I don’t like writing about rumors, but the ARM Mac transition is happening, and it’s almost certainly being announced at WWDC — when else could it be announced? — so here is John Gruber’s speculation on what to expect:
There’s been some reasonable speculation that Apple might use ARM chips only for portable Macs and consumer desktops, and stick with high-end Intel chips for pro desktops, but I think that’s simply based on our never having seen Apple even try its hand at high-performance chips. If you’re going to switch, switch.
If you want to see how this is going to play out, just rewatch Steve Jobs’s 2005 announcement of the PowerPC-Intel transition. As I wrote back in 2018, it’s uncanny how similar the explanation could be: Apple’s in-house ARM-based chips offer vastly superior performance-per-watt compared to Intel’s, and Apple has ideas for future Macs that they can’t build without that superior performance-per-watt. All computers benefit from superior performance-per-watt, not just portables. That was the story in 2005, and it should be the story in 2020.
There are other reasons too. It’s cheaper for Apple to make its own chips than to buy Intel’s. They already make a $400 iPhone that out-benchmarks a $3,000 top-of-the-line MacBook Pro in single-core CPU performance. That’s bananas when you think about it. And there is a cross-platform developer story. With one consistent set of system-on-a-chip designs, all software for all Apple platforms can target the same Metal APIs for the GPUs, and the same neural engine APIs for machine learning and AI tasks.
The rumor mill currently leaves some big questions unanswered, though.
I think Gruber is spot on, especially when it comes to moving the entire line of products to ARM. It doesn’t make sense to split the line, and when you really think about it, most of the popular apps, from Photoshop to Office, have for years been written for different processors — PowerPC and Intel in the 90’s/early 00’s, and today for Intel and Apple’s A-series ARM processors already in iPhone and iPad. Apple’s biggest benefit in going to Intel in ‘05 was improved power and battery life. The fact that Photoshop was now able to run on the same CPU architecture was a happy accident; in no way did it make the transition for anyone easier.
Gruber touches briefly on a very interesting idea in the footnote of this post. He doesn’t name it, but a device that straddles the Mac and iPad — the MacPad — would be a very interesting device. I don’t think this is as simple as macOS on an iPad Pro or iOS on a MacBook, but I think a true merging of both hardware and software that brings the best of both worlds into a single device. A Finder-like filesystem, iOS gesture controls, AppleScript or Shortcuts+, full screen apps and full keyboard support, and Pro apps like Xcode and Final Cut Pro.
MacOS and iPad OS feel like two distance celestial bodies — like the Milky Way and Andromeda Galaxies — destined to merge in the distant future. We can already see some early collisions — keyboard shortcuts in iPadOS 13.5, Catalina’s super lockdown of the Desktop — and while Apple claims they envision two distinct platforms, hardware and software changes. A year ago no one was predicting an Apple-designed scissior-switch keyboard for the iPad, but here I am, typing on it.
But speaking a little closer to present day: If Cook doesn’t do the “About this Mac...” reveal at the end of the keynote showing the Mac was running on an ARM CPU, I’m going to be pretty upset.