Broadwell and ARM Macs
I wrote last month about Broadwell’s delays and the expected impact on the release of new Macs. It appears that is coming true. The recently updated MacBook Pros just received slightly bumped Haswell chips and, as Macworld reports, the speed improvements are small.
Moreover, I’m more convinced than ever that the rumored 12" MacBook Air with retina screen, assuming it exists, will get pushed back until next year when Apple can get the Broadwell chips it needs to put a retina screen in a small MacBook. If Apple were to release a Haswell-based MacBook Air with retina screen, I’d recommend waiting.
There is also more buzz about the idea of an ARM-based Mac. The ARM chips that currently power iOS devices are Apple designed and Apple controlled. If Apple could put those in Macs, they wouldn’t be dependent on Intel for future Mac releases and wouldn’t get saddled with the problems they have with the current Broadwell delays.
The problem is that ARM chips aren’t nearly as powerful as these Intel chips and it would incur a substantial performance hit. Another downside of ARM Macs would be that they don’t run Windows nearly as easily as Intel based Macs do. (However, I have to wonder how important that is as we increasingly move to web-based services and Windows becomes less relevant.)
On the plus side, ARM based Macs would have ridiculously great battery life. When you think about it, a thin, light Mac that sucks at Final Cut but runs Safari and Mail for 24 hours on a single charge may have a pretty large audience. If Apple were to go this route, I suspect that initially they would keep producing high-end Intel Macs for people that need the power.
Yesterday, Jean-Louis Gassée (who knows more about this stuff in his pinky finger than I do in my entire body) wrote that he believes an ARM based Mac may very well lie in the not so distant future. One of the points he made that hadn’t occurred to me is that since Apple is designing the chips, they could create a separate ARM design for the Mac that is a bit more powerful and uses a bit more power. Pound-for-pound though, I suspect Apple would have a hard time matching Intel on the power end, especially now that the Broadwell chip is on a 14nm dye.
Could something like this be already in the works at Apple? To answer that question I’d state that just a few months ago Apple announced an entirely new programming language for the Mac and iOS that they’d been internally developing for years and nobody on the outside had a clue of its existence.