Making the case for an ARM-based MacBook

Making the case for an ARM-based MacBook

Summary: While the new iPhone launch event left me lukewarm, the introduction of a 64-bit processor stirred the technologist in me. The new processor in the iPhone can easily power a MacBook in the not-too-distant future.

ARM MacBook
Image: James Kendrick/ZDNet

Like my colleague Jason Perlow, the unveiling of the 64-bit ARM processor in the new iPhone struck a cord with me. It wasn't an accident that Apple execs kept referring to the new processor as "forward thinking" and that it brings "desktop class" processing to the new iPhone.

Perlow penned a thoughtful piece about the future of Apple's products with a merger of iOS and OS X coming down the road. His view is fascinating and probably spot on. It makes sense for Apple to merge its two OSes at some point in the future. Perlow's article is worth a read for its insightfulness and you should stop right now and go read it.

See related: 2015: 64-bit ARM chips in iPhone 5S serve up taste of Intel-free future for Apple | Apple announces iPhone 5S | The iPhone 5S goes 64-bit: Will it matter? | Apple's iPhone 5S event: By the numbers | Apple reveals iOS 7 development status | Apple's iWork free on all new iOS devices | iPhone 5s with Touch ID is a big win for BYOD security | Apple launches iPhone 5C: Color, aimed at emerging markets

I agree with Jason in regards to where Apple is likely heading with its products. It makes perfect sense to merge both the OSes and the hardware lines as much as possible. The long-term roadmap that Perlow envisions is a solid one, and I wouldn't be surprised if he's right with this call.

Apple could incent developers to port their OS X apps to the new ARM Mac, and even iPad apps could appear on it.

I think there's a short-term scenario that Jason doesn't discuss that I believe Apple could tackle very soon. The 64-bit processor that's in the new iPhone is very powerful, easily powerful enough to run a MacBook. I wouldn't be surprised if Apple is already working on a full port of OS X to this new processor, which makes sense on many levels.

Apple could produce a MacBook Air-type laptop with the ARM processor onboard. OS X is based on a Unix core after all, and that could be ported to the 64-bit ARM processor without a huge effort. Microsoft ported Windows to ARM and that was a much bigger task than bringing OS X would be to ARM. Apple controls every aspect of the processor, OS, and the apps which works in its favor.

This MacBook could be produced more cheaply than Intel versions and avoid the standard heat issues for such a thin laptop. The battery life, admittedly good on the Air with the Intel Haswell processor inside, could be even better with an ARM processor. They typically run cooler and are power-stingy, both good things in a MacBook. Apple could make the ARM MacBook case out of plastic, er, polycarbonate, instead of aluminum which would lower the cost signficantly.

I believe Apple is already working on such a beast secretly, and could bring an ARM-based MacBook to market quickly. It could be cheaper than the Intel versions and become an entry-level Mac for emerging markets, students, and other segments where a cheaper Mac makes sense.

Apple could incent developers to port their OS X apps to the new ARM Mac, and even iPad apps could appear on it. That could mean the ARM Mac might have a touch screen, which would make sense with iPad apps arriving in number.

I can visualize a MacBook with the 64-bit ARM processor inside, and I believe it would be a killer laptop. Running full OS X, it would be a good first step on Perlow's path to merge iOS and OS X.

There is precedent for an ARM laptop. Google has already successfully ported Chrome OS to the ARM platform. The $249 Samsung Chromebook has been a top seller on Amazon for months, partially due to how well that OS runs on ARM. Microsoft has ported Windows 8 to ARM, although Windows RT hasn't been a big seller. That's probably due more to the perceived shortcomings of Windows RT than the ARM platform.

Why would Apple introduce a cheap ARM MacBook which would cannibalize its lucrative MacBook line? The company has demonstrated with the iPad and iPad mini that it is all about profits, not unit sales. An ARM MacBook could be a very profitable product line in this writer's view, and pave the way for Perlow's unified platform.

Topics: Laptops, Apple, Mobility, Operating Systems

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • OMG , the level of BS is SO big on this article..

    It's very powefull ? not even the A7x they will putt on the ipad will reach the performance level of a old i3 ULV. WAY, WAY , WAY behind a i7, at best it will have performance near the upcoming Atoms.
    • actually Intel will cover the range low power to hi performance b4 ARM will

      They would be better off to converge on Intel rather than ARM. This is not even a close contest.
  • Where is your proof?

    You make a claim that porting osx to ARM would be easier than porting Windows. You show no proof. Or will you just say UNIX and pretend that is somehow meaningful?

