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

[Unregistered 4U]

"Advance warning" for me would mean, begin the transition back to Windows or Linux-based Intel systems.

So, that’s actually was I was thinking initially. Especially since you’re already comfortable in those OS’s ANY one of them would be a good base to do all your virtualization from.

At this point, the handwriting is on the wall, and eventually, there won’t be an option to have an Apple ecosystem running on an Intel platform. As I’ve written earlier, Intel likely thinks Apple will transition because they expect to already have received a large contract from Apple from the chips they need next year, and they either haven’t seen that order OR the order is way smaller than Apple usually purchases from them. This would indicate a replacement by some other x86 solution provider (like AMD) or ARM. And, with what they’re doing with Mac Catalyst, the smart money is on some A series variant processor.

Start easing yourself out of the Apple ecosystem, now. So that, when the official news comes, you’ll already be well on your way to the portable virtualization platform that will do what you need for now and grow with you into the future.

 

[Unregistered 4U]

double post

 

[Ledgem]

At this point, the handwriting is on the wall, and eventually, there won’t be an option to have an Apple ecosystem running on an Intel platform. As I’ve written earlier, Intel likely thinks Apple will transition because they expect to already have received a large contract from Apple from the chips they need next year, and they either haven’t seen that order OR the order is way smaller than Apple usually purchases from them.

Unless there have been new articles, the one article about Intel speculating that Apple would move away from them wasn't sourced from high-level management, and it wasn't based on any obvious data. It was someone at Intel speculating the same way that we are here. It only made the news because they worked at Intel.

Is it true that Apple didn't place a large contract order from Intel? If not, perhaps the downturn in Mac sales could explain it (information based on the recent earnings call report), or the fact that Intel doesn't yet have a chip that interests Apple for an upgrade of some of their systems.

I'd also counter by saying that usually when there's some change you'd hear rumors from other sources. There would be benchmarks from test systems showing up, and/or you'd hear rumors from fabrication labs and assembly partners about the new Macs. But so far we haven't heard a single thing on that front: it has purely been speculation on this forum, with the only off-forum source being that one weak article quoting some random, anonymous Intel employee.

Far from the writing being on the wall, there's really no writing to go off of at all. I won't be surprised if it really does happen, nor would I place a bet against it, but I also don't expect it to happen. I'm not going to make any changes at this time, but I will continue to decry the idea.

 

[Samuelsan2001]

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

Windows can already run natively on arm chips you wouldn’t need parallels and bootcamp should be a no brainer to pull off. Remember many Microsoft titles not only already run on arm but already run on iOS and utilise metal. Larger arm chips with cooling could be a huge bonus for the Mac line. I’d imagine the MacBook Air will be first though with a passively cooled upgraded A13X solution.

 

[Nugget]

Windows can already run natively on arm chips you wouldn’t need parallels and bootcamp should be a no brainer to pull off.

This couldn't be farther from the truth.

Nobody runs BootCamp or virtualization because they want to run the Windows operating system. They run BootCamp or virtualization because they want to run Windows Applications. While it may be true that Microsoft have built Windows for ARM (but you can't actually buy it, which is its own problem) and there really isn't good application support for it. Even if you could buy a copy of Windows for ARM (which you can't) there's not much you can do with it if the applications aren't there.

If Macs switch from Intel to ARM it's effectively the end of BootCamp and useful virtualization (this includes Docker Desktop). Apple may decide that they're willing to take that risk, but it's definitely going to deter users who rely on those options for their workflows.

As you say, it might make sense for low-end products like the MacBook Air but it would be a huge setback to products like the iMac or MacBook Pro.

 

[Unregistered 4U]

but I will continue to decry the idea.

