Will Apple move away from Intel?

There is an appealing idea which has been floating around in the internet for a while: Apple could transition from Intel x86 processors to ARM-based CPUs designed by Apple itself.

I think that this moment has finally arrived.

In a certain way this transition sounds to me like the classical tale of disruptive innovation. An emerging technology is not suitable to be used in an existing market because its characteristics are inferior compared to those of the current one. However this new technology might introduce features which makes it suitable for a secondary market. In this secondary market the technology is further developed and thanks to its newness generally there is room for substantial improvement. As the new technology matures it becomes suitable also for the primary market. Moreover the improved characteristics surely make it an optimal choice in the secondary market. At the end, the combination of those benefits could lead the new technology to definitively displace the current one of the primary market.

Need an example? There are fairly good chances that you have one in your hands. If you are reading this article in a laptop consider that its technology was originally designed for the niche of mobile computing. Only at a later time laptops rose to become the most common form of computing. Otherwise if you are reading this article in a smartphone or a tablet, consider how mobile devices are displacing computers.

The A series SoC

Along with the iPad, Apple introduced the “A” series family of SoC (System on Chip) which includes an ARM based CPU to power its mobile devices. “A” series CPUs are optimized for mobile devices and trade off computational performance in favor of power efficiency, smaller dimensions, and an architecture without fans (which are noisy and are a point of failure as they are one of the last moving part of modern computing devices).

What is striking for me is that Apple has been working a lot on the computation performance of its CPUs. Recently, the performance of the A series chip is starting to be good enough for consumer computing: the iPhone 7 A10 chip is faster than some Macbooks currently on sale.

The iPhone 7 is also faster than the iPad Pro. This is interesting because the iPad Pro features the A9X chip, which derives from the A9 found in the iPhone 6s and is roughly 25% more powerful because there is more room than in the iPad Pro. It is possible therefore to imagine that the A10 CPU found in the iPhone 7 could be optimized for laptops and desktops in a similar but more prominent way in which A9 was improved to create the A9X for the iPad Pro.

This would translate into CPUs with performances good enough for the general consumer, which uses the computer mostly for browsing the internet and running office applications, and therefore could enjoy the benefits developed for mobile devices.

Why I think the time has come now?

“A” series CPUs look ready and Macs haven’t been updated in ages. If you have any doubt that something is going on, just look at the MacRumors Buyer’s Guide.

Advantages for the customers

The advantages for the customers would be the following:

• longer battery life
• thinner computers
• no fans
• a completely silent system
• cooler laptops

Besides what I have already said, there is another intriguing idea: the support of macOS on the iPad Pro. This would allow Apple to make a tablet a la Microsoft’s Surface Pro. I’m quite sure Apple would consider this option only if they can find a twist which would allow them to market their solution as completely superior. I think that there is room for Apple to do this. A 13" inch iPad Pro is basically a Macbook Air without a keyboard and a trackpad. Microsoft’s proposal is a good implementation of this idea but it is not flawless yet. Specifically I’m referring to the keyboard attachment design, which is good but still a bit cumbersome, and to the less-than-elegant handling of the coexistence of touch and cursor based interfaces that characterizes Windows.

I think that this is the future for Apple. I don’t see iOS and macOS as two separates operating systems in 10 years from now.

Pros for Apple

• Unique competitive advantage. Having a custom processor would help to further differentiate the Mac in the personal computer category. This would be a very big barrier to overcome for competitors as a huge size is required to develop a custom processor.
• Huge economies of scale which Macs could take advantage of. More than 1 billion of iPhones have been shipped.
• Unification of the computing architecture. Standardization would allow to greatly concentrate the energies currently used to develop parallel features in macOS and iOS.
• Total control of the technology. Apple has always aimed to have a complete control of its technologies. This kind of control allows to introduce features which are difficult for the competitors to reproduce. For example Apple introduced the Apple A10 Fusion processor with the iPhone7. This processor has 4 cores: two guarantee high performances and two are energy efficient. Apple uses a high performance core only for instructions which require computational power. This greatly increases the battery life of the iPhone. Apple can decide when to use the high performance cores because they control the operating system and the tooling used compile the applications. Different pieces of the architecture need to be tuned in order to extract the most of the gains from this feature. Currently there is no other competitor in such a position.
• It is feasible to change the architecture and Apple has already done a similar migration with success: in 2005 Apple switched from the IBM PowerPC architecture to the Intel’s x86.

Cons for Apple

• Applications would need to be ported to the new architecture. However, most of the Apps have been ported to iOS so the majority of code optimized for the Intel platform has been adapted to take advantage of the A series processors features. Moreover a translation layer, which would allow the code to run in a different platform with a performance hit, could be implemented. Apple has already done a lot of related work with a technology called bitcode that they introduced in 2015.
• Prosumers want a lot of power. From the point of view of Apple the advantage in the consumer segment would be more than just offset the dissatisfaction of the prosumer segment, which nowadays is very small in the company customer base. The company has already proved that it is comfortable in accepting this trade off with other choices in the past.

To mitigate the cons apple could also opt to ship its CPUs along Intel processor. This could be done only in the prosumer Macs.

Conclusion

On the 27th of October we likely will discover if I’m correct.

If my predictions will reveal to be untrue, I would hope that an innovation with similar magnitudo will be introduced for the event. Otherwise it would be hard for me to find a reason why Macs have not been updated for such a long time. Actually it would be very fun if the event of 27th of October is not about the Mac but about Apple Watch bands.

If my prediction will reveal to be true, in one way or the other, this strategy will make one of the most interesting case studies of innovation strategy for business schools around the world. It would mean that Apple is cooping with what in management is called the Innovator’s Dilemma.