Apple's Path to Arm-Based Macs Could Start With a New 12-Inch MacBook

[0924487]

I think a much better approach than just a switch is to have dual processors. An ARM and an Intel CPU over the next five years. That way, tasks that only run on Intel allow the Mac to use more power and turn on the Intel processor. In all other cases, it would use a much lower power state of a twelve-core A-Series ARM SoC. This, if implemented well, could allow Apple and developers a path to ensure customers get the best of both worlds. Apple can show its prowess and SoC capabilities without leaving out Intel/x86/Windows and etc.

This scenario has never been offered but it seems to me to allow the best of both worlds. Apple could run its own graphics which would probably destroy anything AMD has available. The SoC can do certain things much faster and better than Intel. But for those Intel-only apps that are power hungry not alienate them.
It’s truly the best of both worlds. It requires some advanced code and a rosette model to emulate all possible for running as much as possible on ARM CPUs. At the same time, when it just isn’t feasible like with Adobe Premiere Pro or other intensive apps, allow Intel to shine.
This strategy would allow a win-win for Apple, customers, developers and etc. as Customer will not be left out in the cold and developers will have time to implement a new instruction set.

anyone want to give their thoughts?

Like having an Intel M3 coprocessor?... Then you would need a dual architecture kernel that supports both x86 and ARM instruction set, which would be a bit of work.

If you are thinking, well, what about T2 and Intel Core i7 etc, this is because Apple runs a separate kernel and OS called BridgeOS on T2 which also powers the TouchBar. They are separate, and interoperate via APIs, to have Intel as a coprocessor, and have a coherent experience, and draw on the same monitor or monitors, that's gonna be some kernel-level work on Apple's part.

 

[cmaier]

They probably had Mac ARM running on Mac OS for the last several years at least.

Great performing chips. And the last noteworthy one is two years old. The A12x is some beasty performer.

I'm looking forward to see what the A14x can do. How Apple announce Mac ARM and what the road map it. And if they give Mac Dev's a Mac ARM device to play with to move iPad apps and 'Mac' apps over to Mac ARM. It doubt it's any more difficult than last time. And M$ and Adobe are already on ARM, anyhow. Already on the App stores.

I'll enjoy pulling the trigger on the Intel iMac at the same time.

Azrael.

To me the only question is whether and how they allow existing x86 binaries to run. Other than that, it’s all good news. (M$ and adobe on ARM aren’t the real apps. But they will port the real apps.)

 

[hanske68]

“Desktop OS” is coming.

In short: DOS

Nice innovation, very much in line how apple is reinventing the laptop with their ipad. haha

In time Apple will 'innovate' into something the new an amazing xBook

 

[Project Alice]

AMD’s x86 license becomes invalid if they are acquired. It’s non-transferable.

But doesn’t AMD license x86-64 to Intel? I’m pretty sure it’s a cross license.

 

[Azrael9]

To me the only question is whether and how they allow existing x86 binaries to run. Other than that, it’s all good news. (M$ and adobe on ARM aren’t the real apps. But they will port the real apps.)

You raise an interesting point.

What's that tech' Valve are using to bring x86 games to Linux? (Fusion? Can't recall. But it allows a wine style running of x86 without WIndows?)

Azrael.

 

[cmaier]

But doesn’t AMD license x86-64 to Intel? I’m pretty sure it’s a cross license.

The poster said “buy AMD.” If they buy AMD, then the license, cross-license or not, goes away. Intel may or may not actually need the license, and it may or may not be part of a cross-license. (Case law is mixed on whether you actually need a license for just an instruction set - in other words, whether or not an instruction set is copyrightable. )

What’s for sure is you need it for the zillion patents that you would inevitably infringe.

 

[Detnator]

The new ARM-Based Macbook is happening.
View attachment 923776

Oh for the love of...

This is a nice mockup but seriously... It’s not going to have MagSafe... or any of the other old ports. Native MagSafe is dead. Get over it.

If you want what MagSafe offers there’s countless USB-C options and that’s exactly why Apple won’t and doesn’t need to do it.

ð usb-c magsafe - Google Search

www.google.com

 

[mrchinchilla]

Cool. Looks like by the time that I buy a new Mac, the transition to ARM will probably be complete.

 

[Lounge vibes 05]

Not at all.