That is absolutely an option. And it’s not like decrying the idea is going to hurt anyone. Plus, you get the added benefit of continuing to stay in the ecosystem for at least another 6-8 months. Win-Win

 

[Unregistered 4U]

Windows can already run natively on arm chips

Regarding any OS other than macOS executing on an ARM Mac, I’d just say “Don’t expect it”. Go to Apple’s website right now and in the search bar, do a search for both “emulation” and “virtualization”. You won’t find any search results under Explore or Support for the emulation of other OS’s. Under Accessories you WILL find Parallels. The search for virtualization doesn’t show anything related to the virtualization of other OS’s and you won’t even find Parallels under Accessories!

These searches used to bring up Boot Camp somewhere on the first page of the results OR would link you to the product page for one of the Macs. But now, you’ve got to actually KNOW to search for Windows or Boot Camp... in other words, you have to already know Macs do that (which today’s buying public doesn’t) in order to find any information about it. You won’t even see a mention of those OR of Windows on the pages for the Mac Pro or iMac Pro... where one would THINK it’s appropriate.

So, while a lot of people see emulation and virtualization as critical features, I don’t think that group of folks are large enough or important enough to Apple for them to even try to maintain this as a feature. Consider this, Final Cut Pro was VERY important to Apple at the time that Steve looked Final Cut professionals in the eye and basically said, ”We’re going in this direction... bye!”

 

[Samuelsan2001]

Regarding any OS other than macOS executing on an ARM Mac, I’d just say “Don’t expect it”. Go to Apple’s website right now and in the search bar, do a search for both “emulation” and “virtualization”. You won’t find any search results under Explore or Support for the emulation of other OS’s. Under Accessories you WILL find Parallels. The search for virtualization doesn’t show anything related to the virtualization of other OS’s and you won’t even find Parallels under Accessories!

These searches used to bring up Boot Camp somewhere on the first page of the results OR would link you to the product page for one of the Macs. But now, you’ve got to actually KNOW to search for Windows or Boot Camp... in other words, you have to already know Macs do that (which today’s buying public doesn’t) in order to find any information about it. You won’t even see a mention of those OR of Windows on the pages for the Mac Pro or iMac Pro... where one would THINK it’s appropriate.

So, while a lot of people see emulation and virtualization as critical features, I don’t think that group of folks are large enough or important enough to Apple for them to even try to maintain this as a feature. Consider this, Final Cut Pro was VERY important to Apple at the time that Steve looked Final Cut professionals in the eye and basically said, ”We’re going in this direction... bye!”

Many Microsoft products are already available on arm Powered Apple hardware. Hell they are also available natively for OSX. The need for virtualisation and bootcamp is a fairly niche need these days not to mention cloud streaming and virtual private networks for other services and internet based apps. I have needed Windows precisely once in the last decade to run a protein modelling software suite, one that is now available online if I am not mistaken. The need for virtual machines and bootcamp is disappearing fast it’s that simple. Final Cut Pro is the driving force behind the Mac Pro it’s hardly been left behind by Apple. Hell you can edit 4K video at 60 FPS on your iPhone these days, and there are a plethora of good software suites for photo and video editing for the Mac and iOS OSes. You can also run premiere and other adobe software on macs iPads etc already. As I say this is a non-issue for the vast majority of people and for those where it is an issue I see no reason to believe the the developers of virtual machine software and emulators will stop or ignore an arm powered app.

 

[k27]

Why? Because my MBP will run cooler and the battery will last longer, and my apps will run faster?

Well...
Intel Macs could run better. Unfortunately Apple does strange things. Maybe even on purpose to make the ARM Macs look better?

Example:

2020 Air Heatsink Modification Thread

Update: went ahead with the heatsink and heat pad mods and oh...my... The performance gains are absolutely amazing, but the temperature drop and stability is even more impressive in my opinion. This has turned my MBA into the powerhouse it ought to be. Left side is pre-mod GeekBench testing...

forums.macrumors.com

 

[cmaier]

Well...
Intel Macs could run better. Unfortunately Apple does strange things. Maybe even on purpose to make the ARM Macs look better?

