bostonidealist's Journal: OS X On ARM

On Apple's January 2013 quarterly earnings call, CEO Tim Cook stated, regarding âoeâ¦somebody who buys an iPad mini or an iPad [as] â¦their first Apple product, we had great experience through the years of knowing that when somebody buys their first Apple product, that a percentage of these people wind up buying another type of Apple product. And so, if you remember, what we had termed the halo effect for some time with the iPod with the Mac, we're very confident that that will happen and we're seeing some evidence of that on the iPad as well. And so, I see cannibalization as a huge opportunity.â

In September of that year, Apple announced a new 64-bit A7 processor, which was introduced in the iPhone 5S and was subsequently included in the iPad Air and iPad mini with Retina display that debuted the following month. Apple's public relations copy consistently touted the A7 as a âoedesktop-class architecture.â

Beginning with the A4 processor in 2010, Apple has released a new custom ARM-series CPU each year in its flagship iPhone and Cook has repeatedly indicated that Apple is working on new product categories that will be available in 2014. While there is a lot of warranted speculation around payment processing, wearables, and television products, which Apple is certainly pursuing, there are indications that Apple may be planning to dramatically shake up its Macintosh computer lineup as well, perhaps by building new OS X devices around its custom ARM-based silicon vs. Intel's CPUs.

Apple is facing threats on multiple fronts in the sub-$1,000 computing marketplace. There are now many competing computing devices available in this space, and there are indications that Apple's iPad sales are slowing as alternative devices, many with lower price points, attract potential customers and existing iPad owners remain complacent and uninterested in rapid upgrade cycles. Despite Apple's clever marketing and the best efforts of many third parties to augment the device's limitations, the iPad was designed first and foremost as a media delivery vs. general computing device. While this led to a streamlined interface with a high degree of ease-of-use, this design is turning into a liability, as it becomes more difficult for existing iPad customers to perceive any advantage in the purchase of a newer model.

Apple knows that while its iOS users have the highest level of customer satisfaction in the mobile industry (70%+ for iOS vs. under 60% for Android and all other competitors), many of these customers remain frustrated with their general computing devices. The problem is that while many prospective customers might be attracted to OS X devices, they're put off by the high price of entry. The cheapest OS X notebook is $1,000, while competing devices begin under $300. Additionally, the lowest-priced OS X system, the Mac mini, starts at $600, requires significant additions for most users (monitor, keyboard, mouse), and has low enough margins that Apple typically excludes it from its employee discount program and other promotions.

While this is purely speculative, Apple could release sub-$1,000 OS X devices running on ARM CPUs in the near future. In particular, an A8-based Mac mini replacement and A8-based notebook seem like a good fit and would bring Apple several advantages.

First, Apple would control its own CPU design for all of its sub-$1,000 products in this scenario. The Apple TV (if it continues in its current incarnation), iPod, Mac mini, iPad, iPhone, and ARM-based notebook would all be based on the same CPU architecture. This would reduce Apple's dependence on Intel.

Second, while recompiling OS X and its own related applications for ARM would be straightforward for Apple, all third-party applications would also need to be rebuilt for the architecture, and Apple could leverage Gatekeeper and the OS X App Store to control distribution of third party software and gain commissions on third party software sales and subscriptions. Customers on the whole wouldn't care about losing the ability to sideload; iOS users would be accustomed to the App Store model, wouldn't have an existing library of OS X software if they're moving from Windows, and would be exuberant about sub-$1,000 Apple notebooks.

Because even the most optimistic projections for the processing capabilities of the A8 wouldn't put it in the same league as higher-end Intel processors and existing OS X users may still want the degree of flexibility the Intel architecture affords, Apple could continue selling Intel-based OS X systems above the $1,000 threshold. The only immediate loss to customers would be the Intel-based Mac mini, but Apple could easily downplay that, as the system hasn't been updated since 2012.

The timing for ARM-based OS X seems ideal. The existing A7 is already a capable processor, and the A8 will likely be even more so. Intel has achieved amazing things with its Haswell CPUs, and many powerful competing devices based on the architecture are available that threaten to lure users away from Apple. The previous power consumption advantages of ARM over x86 are dwindling given Intel's remarkable improvements, so this is one of the last opportunities Apple will have to distinguish battery life advantages of ARM, before Intel's 14 nanometer die shrink in 2015.

The fact that the iPad mini jumped several CPU generations in moving from an A5 to an A7 in the Retina iPad mini last year is further evidence that Apple wants to consolidate its architectures for the long-term, especially since Apple must have known that such a significant CPU upgrade is lost on the average consumer. As Apple converts more of its devices to 64-bit ARM, it can simplify development tools and gradually push ARM CPUs up the product line as the architecture becomes more powerful and third party applications become more available.

Earlier this year, Apple's Senior Vice President of Worldwide Product Marketing Phil Schiller and Senior Vice President of Software Engineering Craig Federighi adamantly professed that working on merging iOS and OS X would be a âoewaste of energyâ and is a âoenon-goal.â While this is true, it's not just for the user experience rationale Apple usually puts forward. Apple is much more deceptive than it lets on: it knows that people want both touch and keyboard + mouse interfaces, but why sell each customer one device when you can sell two?

Apple's switching OS X to ARM will have lots of implications, including the possible loss of Boot Camp and Thunderbolt on non-Intel systems. Apple will stress cost, battery life, theft prevention through Activation Lock, and many other benefits to customers for newer ARM-based OS X systems, and some of these will be real advantages. However, if the above predictions materialize, particularly locking out alternative operating systems and software distributed outside the curated App Store, Apple's removal of the user-accessible Gatekeeper switch wouldn't be about improving security or software quality, it would be exclusively about breaking the 30 year legacy of the Mac as a flexible computing platform so that Apple can enforce totalitarian control over the platform at the expense of its users.

Consumers and the press should be on alert and should look critically upon such a development, but a shiny $700 Apple notebook with 24-hour battery life might prove too appealing for people to care.