• Explore the magic and the mystery!



  • Newsletter Issue #1013

    December 29th, 2020

    APPLE’S M1 MACS: MUST YOU WAIT?

    If you’ve used Macs for a long time, say over 25 years, you might recall the original processor migration. Then it was the journey from Motorola 680×0 chips to the PowerPC, said to be the ultimate processor. But it wasn’t such a friendly change.

    For one thing, just about every app you used, even those from Apple, hadn’t made the trip to the new processor. So they had to run in emulation. So they’d run, more or less, about the same as Macs using the 68030 CPU, such as the IIci or the IIx. If you were upgrading from a Quadra, for example, using the 68040 CPU, you lost the performance advantage till the apps were updated.

    Now if you weren’t using Macs way back then — or weren’t even around — no worries. Most of this should be pretty clear anyway.

    And porting the app wasn’t such an easy process, nor did developers rush to do the work. Forget about waiting years on end for Adobe, Quark or Office for Mac. Even small developers took their sweet time, or just decided that it was time to jump off the Mac train and embrace Windows. I’d like to say which apps got there early, but that was long ago and far away, so let’s just say that the situation grew even more difficult due to Apple’s foolish design choices.

    After a few years, as the PowerPC grew faster with the G4 and — later the G5 — it wasn’t unusual to have native apps on your Mac.

    The PowerPC was rapidly approaching a dead end as the 21st century progressed. In 2005, Steve Jobs announced, at that spring’s WWDC, that the migration to x86, Intel Inside, would commence the following year and be complete by the end of 2006.

    This time the situation wasn’t quite as difficult. After all, most of Apple’s developer community was already building Windows versions of their apps — using x86 architecture of course — so making a Mac version of their products wasn’t quite as difficult as the previous generation. Also, Apple’s sales were higher, meaning more sales potential. Also, the emulation scheme, Rosetta, was actually quite decent and the speedier Intel silicon reduced the performance loss compared to the original PowerPC hardware.

    Making the transition more promising was the arrival of Boot Camp, the ability to run Windows natively in a separate partition on a Mac drive, and via a decent OS virtualization scheme such as Parallels Desktop. Apple finished the migration four months earlier than promised with the arrival of the first Mac Pro in August, 2006.

    Now the Intel move was not just necessary — PowerPC development had stalled — but was productive. There were regular performance boosts each year, though not as much as one might have hoped. So while my Late 2014 27-inch iMac with 5K Retina Display (say that five times backwards) may be pokey when benchmarked against today’s M1 Macs, it is actually not altogether slow at all. Its Fusion Drive has most of the speed of a full SSD, apps open relatively quickly, and general performance is quite snappy with macOS Big Sur 11.1.

    Yes, my iMac is now passé, “vintage” in Apple parlance, which means an Apple Genius isn’t inclined to want to fix it if it breaks, but it’s still perfectly capable of handling my normal workflow. Now if I had to edit lots of 4K videos in Final Cut Pro X, maybe not.

    The long and short of it is that improved performance has hardly been a reason to regularly upgrade your Mac, unless your Mac is getting long in the tooth.

    In contrast, Apple has seen consistent performance upgrades at a decent clip every single year with its own A-series silicon. Partnering with TSMC, who manufactures most of the chips, Apple has been able to build hundreds of millions of them, so it’s a sure thing that they’ve gotten it right.

    That said, there are still critics who toss in fear-mongering falsehoods about the potential of the first crop of Macs with the M1 chip that arrived this fall. It can’t possibly work, and what about all those Intel apps that people need to run? How many years will it take for developers to get in line? What about Adobe and Microsoft, both of whom have taken their sweet time in the past to catch up with Apple’s processor switches?

    But Apple isn’t asking developers to go where nobody has gone before. Its Xcode developer tools have been used for years to compile apps for its various gadgets using both ARM and Intel chips. So making a “Universal” binary, an app that runs native on Intel and the M1, is not as big a jump as, say, making a PowerPC app run native on Intel.

    During its rollout of the new hardware on November 10th of this year, Apple was already demonstrating how many apps had been ported, and how well most Intel apps would run using Rosetta 2 on the new systems. By running at 60-80% of native, emulated apps are, for most people, working faster than on regular Intel systems.

    More to the point, Adobe and Microsoft are doing their jobs. You can already upgrade most Office 365 apps in Universal binaries. There’s already a feature-limited beta vein of Adobe Photoshop for M1 Macs; Lightroom is already available, and most or the rest will probably arrive some time in 2021. Apps from smaller developers are already available, and others are no doubt mostly in the pipeline.

    What this means is that most people who buy the new Macs — and a number of others are expected in 2021 — will not have to suffer much in finding compatible software.

    There are exceptions, of course, and you’re best advised to check the publishers of the apps you need to run to see about their plans. As it stands, the most critical apps in my production workflow are already native to Apple Silicon, and a few of the others run fine, I’m told, in Rosetta 2.

    If you depend on Windows virtualization, though, that’s another story, although third parties have made Windows run quite well on M1 Macs. An official version from Microsoft — which has already made an ARM-based version of Windows that sort of runs on Surface tablets — will probably occur. Microsoft has been very good about making its apps in near-equivalent versions for Macs in recent years, and the speedy arrival of M1 versions of Office clearly demonstrates that Apple’s chosen path is the right one.

    Boot Camp? Well, the few people who need it will have to just wait. But that doesn’t mean Apple won’t have a solution if there’s a demand for it. A virtualization solution will probably do the job.

    Despite what’s really happening, you will still read about the alleged difficulties in making Intel apps compatible with the M1, along with false complaints about the performance advantages.

    Now I understand people get things wrong. But it does seem that Apple gets more than its share of unfair criticisms about things. The underlying assumption is that it must fail with everything it does, and that its decisions are simply wrong.

    I would grant that Apple sometimes does things that do not make any sense, or reflect a level of greed that doesn’t always serve the customer so well. I can tell a few stories, such as the time when my wife’s iPhone 6s malfunctioned, and they wanted to charge me more than twice its original purchase price to repair broken mics. Fortunately I was able to talk them out of that outrageous posture and they agreed to replace the unit.

    But it’s also true that Apple has prospered, confounding the skeptics, and that cannot be ignored. Some of these skeptics do it because it’s hit bait pure and simple, which is why I don’t provide links. They don’t deserve the Google AdSense payments.

    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

    Leave Your Comment