News: 1587665888

  ARM Give a man a fire and he's warm for a day, but set fire to him and he's warm for the rest of his life (Terry Pratchett, Jingo)

The rumor that just won't die: Apple to keep Intel at Arm's length in 2021 with launch of 'A14-powered laptops'

(2020/04/23)


Apple will reportedly release its own Arm-powered Mac computers next year, marking the biggest component change since Cupertino ditched IBM's PowerPC chips in 2005 for [1]Intel's x86 line .

Citing anonymous internal sources, [2]Bloomberg claims Apple has designed three laptop processors based on the upcoming A14 platform, with the codename Kalamata. These will be more powerful than the processors found on the equivalent iPad and iPhone devices, and will gradually be rolled out across the firm's computer line, starting with the least-powerful machines.

Presumably this means the MacBook Air, which represents Apple's entry-level computer since it ditched the MacBook. Although Bloomberg suggested Apple could roll out an entirely new laptop line dedicated to its new Kalamata chips.

The report cites TSMC as the fabricator for the new Arm processors, which will purportedly use a 5nm process and have 12 processing cores – eight high-performance cores, codenamed Firestorm, and four energy-efficient cores, dubbed Icestorm. This will represent a step-change over the current entry-level MacBook Air, which has just two processing cores.

Although a move to Arm would offer some degree of commonality between the Mac and the iPhone/iPad, it doesn't necessarily spell doom for macOS, with Apple reportedly planning to preserve it as the mainstay of its laptop and desktop computing lines.

Famed Apple analyst chances his Arm-based Macs that Apple kit will land next year [3]READ MORE

"Apple will leave Intel for Arm" is the consumer tech equivalent of "20XX is the year of the Linux desktop." For years, [4]prophesies and predictions have swirled about Apple, leaving the loving embrace of Chipzilla to make it on its own, with little substance to back them up. But this time feels different.

Already we've seen a steady flow of Arm laptops, with Google and Microsoft offering compatible spins of ChromeOS and Windows 10 respectively. These efforts have won the support of chip makers like Rockchip and Qualcomm, as well as laptop vendors like Samsung and Lenovo. And that's largely thanks to the narrowing of the performance gap, allowing Arm-based platforms to provide a steady, relatively fluid experience.

Arm-based chips also have several key advantages, particularly on the power efficiency and connectivity fronts. The Lenovo Yoga 5G, for example, packs a built-in 5G modem and a 24-hour battery.

Of course, any move to Arm would present its own challenges – although none insurmountable. There would be inevitable compatibility issues with applications built for Intel, with developers required to recompile them for the new platform, or users required to use an emulation layer.

Apple has experience here, following the move to Intel from PowerPC. The transition was relatively short too, considering the amount of legacy software that's inevitably prevalent on such a popular and storied platform. For a few years, it offered the Rosetta binary translation tool layer as users and developers alike transitioned to the new platform, eventually retiring it with the release of [5]mac OS 10.7 Lion .

But it would also have its own advantages, offering a greater degree of commonality between macOS and iOS, making porting apps somewhat simpler.

From the eyes of this observer, a switch to Arm feels more viable than ever. Not only does Apple have years of chip-design experience, itself bolstered by several strategic acquisitions, but the technology has matured enough to allow a transition from the power-thirsty Intel chips that have represented the [6]mainstay of Apple's computer line for over 15 years. ®

Sponsored: [7]Legacy Modernization: Finding Your Way With Low-Code



[1] https://www.theregister.co.uk/2005/09/20/apple_intel_mac_timeline/

[2] https://www.bloomberg.com/news/articles/2020-04-23/apple-aims-to-sell-macs-with-its-own-chips-starting-in-2021

[3] https://www.theregister.co.uk/2020/02/26/apple_arm_based_macs/

[4] https://www.theregister.co.uk/2017/02/01/apple_weaning_off_of_intel/

[5] https://www.theregister.co.uk/2011/07/21/adobe_lion_incompatibilities/

[6] https://www.theregister.co.uk/2006/08/07/apple_xeon/

[7] https://go.theregister.co.uk/tl/1936/-8553/legacy-modernization-finding-your-way-with-low-code?td=wptl1936

jpo234

>Apple has experience here, following the move to Intel from PowerPC.

And Motorola 68K ==> PowerPC before that.

No surprise then

Steve Davies 3

Intel has failed to deliver most of what they have promised in the last 5 years.

As the article stated, they have lots of ARM design experience and seem to think that they can take the overall performance of a MacBook to levels that Intel can't deliver.

It remains to be seen if they can make it happen. Intel should have seen this coming a mile off.

ExampleOne

Two thoughts:

1) Apple already has an ARM powered laptop: the iPad Pro with the Smart Keyboard.

2) How to completely alienate your professional workstation customers: finally announce a long awaited successor to your workstation line, then effectively EOL it before it has even hit full availability.

