We’re about to enter an exciting, messy transition. Not only will Apple Silicon make better Macs, it will force Microsoft to polish its Windows on ARM act, both hardware and software. In turn, this will cause PC OEMs to reconsider their allegiance to x86 silicon…and that will have serious consequences for the old Wintel partnership.
——
https://mondaynote.com/apple-silicon-the-passing-of-wintel-79a5ef66ad2b
———
by Jean-Louis Gassée
——-
Why should Intel worry about Apple’s decision to base future Macs on homegrown Apple Silicon SoC (System on a Chip) devices? According to Dataquest and IDC estimates, Apple owns no more than 7% of the PC market. Furthermore, Apple doesn’t buy the expensive Xeon chips, used in millions of Cloud servers, that represent a growing proportion of Intel’s revenue. And the company is a headache: It makes demands and complaints way out of proportion with the amount of revenue it generates. Losing Apple is more symbol than substance.
Not so fast. The impact on Intel — and the entire industry — will be felt beyond Apple’s small share of the PC market.
Apple isn’t simply dropping a proudly designed homegrown CPU in place of an Intel chip on Mac motherboards. Moving to Apple Silicon is an expensive undertaking that affects hardware and software engineering, developer relationships, marketing… If the switch to Apple Silicon were a mere CPU replacement, billions of dollars would burn in a bonfire of vanity.
No. Apple sees its SoC as a means to make the Mac better. Of course, “better” is a dangerously vague adjective that needs some evidence.
We’ll start with power dissipation. My MacBook Pro gets hot…really hot. Apple doesn’t specify the exact chip, but it appears to use this Intel iCore 7 processor that has a 28 Watt TDP (Thermal Design Power; think of it as a “power budget”).
Let’s compare this with the latest iPad Pros that feature an A12Z Apple processor.
According to Geekbench tests, A12Z performance matches or exceeds my MacBook Pro. Apple doesn’t disclose the TDP for the A12Z processor, but we can rely on an indirect number, the iPad Pro’s 18W power adapter output. This gives us an idea of what to expect from Apple Silicon in future Macs: Significantly lower TDP without losing processing power.
Next, throughput. Given what we see with today’s A12Z, one can’t imagine tomorrow’s Apple Silicon Macs providing less than a 25% throughput advantage against corresponding x86 PCs. Admittedly, these are speculative, broad strokes assumption for Apple Silicon Macs — think faster, svelter laptops actually lasting 10 hours on a battery charge. If not, once again, why bother burning the billions?
Next comes the matter of software, one that Apple took pains to handle at last month’s WWDC by demonstrating native versions of big standards (Microsoft Office, Adobe Photoshop…) and showing off the Rosetta 2 emulator. More concretely, Apple immediately shipped large numbers of its DTK (Developer Transition Kit) to help third-party developers port their apps. First impressions of the kit, which provides an Apple Silicon Mac prototype running an A12Z processor inside a Mac mini box, are promising: The hardware is fast and the software tools are more mature than one might have expected at this early stage.
We’ll know more when Apple Silicon Macs ship in a few months, but it appears that the hardware and software transition seems to have been carefully planned and executed.
So how will this affect Intel and the industry?
In 2012, Microsoft started the move away from Intel’s x86 processors with its first Surface machine running on an ARM SoC engine. It didn’t work too well. But Microsoft persisted and, late last year, came up with the Surface Pro X powered by another ARM-based SoC and running Windows on ARM. It was an improvement, but many reviewers still weren’t charmed. To cite but one problem, Microsoft’s bread and butter apps didn’t run in native mode. This was made even more embarrassing when Office on Apple Silicon was demonstrated at last months’s WWDC.
This leaves Microsoft with a choice: Either forget Windows on ARM and concede modern PCs to Apple, or forge ahead, fix app compatibility problems and offer an ARM-based alternative to Apple’s new Macs. It’s a false dilemma, of course. Microsoft will forge ahead…with repercussions for the rest of the Windows PC industry.
Specifically, what are Dell, HP, Asus, and others going to do if Apple offers materially better laptops and desktops and Microsoft continues to improve Windows on ARM Surface devices? In order to compete, PC manufacturers will have to follow suit, they’ll “go ARM” because, all defensive rhetoric aside, Apple and Microsoft will have made the x86 architecture feel like what it actually is: old.
This won’t happen overnight and there will be an interesting mess of x86 and ARM SoC machines fighting it out in the marketplace. Large organizations need continuity and would balk at the prospect of servicing two kinds of Windows machines and apps. As usual, they’ll downplay Apple’s advantage and curse Microsoft for causing trouble. But if the newer machines are actually better, rogue members within these organizations will sneak in new devices and software, they always do.
We now come to Intel’s reaction. Not what they’ll say when the trouble really starts, which could be soon.
Intel execs know they missed the Smartphone 2.0 revolution because of culture blindness. They couldn’t bear to part with the high margins generated by the x86 cash cow; they couldn’t see that lower margins could be supported by unimaginable volume. Now, Intel is facing a more serious problem: The x86 commands high margins not because of the chip, but because of the Intel/Windows duopoly, meaning that, all other thangs being equal, chips not running Windows get lower margins than an x86 CPU. Now, that union, that advantage is about to disappear. Intel will face ARM-based SoCs running Windows on ARM with applications, in PC-like quantities, at lower prices.
This leaves Intel with one path: if you can’t beat them, join them. Intel will re-take an ARM license (it sold its ARM-based XScale business to Marvell in 2006) and come up with a competitive ARM SoC offering for PC OEMs. Margins will inevitably suffer as the ARM-based SoC field is filled with sharp competitors such as Qualcomm and Nvidia, sure to be joined by arch-enemy AMD and others, all ushering in a new era of PCs.
I’ve avoided discussing Intel’s high-margin server chips and if/when/how they will be affected by more power-efficient ARM-based chips such as the AWS Graviton. I don’t know enough about the specifics of Intel’s monster (and electricity-hungry) Xeon chips powering forests of Cloud servers to form an opinion.
Wintel’s passing is a heavy-enough topic for today.
— JLG@mondaynote.com