When Steve came back, I think one of the smartest early decisions he made was to radically simplify the Mac line-up. Sure, there were other factors at the time, like Apple not having the money for such a broad line-up. But part of it was also to avoid the paralysis of choice. A 2x2 grid. You want a desktop or a laptop? You're a pro, or a consumer? Done.

Nowadays, Apple has many more products, but some aren't necessary. The 13-inch MacBook Pro with 2 ports probably shouldn't exist. It's an alibi product to say, "hey, the Pro starts at $1299", but really, the Air configured to the same price point is a better product for most people, and the $1799 is the "real" Pro. Almost nobody should get the two-porter.

But it was way worse than that three years ago. Do you get the Air? Oh, but it has no Retina Display. Or the MacBook, that's kind of like an Air? Oh, but it's much more expensive, yet slower. Or the two-port Pro, which was introduced by Phil as Air-esque? Well, that one's even more expensive. And despite being Pro, it lacked many of the things of the other Pros, such as Touch ID.

I was happy that they simplified and consolidated the line-up in 2018 and 2019, and I hope they're not going to step back from that.



I've actually written a ton about this subject.



I do! Because it matters that I can tell someone who wants a Mac "oh, what are your requirements? Alright: this is the one you should get". Right now, I can do that. For almost half a decade, it was hard, in part because of the above, and in part because of keyboard reliability issues. "Oh, I guess just get the old Air; at least it has a good keyboard."



Those don't overlap much, so I fail to see your point.



That is exactly what they did both with the 68k to PowerPC and PowerPC to Intel migrations, and it wasn't a big issue, so yes, I think they should replace products rather than add to them. But that's also way besides my point.
[automerge]1592077497[/automerge]


Maybe, but I don't think "we have a great transition plan for developers: buy two Macs!" is going to be a popular message.

while I agree that for a small time in the late 90s the Mac lineup was simple, it quickly got complicated.
In January 1999 he introduce the power Macintosh G3 blue and white, then barely 6 months later the power mackintosh G4.
Then there was the power mackintosh G4 cube.
Then the 15 inch titanium PowerBook G4
Then the eMac G3 and the iMac G4.
Then they introduced the 12 and 17 inch PowerBook G4, but the 15 inch was still titanium while the other two were aluminum.
Then came the iMac G5, and the Mac energy forThen came the iMac G5, and the Mac mini G4.
So by 2005, there was a iBook G4, a PowerBook G4, a Mac mini G4, an eMac G4, an iMac G5, and a power mackintosh G5, with a server machine as well.
Then for a while, the lineup was pretty simple with the MacBook, MacBook Pro, Mac mini, iMac, and Mac Pro.
But then it got super complicated. By 2008, there was a White plastic MacBook at $999, an aluminum MacBook at $1199, A 15 inch MacBook Pro, an outdated 17 inch MacBook Pro, MacBook Air that was more expensive than the MacBook and the 15 inch MacBook Pro and super under powered, a Mac mini that haven’t been updated in a long time, an iMac, and the Mac Pro.
Then, it was simplified again. In July 2011, there was a MacBook Air that came in 11 and 13 inch, a 13, 15, and 17 inch MacBook Pro, the Mac mini, iMac, and Mac Pro.
But they couldn’t stay satisfied with that lineup for too long, because in June 2012, they introduced the MacBook Pro with retina display on top of the two MacBook Pro’s that already existed.
Basically, the point I’m trying to make is that the Mac lineup has never ever been extremely cut and dry. It’s always been complicated

 

[radiologyman]

I waited to see if Apple will bring a successor to MB12 but eventually bought IMHO a next best thing, HP Spectre 360 13.3in. It has very narrow bezels but slightly bigger and heavier than MB12 given larger screen. With gen 10 Intel processor it rarely if ever uses fan but still provides decent CPU and GPU performance. Also has convertible touchscreen with great keyboard and slick design but most importantly to me an absolutely amazing 3840 x 2160 OLED screen. I will still buy an ARM MB12 when it comes but hopefully Apple will add miniLED along with second USB-C connector.

 

[MikeZTM]

I suppose an "auxiliary" x86-64 chip would only run applications, and the main OS would run completely on the ARM chip. If emulation is good enough to offset simply installing another chip, then we will see that, though I would imagine on the higher end models (iMac /iMac Pro/MBP16/Mac Pro/MBP13 4ports) it would at least be an option to maintain compatibility and the highest performance possible for applications that have not migrated to ARM.