Apple may well be working on a new 5nm higher performance ARM chip, but I somehow doubt they are planning to kill the Intel laptop and desktop lines. They can easily enough run both, given the slight split in their market between consumer and production devices.

werdsmith

iOS or iPadOS is not really as enabled as MacOS, for instance, it doesn’t really support native coding/compilation.

enabled?

mevets

The i*OS are all based on the same kernel as macos -- darwin. Do you imagine that some apple engineer thought one day "hey, I'm going to cross compile the toolchain and run them on my ipad?"; even if just for kicks? The Ipad has enough ram, cpu and storage to use as a dev machine; why wouldn't they?

Re: enabled?

doublelayer

Possible? Certainly. Do I think it happened? No. If it did, we'd have it. IOS is a functional OS for mobile devices, but there are tasks it doesn't handle well. One of those tasks is manipulating lots of files, keeping them organized, etc. One of those things you do a lot when writing software. Another is spinning up a terminal session to run multiple small tools on files. IOS doesn't do that either. For nondevelopment purposes, you don't need those things and most users won't notice their absence, but devs would.

You can call nearly anything portable a laptop. As long as it has a processor in it and can be carried with you, it qualifies. There's still a major difference between a traditional laptop that runs a desktop operating system and other devices that do less. Even when it's shaped like a laptop with a keyboard and everything, most tablets are still just tablets with a keyboard.

How about both?

IGnatius T Foobar !

I wonder if it would be viable to place Intel *and* ARM processors in the same computer, similar to how the Commodore 128 had both 6502 and Z-80 onboard. I imagine it would require some sort of special-purpose hypervisor to keep everything running in symphony.

Re: How about both?

Anonymous Coward

Reminds me of the Dec Rainbow PC in the 80’s - Z80 and 8086, able to run either CP/M or MS-DOS. I used one for 5 years, running Lotus Symphony (under MS-DOS). Had CP/M boot disks but only once booted that up (out of curiosity).

Re: How about both?

tip pc

Was that tongue in cheek?

Are you suggesting something like the Pc dos card from 1995?

https://en.wikipedia.org/wiki/Power_Macintosh_6100#Overview

rcxb

Arm-based chips also have several key advantages, particularly on the power efficiency

ARM looks good only in apples-to-oranges comparisons. Extremely low performance ARM CPUs have very low power usage, but the higher performance ones use more power than an Intel Atom or some mobile ULV CPUs. Those touting the low power and high performance of ARM CPUs are either lying or are themselves falling for the lies of interested parties, and sadly propagating them.

There isn't a single ARM core out there which comes anywhere close to the performance of a single modern x86 core. Either Apple will have to wait for that to change, or they will have to force all developers of MacOS apps to continually maintain fat binaries for two architectures indefinitely, or Apple will just have to surrender the high performance PC market entirely, which won't make graphical designers and video editors very happy.

If Apple was smart, they'd only consider releasing an ARM laptop as an iPad-with-keyboard type device running IOS. Call it an iClam... That would avoid the risk of hamstringing their MacOS business in the process.

Geekbench?

mevets

Geekbench disagrees with you. The A13 single core beats the i9-9900 by about 7%. The multi-core is quite a bit different, but heat dissipation is a big issue; if you exclude imac and mac pro; the macbook pro (8 core) has about a 50% lead; but still has a lot better heat dissipation than an iPad.

Do remember, Apple make in-house designed ARM CPUs, that are quite a bit different from the more generic line.

IGotOut

"but the higher performance ones use more power than an Intel Atom or some mobile ULV CPUs. "

I'm confused, you've saying high performance ARMs use more power than an Atom or ULV.

That's like saying a Porsche uses more fuel than a moped.

thames

I don't see why Apple would bother with fat binaries. The concept was made obsolete years ago by the Internet. It only made sense when you bought your software on CD in boxes.

Linux manages to support multiple chip architectures without fat binaries. The OS has a standard installer and it knows what chip architecture it's running on. It just fetches the correct binary for that architecture from the correct repo. I run Ubuntu on x86 and Ubuntu on ARM without fat binaries and it's completely transparent to me as a user. I'm sure that Apple is capable of copying that idea if necessary.

As for performance, if you care about per core CPU performance above all else you're not likely to be running an Apple laptop anyway. I don't seriously see it as an issue for the bulk of their target market. Graphics performance is what will matter most, and that is down to the GPU.

For overall performance, they will rely on having a large number of cores for peak loads, and throttling down to a few lower powered cores whenever possible. Have a look at all the work that Apple has spent the past few years putting into multi-core process dispatching. They won't have been doing that work if they didn't intend to make use of it.

mark l 2

I don't think that Apple wanting to move over to ARM for their laptops (if the rumour is true) is necessarily that they think they can make an Intel CPU killer, Although the performance gap between Intel and ARM is a lot closer than it used to be. But instead that they can save costs by designing their own CPU in house rather than buying expensive Intel chips which Intel have not been great at getting any decent yield at smaller die sizes. Apple can also optimize their own designed CPU better for MacOS, as Intel chips come with all those old X86 instructions that Intel have to include to support legacy code that MacOS doesn't need. Obviously a move to ARM would mean dual booting to Windows 10 would no longer be an option, unless MS were to release an ARM compatible Windows 10 version that supported Apple hardware.

thames

What proportion of Apple's user base are still dual-booting Windows? It's something that I read about as being possible but rarely hear about people actually making use of these days. I doubt that Apple are going to cripple the future development of their product line by maintaining a legacy use case that most of their users don't care about.

