The Intel Problem 


With lots of chatter this week about Apple and Intel, the costs of Apple switching to ARM have been on my mind.

I firmly believe Apple has ARM-powered Macs running ARM-ready OS X in a lab somewhere. This isn’t a technical problem; its a philosophical one.

With the platform transition from PowerPC to Intel, Apple had Rosetta, a software layer that let code compiled for PowerPC Macs run at neat-native speeds on Intel machines. It was an impressive feat, and one that Apple could certainly pull off again if OS X were to transition to ARM. However, I think there are three big reasons the Mac is going to stay x86-based for the foreseeable future.

Performance

When Apple announced its move to Intel in 2006, Steve Jobs said:

When we look at Intel, they’ve got great performance, yes, but they’ve got something else that’s very important to us. Just as important as performance, is power consumption. And the way we look at it is performance per watt. For one watt of power how much performance do you get? And when we look at the future road maps projected out in mid-2006 and beyond, what we see is the PowerPC gives us sort of 15 units of performance per watt, but the Intel road map in the future gives us 70, and so this tells us what we have to do.

There’s no doubt this topic is still on Apple’s mind, but comparing iPad-class ARM chips and Mac-class isn’t exactly Apples-to-oranges, as Siracusa pointed out on last night’s ATP.

The iPad Air has a 1.4 GHz A7 chip coupled with a 32.4-watt-hour battery that provides 10 hours of battery life. The 13-inch MacBook Air ships with a 1.4 GHz Intel Core i5 chip and a 54‑watt‑hour battery that provides 12 hours of battery life.

While the clock speeds are the same, the MacBook Air is computationally more powerful (roughly twice as powerful, according to tools like Geekbench. The iPad Air is a good bit more power efficient than the MacBook, but that difference isn’t a good talking point with such a wide gap in raw performance.

(When thinking about ARM-powered Macs, many think of the MacBook Air not only because it’s become the default Mac, mindshare-wise, but that the thought of an ARM-powered iMac or Mac Pro seems laughable.)

In short, Apple’s ARM-based A7 isn’t a good choice for a MacBook Air at this point. While I don’t think it’s outside the realm of possibility for Apple to ship a slower notebook if the tradeoffs were right, but that current performance gap isn’t enough to justify the possible gain in Cupertino’s beloved performance per watt metric.

Virtualization

One huge gain Apple made when switching to Intel in 2006 was the ability to natively run Windows. While Apple certainly made fun of the Microsoft operating system (like in the ad above), the ability to run OS X and Windows side-by-side is a huge selling point for power users.

For example, I use modern.IE virtual machines to QA websites at work. If I were to be running an ARM-based Mac, I wouldn’t be able to do this. As much as Apple might like to think its customers can be OS X-only, I’m not alone in these needs.

(Assumedly, Windows RT could run natively on an ARM-based Mac, but that’s not a robust enough solution for people who need real Windows apps.)

Product Confusion

Of course, the first two reasons can be brushed aside by a simple statement:

Apple could keep Intel chips in the MacBook Pro, and use ARM chips in the MacBook Air.

I think that line of thinking has some serious flaws.

Apple likes simplicity in its products and its product lines. Selling one family of machines that can’t do what another does isn’t a very Apple-like thing to do. A Mac mini can do what a Mac Pro can do, even if its not as fast. How does Apple pitch a slower, less-featured MacBook Air?

Now what?

For these reasons, I don’t think we’re going to see an A7 or A8-powered MacBook Air anytime soon. The Broadwell delay is annoying, but not a big enough reason to do something as crazy as moving away from Intel. Cupertino might not like spending its time waiting on new silicon from Intel, I don’t see what choice it has.