a relatively small single digital millions ( versus two orders higher 100M range ). Apple could possibly toss "hand me down" A__X processors over the wall for a MacBook and perhaps a MacBook Air over time, but Mac product line up isn't going to pragmatically be viable with just a single processor design ( with maybe 2-4 cores tweaked on/off ).

So once on the subset of the Mac product range that has a run rate of 100k per quarter whats the point of some custom team in additional the strategically critical iOS targeted team? The relatively low volume custom ARM has the problem being pointed out. For example, if there are 3M per Quarter low end laptop users and 3k/Q high end laptop users then why switch from x86 ( Intel or AMD) as a supplier on the latter. The desktops are more skewed toward the higher end where "hand me down" 100 iPhone processors aren't going to be viable.

Apple doesn't have to care about the folks who want what is in the broad range that the Mac line up has historically covered. The Mac line up isn't suppose to just solely be about the thinnest, lightest, laptop Apple has ever sold. Apple ARM efforts have decreasing limited impact once move away from that relatively narrow corner case.

Apple tossed the A10X into the AppleTV probably hoping to get some volume momentum out of that. That probably hasn't worked out as planned. Tossing something of that class into the MacBook probably would work incrementally better. But if Apple is scraping up volume for the iPad Pro chip ... how is the even more performance fragmented Mac line up going to pose even less of a "too small don't care" problem?

Coupling a sizable chunk of the Mac product line up easily gets over much of these "too small don't care" issues on finding enough volume to cover costs for yet another full silicon pipeline set of resources. In short, share R&D costs with a large pool of players. Intel stumbling on their face pragmatically doesn't mean as much if AMD isn't. Apple having two competent suppliers isn't a critical business problem. If Intel was the only 'possible' vendor then perhaps, but they aren't. Apple could go down to zero Intel chips in 2021 in macs and still not being buying from Intel or "themselves".

 

[jwdsail]

Noooo...Desktops should stay intel, portables can go the way of the overpriced lappad...

Ok, I'm sorry, but did you even read my post before replying?

I preposed two Mac lines, Consumer, and Pro, that would fall neatly into line with what Apple's been doing recently. Each with both desktop and laptop models? Each with both head-less and AIO desktop models.

Consumer/A-Series-powered:

1: Mac mini
2: iMac 21" 27"
3: Macbook/Air 11, 13, 15

Pro/Intel-powered:

1: Mac Pro
2: iMac Pro 27" ?
3: MacBook Pro 13, 15, 16-17

Hell, maybe this would allow a thicker/more upgradable MBP be introduced, Ram and SSD not soldered in place? Now THAT would be INSANELY GREAT! Just imagine! With a KB that didn't suck like a Hoover to type on for anything longer than a tweet? Shut up and take my money!

 

[applebro]

So macs will be bigger iPad pros with built on keyboards lmao

Since I do more than the typical photo editing Facebook surfer I’ll provably have to switch back to windows. Oh well their surface book doesn’t look that bad

 

[BlandUsername]

I think you are right. The reason the last transition was embraced was because it made the Mac (sort of) standardized with the existing world of PCs. It made the transition easier, it made emulation easier, etc. Apple made the transition for performance and cost reasons, but Apple users accepted it for the standardization reasons. I sure Cook remembers the cost and performance benefits. What Cook never understood was the standardization issues and why it was accepted by users.

This transition makes Apple Macs non-standard again. So unless Windows, Linux, and Unix are released for the Apple ARM hardware, this blows and Cook has no idea. He thinks Apple is Apple and can do anything. Sure a few people will buy Apple Macs with ARM processors and never know the difference. But most of us pro users etc. will not. Of course, this probably why Apple has been pushing Pro users away in recent years.

This just means the secondary market for used Mac will boom and at the very last second the Hackintosh crowd will explode. I have to think nothing would slow them down once the last build supported on Intel gets announced.
[doublepost=1550860016][/doublepost]

Ok, I'm sorry, but did you even read my post before replying?

I preposed two Mac lines, Consumer, and Pro, that would fall neatly into line with what Apple's been doing recently. Each with both desktop and laptop models? Each with both head-less and AIO desktop models.

Consumer/A-Series-powered:

1: Mac mini
2: iMac 21" 27"
3: Macbook/Air 11, 13, 15

Pro/Intel-powered:

1: Mac Pro
2: iMac Pro 27" ?
3: MacBook Pro 13, 15, 16-17

Hell, maybe this would allow a thicker/more upgradable MBP be introduced, Ram and SSD not soldered in place? Now THAT would be INSANELY GREAT! Just imagine! With a KB that didn't suck like a Hoover to type on for anything longer than a tweet? Shut up and take my money!

I did. I invest in Multiple Mac Mini's , I use them for work, I have multiple sessions to remote sites open all the time, The desktop world should be left alone ...period, They want to move users to a more cloud oriented POV, do it on the the portable devices only.

 

[Spock]

If Apple announces this new iOS update for iPads that makes the iPad more like a desktop computer we will be able to see how close we are to a merging of iOS and MacOS. That should scare people more than ditching intel.