If we look at Windows 10 ARM today we could see the performance of emulating x86 software is pretty good.
They got 50% performance compare to native arm64 software. And we all know windows legacy software is much harder to emulate than unified API macOS software.

If A14 have a faster single core performance than current A13@2.6GHz(euals to about 90% of a 9900k single core in SPEC2006) even after 50% it will be on par with today's MacBook Pro running at 2.3GHz.

There no need for a real x86 CPU inside the machine to do anything. Just like how a PowerPC core is useless inside a Intel Mac 15 years ago.

 

[chucker23n1]

If A14 have a faster single core performance than current A13@2.6GHz(euals to about 90% of a 9900k single core in SPEC2006) even after 50% it will be on par with today's MacBook Pro running at 2.3GHz.

There no need for a real x86 CPU inside the machine to do anything. Just like how a PowerPC core is useless inside a Intel Mac 15 years ago.

Well, to be fair, the first Intel iMac was a massive performance jump compared to the iMac G5 from three months prior. I don't think we'll see the same kind of jump this time. So I do think, if emulated at all, x86 perf will take a hit.

 

[MikeZTM]

Well, to be fair, the first Intel iMac was a massive performance jump compared to the iMac G5 from three months prior. I don't think we'll see the same kind of jump this time. So I do think, if emulated at all, x86 perf will take a hit.

The performance jump will be quite amazing thou.

We all know current line up of 13 and 16 inch MacBook Pro all runs at really low clock speed for thermal and power reasons. They are nowhere comparable to Intel's desktop CPU models. And a A13@2.6GHz single core is reaching desktop 9900k's single core boost at 5GHz performance.

Just put that dual core A13 in a 12inch MacBook you already got at least double the performance than a MacBook Air 2020 dual core.
That put the emulated performance at least faster than a MacBook Air 2020 dual core.

If the rumor for a 8 big core A14 is true then it will be faster than a MacBook Pro when emulating x86.

 

[chucker23n1]

The performance jump will be quite amazing thou.

We all know current line up of 13 and 16 inch MacBook Pro all runs at really low clock speed for thermal and power reasons. They are nowhere comparable to Intel's desktop CPU models. And a A13@2.6GHz single core is reaching desktop 9900k's single core boost at 5GHz performance.

There isn't really that much difference in single-thread performance between Intel's desktop and laptop CPUs. The 9900K is only 19% faster than the highest-end 16-inch MBP, and only 9% than the 13-inch (which is ahead of the 16-inch due to being on Sunny Cove).

You're right that the A13 is pretty close to high-end Intel. But 9% better is nowhere near enough to make up the emulation penalty.

Just put that dual core A13 in a 12inch MacBook you already got at least double the performance than a MacBook Air 2020 dual core.

Hard to say since an A13 is 2+2 cores, but on single-core, it's around a 30% difference.

 

[cmaier]

There isn't really that much difference in single-thread performance between Intel's desktop and laptop CPUs. The 9900K is only 19% faster than the highest-end 16-inch MBP, and only 9% than the 13-inch (which is ahead of the 16-inch due to being on Sunny Cove).

You're right that the A13 is pretty close to high-end Intel. But 9% better is nowhere near enough to make up the emulation penalty.



Hard to say since an A13 is 2+2 cores, but on single-core, it's around a 30% difference.

I expect 20% improvement > than each intel part an apple arm is replacing.

I wonder if Apple stuck anything interesting in their hardware to make compatibility a little easier. One thing that comes to mind is the weird intel floating point (80 bits? It’s been awhile since i worked on x86 FP and drawing a blank, though I do remember it was a real pain after having worked on FP on powerpc and SPARC). Another is extra registers dedicated to emulation (both general purpose and things like stack pointer). If someone asked me to design an ARM cpu that could emulate x86 well, I can think of some things I could do with not-too-many transistors, particularly if I am already supporting heterogenous cores and I already have an SoC methodology which lets me build extra little cpus and stick them in the package.

 

[firewood]

That's what the iPad Pro is for.

Unless they're one of the few who don't like touchscreens, or one of the many who need to connect large (27"+) displays due to aging eyeballs.
[automerge]1592174316[/automerge]

To me the only question is whether and how they allow existing x86 binaries to run.

They do a deal with MS for subscription remote Windows desktops on Azure, et.al. ?

 

[sidewinder3000]

Not at all.

