• DreamHost



  • Newsletter Issue #995

    June 14th, 2020

    DOES APPLE REALLY WANT TO ALIENATE CUSTOMERS WITH ARM ON A MAC?

    A lot of what passes for tech journalism — or mainstream journalism for that matter — fails to recognize the history behind a matter. So for several years, as Apple produced faster and faster custom ARM-based CPUs that vanquished competing silicon, there has been growing speculation that the Mac will get them next.

    On the surface, it makes plenty of sense sense. In recent years, Intel has begun to hit the wall in improving the performance of its own CPU. Its efforts to build mobile processors haven’t gone so well. At one time, the low-power chip that came to be known as Atom was considered as a possible contender for use in the iPhone, or perhaps other Apple mobile gear.

    It never happened, and Intel’s efforts to move from its core competency – PC processors — haven’t gone so well. For a time Apple even bought billions of dollars of baseband modems for iPhones from Intel, but the company hasn’t been able to scale up to 5G. In the end, Apple settled a simmering series of lawsuits with Qualcomm to buy its hardware instead.

    And they also acquired Intel’s cellular modem technology, so Apple may just succeed where Intel failed.

    This isn’t to say that Intel is an unsuccessful company, or that its future is in doubt. It will continue to power hundreds of millions of personal computer each year for many years to come. It’s even possible that its difficulties in reducing a chip’s die will succeed.

    But Apple is in the enviable position of being quite capable of adopting its own technology. Recent Macs already include an ARM-based T2 chip that delivers low-level support for such functions as media encoding and storage technology. Intel still provides the CPU, but how long will that last?

    Switching from Intel to ARM is certainly no easy task. But Apple has been unique in the PC business in already doing processor transitions twice. In 1994, with the performance of its Motorola-built 680×0 processors hitting the wall, Apple launched the Power Macintosh, powered by PowerPC chips designed by IBM and Motorola.

    Apple didn’t rest on its laurels. Mac OS X was designed to be processor independent, and even when it was first released, beginning in the fall of 2000 with the Public Beta, Apple was working on alternatives to the PowerPC.

    Over the next few years, the PowerPC failed to realize expectations. The Mac was the only PC platform to adopt the chip, so development was focused more and more on embedded processors for the auto business and other needs. While Apple touted the G5 as the ultimate technological achievement, it was a difficult leap.

    The most powerful professional Macs with G5s inside required liquid cooling. The G5 was never tamed for use in a PowerBook, which continued to lag behind performance of PC notebooks.

    During the 2005 WWDC, Steve Jobs announced the move to Intel beginning in 2006. He said it would take a year to compete, but the job was done by summer with the release of the replacement for the Power Mac G5, the Mac Pro.

    And so it went.

    The latest iPad Pro offers performance essentially equivalent to a MacBook Air and the low-end MacBook Pro except for graphics performance, which is better. And that’s not even the latest A-series chip, the A13 Bionic. An A14 is expected for the iPhone 12 launch later this year.

    So just what level of performance can Apple achieve with its own CPUs?

    Let’s not forget that the chips run in a very resource constrained environment on an iPhone and an iPad. They have to provide decent performance and all-day battery life in these tiny devices. The largest iPad, a 12.9-inch Pro model, weighs around 22 ounces. Compare that to a three-pound notebook computer, and consider how much Apple can boost performance when installed in a much larger box.

    As I write this, the virtual WWDC keynote is scheduled for Monday, June 22 and 10:00 AM Pacific Time. It is widely expected that plans to move the Mac to ARM will be announced there, and a MacBook with the new CPU could arrive by the end of the year.

    Of course, few would doubt that Apple has been working on such a processor switch for several years. Few would doubt that the initial announcement will include claims that the new Mac gear will be much faster than Intel-based models they replace.

    There are, of course, questions yet to be answered, and some members of the tech media have resorted to fear-mongering as a substitute for ignorance.

    Now I don’t pretend to have any secret information, but the history of Apple’s previous processor transitions should serve as clues.

    So there will be some sort of Intel emulation capability to allow you to run most or all existing apps until developers have the chance to switch. It will likely be hardware-based to reduce or eliminate the performance penalty, especially if the new Macs are much faster than the Intel models.

    Apple’s Catalyst technology allows a developer to build an iPadOS app and have it also run with little or no change on Macs. Apple has released a handful of Catalyst apps for macOS Catalina, such as News, Podcasts and Music. They run well enough, but it may take a year or two for any decent number of cross-platform Mac apps to appear.

    But one tech pundit claims that Apple will limit ARM Mac support strictly to Catalyst apps, meaning the vast majority of software won’t work. That hardly makes any sense, since it would doom the product before it went on sale.

    Another theory has it that the apps would have to be sandboxed, meaning they are eligible for sale via the Mac App Store. Again, this would cripple the product and not allow tens of thousands of apps to be used. Consider Rogue Amoeba’s Audio Hijack, a mission-critical tool for podcasts and other media productions as a key example.

    Instead, I expect Apple to make the transition as seamless as possible. The vast majority of ARM Mac buyers will be able to run the same software they do now as developers build updated apps to exploit the new technology.

    The main question marks are Boot Camp, which lets you run Windows in a native environment, and virtual machines. At one time, being able to run two or more operating systems on your Mac was a critical feature, but published reports indicate that Boot Camp is only used on 2% of current Macs.

    But that doesn’t mean Apple will abandon the technology even if there’s some level of a performance hit. Indeed, the promise of high-performance emulation may be a feature of the new hardware.

    If all of most of this comes to pass, it’s very likely Apple will have some surprises in store too. If an iPad Pro can meet the speed of an Intel-based notebook, what about an ARM-based Mac without the constraints? Twice as fast? One report suggested a 12-core MacBook, with eight cores being high powered, and four cores being low powered for maximum battery life.

    Don’t forget that powerful multicore ARM silicon is already used by servers and datacenters. So don’t underestimate where Apple can take this technology.

    Again, speculation has it that entry-level Macs will go ARM first, and it may take a while to see those chips in an iMac Pro or a Mac Pro?

    Or maybe not? Apple completed its Intel transition in roughly eight months. Will the ARM migration move even faster?

    It’s never a good idea to bet against Apple.

    THE FINAL WORD

    The Tech Night Owl Newsletter is a weekly information service of Making The Impossible.

    Publisher/Editor: Gene Steinberg
    Managing Editor: Grayson Steinberg
    Marketing and Public Relations: Barbara Kaplan

    Copyright 1999-2020 Making The Impossible. All Rights Reserved.

    Privacy Policy: Your personal information is safe with us. We will positively never give out your name and/or e-mail address to anybody else, and that’s a promise!



    Share
    | Print This Issue Print This Issue

    One Response to “Newsletter Issue #995”

    Leave Your Comment