Example:

2020 Air Heatsink Modification Thread

Update: went ahead with the heatsink and heat pad mods and oh...my... The performance gains are absolutely amazing, but the temperature drop and stability is even more impressive in my opinion. This has turned my MBA into the powerhouse it ought to be. Left side is pre-mod GeekBench testing...

forums.macrumors.com

because Intel didn’t hit the thermal performance it promised. Speed doesn’t come in a vacuum. Sure, Apple could redesign its thermal solution to accommodate intel’s failure to meet the performance it promised. But that’s the problem, isn’t it? Apple can design the boxes it wants to design the way it wants them to look and with the reliability it wants with its own chips and beat Intel, because apple’s speed per watt will be 20 percent higher.

 

[k27]

No, the real problem is that Apple sometimes makes strange decisions. Too much functions follows design decisions. A lot also for economic reasons, to push customers towards certain products. Otherwise, there would be a Mac Midi, for example.
This can also happen with Apple's ARM-SoCs. The cooling system can be bad there as well.

And we do not want to start talking about poor software quality and ignoring bugs. This is bad. I expect more from Apple. Especially with the high prices and the limited product line.

 

[Darmok N Jalad]

Frankly, Intel has really blown it. They’ve been recycling the same architecture for 4 years now and are still stuck on the same 14nm node. They have not really been able to accommodate Apple’s ULV needs, and even all the way up to Xeon in the Mac Pro, AMD has much better and more affordable options. Combine that with the fact that the chips Apple sources from Intel are of the more expensive variety, and it pretty much begs Apple to find an alternative. The exact same thing happened with IBM and the departure from PowerPC—it just wasn’t competitive any longer, and how long to do you wait? I’m a little surprised Apple never took a chance with Ryzen, but maybe Apple is comfortable controlling their own destiny now and have decided to make their own.

 

[timidpimpin]

I blame AMD and Intel for Intel's shortcomings. Not for AMD doing better now, but for them not forcing Intel to be better since about 2011. Between Athlon (1999) and Ryzen (2017), AMD did nothing at all to really motivate Intel to do better. Sandy Bridge in 2011 (2nd gen i3 - i7) is really the last huge push forward Intel has had.

 

[dspdoc]

I blame AMD and Intel for Intel's shortcomings. Not for AMD doing better now, but for them not forcing Intel to be better since about 2011. Between Athlon (1999) and Ryzen (2017), AMD did nothing at all to really motivate Intel to do better. Sandy Bridge in 2011 (2nd gen i3 - i7) is really the last huge push forward Intel has had.

Healthy competition certainly does push innovation! Intel seems to have lost their way and gotten comfortable. They are probably very concerned now.

 

[timidpimpin]

Healthy competition certainly does push innovation! Intel seems to have lost their way and gotten comfortable. They are probably very concerned now.

Word around the industry has said that over the years Intel made pretty big R&D cuts, because the executives saw it as a waste when they were so far ahead of the competition. And it wasn't until around 2018 that they really started to fully staff R&D again, so they have a lot of ground to make up, and I imagine it will take at least 3-5 more years for them to truly be competitive again.

The one saving grace they still have are their huge multi-year deals with many big PC makers. AMD really only dominates the custom PC market still. So that will carry Intel until they can engineer a true game changer again.

 

[dave255]

Perhaps apple won't even bother porting osx to arm?

Maybe they will just have a new ios/iPadOS branch for arm macs (macOS?) with better multitasking/handling of multiple app windows. Would save them the hastle of porting an os & creating x86 emultation..and overtime just have all devices running a flavor of ios and phase out osx?

 

[dspdoc]

Perhaps apple won't even bother porting osx to arm?

Maybe they will just have a new ios/iPadOS branch for arm macs (macOS?) with better multitasking/handling of multiple app windows. Would save them the hastle of porting an os & creating x86 emultation..and overtime just have all devices running a flavor of ios and phase out osx?

Federighi has point blank stated that Apple will not cannibalize their lineup like this.

 

[timidpimpin]

Perhaps apple won't even bother porting osx to arm?

Maybe they will just have a new ios/iPadOS branch for arm macs (macOS?) with better multitasking/handling of multiple app windows. Would save them the hastle of porting an os & creating x86 emultation..and overtime just have all devices running a flavor of ios and phase out osx?