When Steve came back, I think one of the smartest early decisions he made was to radically simplify the Mac line-up. Sure, there were other factors at the time, like Apple not having the money for such a broad line-up. But part of it was also to avoid the paralysis of choice. A 2x2 grid. You want a desktop or a laptop? You're a pro, or a consumer? Done.

Nowadays, Apple has many more products, but some aren't necessary. The 13-inch MacBook Pro with 2 ports probably shouldn't exist. It's an alibi product to say, "hey, the Pro starts at $1299", but really, the Air configured to the same price point is a better product for most people, and the $1799 is the "real" Pro. Almost nobody should get the two-porter.

But it was way worse than that three years ago. Do you get the Air? Oh, but it has no Retina Display. Or the MacBook, that's kind of like an Air? Oh, but it's much more expensive, yet slower. Or the two-port Pro, which was introduced by Phil as Air-esque? Well, that one's even more expensive. And despite being Pro, it lacked many of the things of the other Pros, such as Touch ID.

I was happy that they simplified and consolidated the line-up in 2018 and 2019, and I hope they're not going to step back from that.



I've actually written a ton about this subject.



I do! Because it matters that I can tell someone who wants a Mac "oh, what are your requirements? Alright: this is the one you should get". Right now, I can do that. For almost half a decade, it was hard, in part because of the above, and in part because of keyboard reliability issues. "Oh, I guess just get the old Air; at least it has a good keyboard."



Those don't overlap much, so I fail to see your point.



That is exactly what they did both with the 68k to PowerPC and PowerPC to Intel migrations, and it wasn't a big issue, so yes, I think they should replace products rather than add to them. But that's also way besides my point.
[automerge]1592077497[/automerge]


Maybe, but I don't think "we have a great transition plan for developers: buy two Macs!" is going to be a popular message.

I appreciate your well written replies, and you seem like a decent guy, but I still fail to understand all the handwringing you and others seem to have over Apple’s Mac lineup. Especially since all we are talking about here is introducing a single ultralight to the lineup that can serve as a light use laptop for people with basic needs and can easily be powered by the first ARM based Mac chip Apple rolls out. The expectations for the power of such a small computer will be low, and so Apple will be able to exceed those with what will undoubtedly be an impressive ARM chip, without disrupting current workflows. Especially for schools and corporations, who won’t want to switch to ARM chips until they are available across the lineup, so that they can minimize the expense for their IT departments. Introducing it in the MacBook Air or Pro could make things difficult for businesses. This will allow a much more gradual and less draconian transition.

Introducing ARM to the Mac lineup via a revamped MacBook seems like a wise idea. If they’d taken that strategy with Final Cut Pro X when they launched it, and came out with A light version called Final Cut Pro X Express first, then the launch wound’t have been one of the greatest shitshows, and massive exoduses, in Apple history. I speak from experience on this one.
[automerge]1592183257[/automerge]

Not at all.

When Steve came back, I think one of the smartest early decisions he made was to radically simplify the Mac line-up. Sure, there were other factors at the time, like Apple not having the money for such a broad line-up. But part of it was also to avoid the paralysis of choice. A 2x2 grid. You want a desktop or a laptop? You're a pro, or a consumer? Done.

Nowadays, Apple has many more products, but some aren't necessary. The 13-inch MacBook Pro with 2 ports probably shouldn't exist. It's an alibi product to say, "hey, the Pro starts at $1299", but really, the Air configured to the same price point is a better product for most people, and the $1799 is the "real" Pro. Almost nobody should get the two-porter.

But it was way worse than that three years ago. Do you get the Air? Oh, but it has no Retina Display. Or the MacBook, that's kind of like an Air? Oh, but it's much more expensive, yet slower. Or the two-port Pro, which was introduced by Phil as Air-esque? Well, that one's even more expensive. And despite being Pro, it lacked many of the things of the other Pros, such as Touch ID.

I was happy that they simplified and consolidated the line-up in 2018 and 2019, and I hope they're not going to step back from that.



I've actually written a ton about this subject.



I do! Because it matters that I can tell someone who wants a Mac "oh, what are your requirements? Alright: this is the one you should get". Right now, I can do that. For almost half a decade, it was hard, in part because of the above, and in part because of keyboard reliability issues. "Oh, I guess just get the old Air; at least it has a good keyboard."



Those don't overlap much, so I fail to see your point.



