Why Apple Won’t ARM the MacBook
Real World Tech’s David Kanter has posted an eloquent, detailed, and thoroughgoing analysis of why the rumor buzz launched last week by the SemiAccurate site to the effect that that Apple would abandon x86 and migrate all their notebooks (MacBook Air and Pro) to ARM silicon over the next two years doesn’t hold water, so to speak, observing that it would cause a massive disruption in the PC ecosystem, and calling it an exceptionally unlikely scenario.
I agree. Kanter examines several reasons Apple might have for switching to ARM CPUs in their notebooks. One would be enhanced potential for convergence of the iOS and OS X platforms toward an eventual merger. Another would be tighter control over hardware – an abiding Apple obsession – and potentially greater integration of hardware and software engineering, also consistent with Apple tradition. Apple would find it much easier to steer the direction of ARM to suit its needs than is possible with Intel, and as a bonus save money by designing and manufacturing their own ARM chip.
Kanter also notes that Apple has switched CPU platforms twice before and mamaged it reasonably elegantly both times, and conceivably could pull it off again, but allows that studying the history of Apple’s hardware choices and their approach to switching platforms actually this time reinforces the unlikelihood of an x86 to ARM migration, noting that from a technical perspective, the performance and compatibility barriers are huge.
However, Kanter contends that the most formidable obstacle to such a switch is would be performance, noting that the MacBook Pro especially is intended for performance hungry professional applications, and ARM has nothing anywhere near as powerful and refined as Intel’s quad-core Core “i” Nehalem, and that there are no ARM microarchitectures even on the horizon that can compare to Intel’s Sandy Bridge or AMD’s Bulldozer, with current ARM designs are at roughly the same performance level as x86 was in 2000, and suggesting that migrating to ARM would not just require matching, but exceeding x86 in performance. Then there’s Intel and Apple’s recent joint announcement of the Thunderbolt I/O interface, with Apple likely intending to consolidate and replace multiple I/O interfaces (e.g. USB, Firewire, DisplayPort) with a single Thunderbolt port, and Intel having little motivation to license the patents to ARM.
He also explores a number of business reasons why Apple switching its laptops to ARM makes no sense, and summarizing that the strongest argument for Apple sticking with x86 is that it meets their needs quite well, and maintaining that while it’s an intrigueing thought experiment, Apple will not be switching from x86 to ARM for notebooks in next few years, and venturing that it would not even be possible, but suggests that a plausible scenario would be for that Apple to develop some sort of hybrid system to enhance areas like boot/wakeup performance and facilitate more iOS/OS X integration, and 5-10 years from now, who knows?
Great article! You can check it out at:
http://bit.ly/kG4jFQ