We have heard rumours about this forever. The earliest leak by a credible source was from way back in 2011, when SemiAccurate claimed Apple was dumping Intel from laptops. At the time, it was still some ways away, but other leaks started corroborating this, and it was widely believed Apple will move at least the MacBook Air to Arm by 2013-15. Well, those time frames came and went. Apple introduced a fanless MacBook which would have seemed like a prime candidate for Arm, but it was still powered by Intel. What happened? While it's certainly possible that the rumours around 2011-12 were just that, I think they delayed the plans due to one simple reason. And that is Windows RT.
Microsoft had made the move to Arm in 2012 with Windows RT, a clean start that did not run old x86 apps. Instead, it had to rely on native Arm apps, of which adoption was scant. Combine that with the anemic performance of Arm SoCs at the time, and Windows RT was a huge flop. It's easy to see how Apple must have had similar plans, and were clearly spooked by Windows RT's failure.
Of course, Microsoft didn't give up, and neither did Apple. They finally released a proper Windows 10 on Arm in 2017. This time, it was fully featured Windows 10, with app virtualization for x86 apps. Adoption has been slow, but that's because there's a wide selection of Intel and AMD systems available. With Apple, you don't get any choice, you're at their mercy, so adoption will be fairly rapid. Like Windows 10, the next macOS will also support virtualization for x86 apps, though there'll be a performance cost. Apple goes one step further with offering dynamic translation with Rosetta 2 (remember, Rosetta? Used for the PowerPC to Intel transition 15 years ago). I'd expect developers to adopt macOS on Arm much more quickly than Windows 10 for native code, for the aforementioned reasons.
What about timing? Apple claims the transition will happen over a 2 year period. We'll see the first Arm-based PCs (yes, they are PCs, deal with it) from Apple late this year. It's fairly safe to assume this will be a refresh of the MacBook and MacBook Air lines, as Apple already has Arm SoCs comparable to the Intel ones powering those. It's very likely to be an A14X or a A14Z or something and will almost certainly be a performance leap over Intel's current Ice Lake parts. The performance here is not in question - it's more about compatibility, and how soon developers will adopt native Arm code. Of course, first-party apps like Final Cut Pro will be ready at launch. The next product to move will be the MacBook Pro 13.
The more interesting transitions will be to the MacBook Pro 16, iMac and Mac Pro lines. These feature higher performance CPUs, and more importantly, discrete GPUs. MacBook Pro 16 and iMac are pretty much the same internally, so they'll like feature the same SoCs - I'd guess these will land in late 2021/spring 2022. Can Arm really match Intel's 8-core parts? I'd say, definitely. Firstly, AMD itself is far ahead of Intel at this point, which offers some clues. Apple could certainly make a SoC with 8 of their big cores, let's call it A14XXX+++, and that'd very likely be competitive with what Intel or AMD offers. Apple can go one step further and make larger dies and cleanly outperform AMD and Intel, even. AMD and Intel have to account for such worldly things as cost effectiveness, and thus have to balance die size with performance. Apple has no such worries, Mac users have paid hefty premiums in the past, and will continue to do so.
Of course, this raises the question of - are Apple going to make their own discrete GPUs? Certainly, these SoCs will feature integrated graphics, which have proven to be quite capable for iPads and iPhones. However, these do not hold a candle to the state-of-the-art AMD and NVIDIA offer. My best guess is that for the foreseeable future, Apple will continue using AMD Radeon graphics. One outside possibility is they will license AMD's graphics architecture and rebrand them as their own - much like they do with Arm. Of course, it's also possible that they have indeed developed their own graphics tech - but I just don't see that happening just yet. Maybe a decade down the line? Sure.
Finally, we get to the Mac Pro. This will obviously be the last product line to transition to Arm. Today, the fastest Mac Pro comes with a 28-core Intel CPU which offers performance leaps and bounds ahead of any Arm processor. And that itself is very anemic compared to AMD's latest 64-core Zen 2 CPUs. But once again, it's a matter of piling on the cores, and there's no reason to believe Apple couldn't make a huge Arm die with 64 cores. More likely, they'll copy AMD's chiplet approach and go with multiple small dies. Indeed, I won't be surprised if there entire line-up is chiplet based. MacBook Air ships with 1 8-core chiplet, MacBook Pro and iMacs ship with 2, and the Mac Pros can go up to 8 chiplets or more at 64 cores. Of course, all of this is just wild speculation, but AMD is clearly leading the industry in that direction.
Finally, there's macOS 11. Yes, we're finally past the X era, and I can stop arguing with people who keep calling it X, the alphabet, where its clearly 10 in Roman numerals. The major changes in 11 is the underlying OS now updated to harness Apple's Arm silicon. But the user interface also sees a complete overhaul. I don't like the look of it, it very much is an iOS-ification. I'm sure it's no big deal, people will get used to it. I plan to remain a Windows 10 user, with some stuff line PopOS (Ubuntu distro), but I'll definitely be buying a macOS 11 device to keep up with the developments. If you're a developer, Apple is already selling Mac Minis powered by A12Z - the same SoC powering the latest iPad Pros. Quite an odd choice, in hindsight, as the A14-class devices powering the first Arm Macs will be far more powerful, but they're simply not ready at this time.
Overall, it's been a long time coming, and it's finally here. If you're a Mac user, I don't think there are many concerns just yet, though if you're using niche professional apps, it's time to get them ported over to Arm. For most of the world using Windows, this can only help Windows' own adoption of Arm. Of course, I'd expect a vast majority of Windows and Linux users to continue using Intel and AMD, but the new breed of Arm CPUs will definitely expand choice.
More information in Apple's press release here: https://www.apple.com/newsroom/2020/06/apple-announces-mac-transition-to-apple-silicon/
There was a segment on Linus's wan show on the weekend that mentioned that perhaps Apple were hamstringing thermal solutions on the Intel chips to allow their ARM based chips to seem more performance by running cooler.
This was amusing, in a follow up video, they removed the heat sink from a modern apple laptop, which wasn't making correct, or proper contact with the chip die - leading to incredibly poor thermal performance.
I watched the video. Of course, Linus was joking, they are still competing in the market today. Another conspiracy theory was that these laptops were designed for Arm, but they couldn't get their silicon ready in time, so had to shoehorn in higher power Intel processors. This one doesn't make much sense either, as there's cTDP down options on Intel.
Apple have had some abysmal thermal solutions over the years. Some of the engineering decisions made by their hardware division for Macs have been really baffling - if it was any other company, they would probably have gone out of business. Fortunately, it looks like the latest MacBook Pro 16 seems to have finally fixed their long time glaring flaws - chiefly the unreliable keyboard and the inadequate thermal solution. It's still not up to par with their competitors, but at least now it's not throttling hard.
It's entertaining :)
Will be very interesting to see where they end up. Especially with AMD's new mobile chips being very tasty for productivity.
Thanks for your contribution to the STEMsocial community. Feel free to join us on discord to get to know the rest of us!
Please consider supporting our funding proposal, approving our witness (@stem.witness) or delegating to the @stemsocial account (for some ROI).
Please consider using the STEMsocial app app and including @stemsocial as a beneficiary to get a stronger support.