There's also nothing preventing Apple from keeping a few legacy x86 versions of their laptop line around for those customers who have a genuine use case for them, at least for a while. I say 'x86', because these could quite as easily be AMD and Intel.

Around 95-99% of professional users

Anonymous Coward

And probably <1% of consumers. Maybe 0.1%

It's been painfully obvious for many years that Apple just want the professionals to go away. Yes, they buy the really expensive Macs, but there aren't many of them compared to the consumers buying Airs, iPads and the like.

A lot of professional software isn't available for macOS anymore because Apple removed the features it needs, and the cost of redevelopment far exceeds the likely return.

So most professionals switch over to Windows (or Linux, surprisingly enough) to get some of their work done. They complain of course, because they happen to like the OSX GUI.

But they're using the machine to do a job, and they'll buy and use the hardware and software needed. Dropping 32bit forced many over, because Photoshop.

In many cases they've not bought a new macbook for years because the newer ones were terrible.

Stuart Castle

Still slightly dubious about this..

They've just launched an extremely expensive upgrade to the Mac Pro. Even Apple aren't going to want to piss off customers who will happily spend tens of thousands of pounds on each workstation.

Yes, Apple have changed CPU architectures several times. Generally not without providing some sort of emulation layer in the OS to enable compatibility with the previous architecture. Can any existing arm emulate the x86 architecture at the same speed as a core i5, i7 or i9? What about the Xeons? Bear in mind that the target markets for the Mac are likely to require tons of processing power (not so much the print and audio design, but video work and 3d work can bring most computers to their knees, pc or mac).

Not all of it.

Richard 12

It'd only need to emulate part of amd64.

They already killed x86-32, and they control the only supported compiler, so they can have been ensuring it uses a specific subset of opcodes.

They've also forced every developer in the world to send them a copy of every piece of software for macOS, including alpha and beta builds, so they have a huge pot of IP to steal applications to test against a compatibility layer.

Plus arbitrarily refusing applications that happen to use opcodes they don't feel like bothering with, perhaps.

This Time I'm Telling The Truth

Imhotep

I don't believe anything when the sole source is Bloomberg. They've been too wrong, too often.

The rumor needs to die before the Macintosh does.

VicMortimer

If Apple does this, it's because they've given up on the Mac.

Apple's earlier architecture changes made sense. They were painful to go through, but there were good reasons for them. The PPC was a huge speed win from the 68k. The G5, great chip that it was, just never got low-power enough to put in a laptop. The whole fat binary problem, the slow and crashy emulators, and worst of all was when the PPC emulation was finally really good, and could have kept legacy software going for decades, Apple stripped it out with no warning. But this, this one is utter stupidity if you're trying to be a computer company. We're finally at the point where there's an industry standard architechture for CPUs, one that works well, and one that isn't even single source unless you're Apple (and the Intel exclusivity deal ended years ago, but we still have no AMD Macs, and no NVidia GPUs in them). If Apple wants another chip, there's one right there at AMD. As far as that goes, Apple could buy AMD with cash on hand, and have enough left over to buy Adobe.

The only reason they'd try is because they know it'll kill the Mac, and that's what it seems like they ultimately want, a sealed box that's not even really a computer, just a content consumption tablet that can be used for some light word processing, and a few industry-specific apps, but spends most of its time as a toy.

What they don't realize is that killing the Mac WILL eventually kill Apple.

Hopefully Cook comes to his senses and kills this garbage before it ever gets close to a release.

Re: The rumor needs to die before the Macintosh does.

doublelayer

I also like having a standard instruction set that is generally open, but your characterization of it being multi-supplier is a little strong. Basically, the only available options are Intel and AMD, with other companies not being allowed to join the party. With ARM, there are many manufacturers and a few designers of processors implementing that instruction set. Qualcomm, Broadcom, Samsung, TI, Apple, ARM themselves, Huawei, and a couple small places that don't make many chips. ARM has many other problems, like not having a consistent method of booting firmware--I can take virtually any X86 chip and run arbitrary code on it, but not so with an ARM one, but in terms of suppliers and lock-in, ARM is probably better.

Intel eclipse, next ?

luis river

I believed chip A14 (will be total surprise ) it offer more performance than eagerly awaited for press. Traductor "Rosetta" applications it need extra-power (offering by CPU A14 ) for at least equal performance Intel´s CPU. Intel had catch "suitcase" and to go very far, this time is begin of the Intel eclipse?

The more complex the mind, the greater the need for the simplicity of play.
-- Kirk, "Shore Leave", stardate 3025.8