You don't seem to understand that iOS/iPad OS are macOS. Not the other way around. Your concept is horrifying to all but the most computer skill limited consumers. Do you honestly think Apple avoided a touch Mac until now because an Intel can't do that? I hope not.

They keep things separate for a reason. Pretty much every current Mac is geared to true computer users, and out of user demand. Your logic is very flawed.

 

[dspdoc]

You don't seem to understand that iOS/iPad OS are macOS. Not the other way around. Your concept is horrifying to all but the most computer skill limited consumers. Do you honestly think Apple avoided a touch Mac until now because an Intel can't do that? I hope not.

They keep things separate for a reason. Pretty much every current Mac is geared to true computer users, and out of user demand. Your logic is very flawed.

Thank you for saying this! Some of us want to continue to use ACTUAL computers and not some watered down version of a full OS.

 

[chucker23n1]

Frankly, Intel has really blown it. They’ve been recycling the same architecture for 4 years now and are still stuck on the same 14nm node.

They've definitely screwed up the 10nm transition and lost almost half a decade, but they seem to be on the path to recovery. Ice Lake on the Air was a big jump forward compared to Amber Lake (14nm), and Tiger Lake seems to continue that path. They also canceled most of Cooper Lake (another 14nm server platform) in favor of moving much of Xeon straight to Ice Lake this year.

Combine that with the fact that the chips Apple sources from Intel are of the more expensive variety, and it pretty much begs Apple to find an alternative. The exact same thing happened with IBM and the departure from PowerPC—it just wasn’t competitive any longer, and how long to do you wait?

Right.

I’m a little surprised Apple never took a chance with Ryzen, but maybe Apple is comfortable controlling their own destiny now and have decided to make their own.

That, and we also don't know NDA'd details: what does Intel's, AMD's, and Apple's own roadmap look like? Maybe Apple's is far more promising than AMD's, so why bother?

I blame AMD and Intel for Intel's shortcomings. Not for AMD doing better now, but for them not forcing Intel to be better since about 2011. Between Athlon (1999) and Ryzen (2017), AMD did nothing at all to really motivate Intel to do better. Sandy Bridge in 2011 (2nd gen i3 - i7) is really the last huge push forward Intel has had.

You're leaving out AMD64, which was so much better in practice than Itanium, Intel had to license it and now uses it exclusively.

 

[theluggage]

No, the real problem is that Apple sometimes makes strange decisions. Too much functions follows design decisions. A lot also for economic reasons, to push customers towards certain products. Otherwise, there would be a Mac Midi, for example.
This can also happen with Apple's ARM-SoCs. The cooling system can be bad there as well.

There are many ways that Apple could mess up an ARM transition (too sudden, too expensive, form-over-function, locked-down, dumbed-down, buggy software, not actively working with key developers...) but, frankly, none of those mistakes require an ARM processor. Until we see what Apple actually does over the next year (I'm sure everything announced on Monday will sound... magical) all we can do is look at what is technically possible.

In technical terms, there's no reason why all software that is still a "going concern" shouldn't get re-built for ARM over the next 12 months, and for modern software there's no reason why that should be a particularly onerous task c.f. the usual annual maintenance - assuming Apple does a competent job of making MacOS for ARM. Equally, there's no reason why MacOS ARM shouldn't have half-decent x86 emulation/translation facilities to smooth over the transition. This is MacOS we're talking about - not Windows - so it's not like users expect 20 year-old software to run, and the 64-bit switch just culled most of the "abandonware".