That is exactly what they did both with the 68k to PowerPC and PowerPC to Intel migrations, and it wasn't a big issue, so yes, I think they should replace products rather than add to them. But that's also way besides my point.
[automerge]1592077497[/automerge]


Maybe, but I don't think "we have a great transition plan for developers: buy two Macs!" is going to be a popular message.

Also, I’ve been recommending Macs to people since 1984, and nothing about the segmentation of their lineup has made it any harder over the past 5-10 years. The MacBook was a weird, underpowered computer, and so I didn’t recommend it to anyone. Boom. How hard was that?

The keyboard thing has certainly been an issue in regards to overall quality, but that has nothing to do with their overall lineup. At one point, all the keyboards sucked. Luckily, we’re over that nightmare now. I’m typing this on an iPad Pro Magic Keyboard and am happy to report that all is well again in Apple Keyboard land.

 

[Zorori]

If we look at Windows 10 ARM today we could see the performance of emulating x86 software is pretty good.
They got 50% performance compare to native arm64 software. And we all know windows legacy software is much harder to emulate than unified API macOS software.

If A14 have a faster single-core performance than current A13@2.6GHz(equals to about 90% of a 9900k single core in SPEC2006) even after 50% it will be on par with today's MacBook Pro running at 2.3GHz.

An interesting thing to note with the Windows machines is that the Surface X's Geekbench performance is closer to an A10X than the A12X. Slower than the A10X in single-core, but faster in multi-core:

	Single-Core	Multi-Core
Surface Pro X	3500	11600
A10X	3900	9500
A12X	5000	18000

Surface pro x - Geekbench Search - Geekbench

 

[SocialKonstruct]

Whatever Apple comes up with I don't want a glorified iPad Pro as the Macbook 12

 

[Gerdi]

You raise an interesting point.

What's that tech' Valve are using to bring x86 games to Linux? (Fusion? Can't recall. But it allows a wine style running of x86 without WIndows?)

Azrael.

You are mixing things up here. Issue hat hand is different ISA, you refer different OS but same ISA with your Valve example.

 

[cmaier]

Whatever Apple comes up with I don't want a glorified iPad Pro as the Macbook 12

SpaceX capsules also run on ARM. Does that mean that if a macbook uses ARM it must be a “glorified” version of a spacex capsule?

 

[firewood]

Gruber's Daring Fireball blog is quoting Brendon Shanks hypothesis that it's going go be the A12Z SOC from the iPad Pro inside a Mac Mini or Apple TV chassis, which requires less engineering to create a developer-only box than a full MacBook. If so, I wonder if the A12Z can run at a higher thermal envelope then when wedged inside an iPad? Or if it's going to be a chip in a different package with something like a bigger heat spreader? Maybe real (non-dev) ARM Mac customers will have to wait till next year for some higher wattage variants of the A13 or A14.

Dalrymple's Loop blog might be passing along a whisper that Apple already has these systems working, ready to go, and that they are "fast".

The hints are out that Apple wants developers to really pay attention to the coming Keynote and Platform sessions.
[automerge]1592196957[/automerge]

SpaceX capsules also run on ARM.

But the fastest (known) CPU is on the Voyager 2 spacecraft.

 

[smulji]

Whatever Apple comes up with I don't want a glorified iPad Pro as the Macbook 12

I agree with much of this article. You won't have to worry about that @cmaier

Re-engine, Not Re-imagine

My predictions for the ARM Mac transition

bslabs.net

 

[incoherent_1]

Apple is going to tell us that macOS and all Apple’s 64-bit apps have been living a secret double life for years and we just had to phase everyone to 64-bit only with Catalina. Plus most of the key third party devs have all their 64-bit apps working on ARM aLrwady, and for the rest it’s just a checkbox in XCode to recompile for Apple’s new ARM chips. Point being, the two switches are effectively the same and all the current up to date 64-bit apps that we all had find as replacements for our old 32-bit apps for Catalina already run on ARM so we don’t have to find more replacements.

At least it’d be nice if that’s how it goes. 😉. It’s probably wishful thinking.

I hope you’re right!

Although it’s just 2-3 applications, I have some legacy programs that haven’t been updated in awhile but are still very important. Spending time and $ to find replacements isn’t something I relish.

 

[smulji]

Dalrymple's Loop blog might be passing along a whisper that Apple already has these systems working, ready to go, and that they are "fast".

That post was written by Shawn King. How much insider info does HE have? I know Dalrymple knows people on the inside.