You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!

  • One line of macOS code could cap a 20-year pivot

    When Steve Jobs made it clear at the 1997 Apple Worldwide Developer Conference he was taking back his company, he tossed the now famous line in his opening monologue: “Focusing is about saying no.” Approaching 20 years later, that decision still reverberates.

    Jobs had deeply considered what to focus on, and the answer wasn’t Mac hardware – it was software. NeXT technology was being deployed on a new operating system project, Rhapsody, that had to win big with developers and third parties. Focusing on Rhapsody meant defocusing something else, something the ARM community had worked very hard to make happen. From “Mobile Unleashed”, p. 102:

    After more Mac-related questions, there was a curveball: What do you think Apple should do with Newton? Jobs recoiled and laughed uncomfortably, muttering, “You had to ask that.” He propped himself against a display table and paused an agonizing 15 seconds, staring at the floor. He sighed deeply, and finally slightly nodded his head in self-affirmation of what he was about to say.

    In very measured tones, Jobs said rarely can companies manage two large operating systems, and he could not imagine succeeding with three. With Mac OS and Rhapsody, Newton made three – it had nothing to do with how good or bad Newton was, it was unsustainable.
    How do you kill off your mobile strategy when you know full well a mobile strategy is your company’s future? The only answer was to make the operating system for the mobile strategy a fork of the desktop strategy. In the beginning, that was nearly impossible. ARM processors simply did not have enough processing power to run a desktop operating system. Avoiding the problem temporarily, Apple bootstrapped the iPod around the Pixo application framework with important tweaks to the user interface – including a handwritten note from Jobs to change the font to Chicago to match the look and feel of the Mac.

    When “Project Purple” got serious about designing an Apple phone in November 2005, the operating system decision came front and center again. By then, ARM processor performance had increased to the point where running real operating systems was a possibility, and the Purple team floated the idea of using Linux. Jobs nixed that thought, insisting instead on modifying the Apple operating system code base. Mac OS X would need heavy modifications to run on a small device, and most of Apple’s software team was tied up porting Mac OS from PowerPC to Intel architecture. Project Purple stole Mac OS X engineers as they were freed up and ripped out code to create what would become iOS.

    This week, Dutch journalists crawling through release notes for the just released macOS 10.12 “Sierra” found what they think is evidence a switch to ARM is really close.

    Article: Samsung going vertical Qualcomm cry CEVA laugh-one-line-arm-support-macos-10-12-jpg

    The name of the “big” core inside the Apple A10 hasn’t been outed just yet. “Hurricane” would certainly be consistent with recent Apple-designed ARMv8 core names including “Cyclone”, “Typhoon”, and “Twister”. Observers point with glee to Geekbench and other benchmark results showing Apple A10 chips ahead of comparable Intel notebook class parts. We also haven’t seen the A10X chip yet, the next logical increment of higher performance parts for iPads with increased memory bandwidth and bigger display capability.

    Still, the rumors are rampant a MacBook with an ARM chip inside is near. We know it takes up to a couple of years to move the operating system from one processor architecture to another. The PowerPC to Intel change was exacerbated by endianism issues, and with more experience on ARM architecture porting may go faster this time.

    “Hurricane” may also be an entirely new core in an entirely new chip. With Intel’s recent announcement of foundry capability for ARM technology, Apple could be lining up an all-new chip fabbed exclusively by Intel for a MacBook while staying the course on iPhone and iPad chips with TSMC.

    And, they could be planning a merge of the MacBook and iPad Pro into a 2-in-1, more along the lines of the successful Microsoft Surface Pro 4. That would be news. A few years ago, I wrote that we haven’t exactly seen ARM and Intel side by side yet. Running Geekbench is one thing; succeeding in an ecosystem with consumer popularity is another. The ARM-based Surface lines died a painful death simply because consumers wouldn’t accept the software incompatibility.

    What Apple is trying to pull off depends on getting right what Microsoft messed up. The theory was good: Microsoft Windows 10 would be one code base running on desktop, tablet, and mobile. My wild guess that Microsoft would put an Intel-based mobile chip in a phone to make that easier appears to have also died a painful death.

    I’m sure that Apple is out getting macOS 10 apps ported from Intel to ARM architecture right now; they’ve announced some changes in how developers submit apps that would be friendly to a platform change. In the last cycle, Adobe Photoshop was the gold standard of application ports and was initially a step back on Intel compared to PowerPC. Eventually that was solved, especially as Intel vector instruction sets got better.

    What would be really interesting is if Apple is running a long game and “Hurricane” is an all-new core supporting ARMv8-A SVE (Scalable Vector Extensions). That, plus a re-architected SoC designed for a 2-in-1 and fabbed on Intel 10nm could be what Steve Jobs said he wanted for the Mac 20 almost years ago – not just better, but much better than alternatives. It could also be the real reason behind Intel signing an ARM foundry deal.

    I’m guessing the ARM-based MacBook shoe doesn’t drop next month as some have speculated, but instead sometime next year when ARM-macOS, the app ecosystem, and Intel 10nm capacity all line up. It would be an epic 20-year pivot, no?