The unavoidable casualty is going to be running high-performance x86 Windows software on Mac hardware - I'm sure that ways of running your tax software (or whatever) will emerge - even under full emulation, or via ARM Windows' built-in x86 emulation. As for Linux, the bulk of the big open-source projects already support ARM64. Docker etc. should be able to run either with an ARM Linux VM (MacOS has a built-in virtualiser which Docker for Mac already uses) or, potentially, with an emulated x86 Linux VM. Yes, the former will need Docker images built for ARM, but that is already a "thing" - and those images are mostly compilations of open-source software that already builds for ARM. The latter - well, if you're developing and testing server-side stuff - no GUI, or the GUI runs on the client, and your testing only uses a fraction of the software's capacity - will x86 emulation be so bad? Will it make a practical difference to your workflow if the container is running on a virtual server in the cloud (or a PC tucked in a cupboard somewhere)?

Sure, that won't happen if Apple turns the Mac into a locked-down clamshell iPad running a castrated version of MacOS - or if they junk the entire Intel mac line on Monday forcing you to "upgrade" as soon as your trusty 2012 MacBook fails - but let's see what they do.

 

[Darmok N Jalad]

I blame AMD and Intel for Intel's shortcomings. Not for AMD doing better now, but for them not forcing Intel to be better since about 2011. Between Athlon (1999) and Ryzen (2017), AMD did nothing at all to really motivate Intel to do better. Sandy Bridge in 2011 (2nd gen i3 - i7) is really the last huge push forward Intel has had.

Healthy competition certainly does push innovation! Intel seems to have lost their way and gotten comfortable. They are probably very concerned now.

That’s the thing, competition wasn’t healthy, and while AMD missed big-time on the Bulldozer architecture, Intel was also using anti-competitive practices during the time when AMD actually was competitive. It was bad enough that Intel settled out of court to the tune on 1.25B USD. When you are significantly harmed in a super-competitive industry, it’s hard to deliver. AMD recovered anyway, largely thanks to a very solid design and roadmap, good execution, and Intel’s 10nm woes. Zen 3 is still on track, so here’s hoping AMD can keep the foot on the gas and gain marketshare. Just look at how much CPUs have gained since Ryzen. Gone are the days of premium-priced quad cores—those are now budget CPUs.

 

[cmaier]

I blame AMD and Intel for Intel's shortcomings. Not for AMD doing better now, but for them not forcing Intel to be better since about 2011. Between Athlon (1999) and Ryzen (2017), AMD did nothing at all to really motivate Intel to do better. Sandy Bridge in 2011 (2nd gen i3 - i7) is really the last huge push forward Intel has had.

Your dates are wrong. AMD put a ton of pressure on Intel in the early 2000s with Athlon 64 and Opteron. That destroyed itanium and forced Intel to license Amd’s 64 bit isa. It also sold well and got a lot of design wins, including the first Dell design win.

Talk about pressure.

but after that AMD forced out the team responsible for that success. Because AMD loves to shoot itself in the foot
[automerge]1592750529[/automerge]

Perhaps apple won't even bother porting osx to arm?

Maybe they will just have a new ios/iPadOS branch for arm macs (macOS?) with better multitasking/handling of multiple app windows. Would save them the hastle of porting an os & creating x86 emultation..and overtime just have all devices running a flavor of ios and phase out osx?

macOS is already ported to arm. It’s run on arm for years.

 

[smulji]

Federighi has point blank stated that Apple will not cannibalize their lineup like this.

That's not what Federighi said. What he said was that iPadOS and macOS are not merging.

 

[ChrisA]

And we do not want to start talking about poor software quality and ignoring bugs. This is bad. I expect more from Apple. Especially with the high prices and the limited product line.

As it turns out the money is with iPhones. They started ignoring the Mac line when the iPhone sale took off. They even moved their best engineers away from Macs to iPhones. There has been little innovation with MacOS in the last 10 years. OK, some color changes on the screen and other trivial changes. But only one substantial change, that would be the new file system APS,

In the old days every new version of MacOS was faster and worked better and their suite of apps (Pages, Numbers, iMovie,...) got major updates every 6 to 8 weeks. That all stopped years ago.

Then they dropped Aperture. This was a signal to all users of their professional line that Apple could abandon them and you better have a plan-B for when they do. I suspect they drop Final Cut X and Logic at some point. and then merge the iPad and Mac line to both run the same OS on Arm.