    Speaking of Windows, yes, I hope apple does this. Watch sales crash and returns spike when the 92% of people who run Windows with bootcamp or parallels switch back to better PCs.

    Speaking of poor sales, RT didn't sell well because it couldnt run legacy programs. Oh, just like an ARM powered mac.

    3 strikes, you are out James.
    • 92%? In your Wet Dream

      I'd say the majority of Mac users do not run Windows in any way, shape or form.

      And given the transition from PowerPC to Intel, Apple does have the technical chops to move to ARM. I was a Mac user during that time and Universal apps were great.
      • Prove it

        You have absolutely no knowledge and no proof of what you say. Just because YOU don't run this setup doesn't mean that other don't. Stop acting like an authority on something you obviously are not.
        • WTH?

          He was just replying to a comment that made an even more authoritative claim with absolutely no proof to back it up. Why are you attacking BruinB88 for saying that IN HIS OPINION the majority of users probably do not run Windows apps via Bootcamp, when toddbottom3 presented the claim that 92% of Mac users run Bootcamp as INERRANT BIBLICAL FACT?
        • Just my opinion

          There is NO WAY 92% of people run Windows on their Mac. I would bet 92% do not run Windows on their Mac. I'm always the one people come to when they need Mac help, and I know that almost none of them use Windows. The ones who have Windows on their Mac never use it and wouldn't care. I use a lot of esoteric Windows programs, and from about 2007-2010 I ran Windows on my Mac, but I almost never used it and haven't had it installed since.

          I don't know if any adaquetely sampled survey has been taken, but in another ZDNet article, they were surprised to find that 97% of Mac users wouldn't buy Windows 7. And 76% of Mac users would try Windows 7 if it were given to them free.

      • The problem is

        Apple went from the powerpc to intel, intel processors were more powerfull than the ppc, therefore rosetta would run with just a minor performance penalty. If you believe such a scenario is possible with the seriously underpowered ARM processors, you are well and truly dreaming.

        By the way, I would believe that the majority of mac users don't run either bootcamp or virtualising Windows to be higher than what you claim.

        That entry level macbook would truly be entry level, no current OSX applications would run, and the few that would get ported might have lousy performance.
    • I'm willing to bet

      that you pulled 92% out of thin air. Thanks to the internet, it's a known fact that 80% of figures are made up.
    • Technically it would be slightly easier

      Simply because Windows is a larger codebase, with more layers of compatibility. However, each has a fairly similar design... API abstractions riding above a kernel, which rides above a hardware abstraction layer.

      Dave Cutler designed Windows NT to be portable - it was originally written not for Intel at all, but for Digital's Alpha. But then, so did NEXT. When they got out of the workstation business, they made a big point of building the NS/Foundation layer (NextSTEP) so abstractly that it can ride different operating systems (Sun experimented with running NextSTEP on top of Solaris.)

      Interestingly, Apple once played with the notion of running it over top of the NT kernel, and I gather they got a prototype of it working.
    • Hello?

      I think you are forgetting a bit of history in that apple has always used a different chip then everybody else and that I'm sure they can resurrect an x86 to ARM version of Rosetta stone and since OS X is based off of BSD and BSD has ARM support, porting programs shouldn't be a problem or even maybe bringing back a type of "classic" mode for some light x86 emu.. Just for the record I love windows 8 and am no apple fanboy.. Just figured I'd point a few things out.. ;-)
      Nick Zamparello
  • The Truth about X86 versus ARM

    Here's an article that shows "Why you will probably never own a Mac with an ARM processor in it". It explains the real truth about ARM V.S. X86.

    Basically the article states that bit per bit, and clock per clock ARM doesn't even come close to X86.

    The only reason that ARM can keep up with Intel in smartphones is because your dumbing down the Intel processor by feeding it RISC. (Not taking advantage of Intel's memory modules and/or memory cache. In an ARM processor the CPU makes up roughly 80% of the processor and the rest is mostly memory modules and memory cache. In an Intel processor, only 25 to 35% of the processor is the CPU and the rest is mostly memory modules and memory cache. It's those memory modules and memory cache that give Intel its power.)

    The reality is the Baytrail and Haswell are the beginning of the end for ARM and next couple of generations of Intel Processors will further diminish the need for ARM.
    Don't fear the future
    • Interesting

      I don't see ARM going away, but intel will catch up lost ground for sure.
      While ARM can keep an advantage (even if small) on power/performance ratio, there will be a place on mobile devices for them - batteries are getting better really s l o w.
      From some reading I've made, I guess top ARM processor must be reaching i3 performance, while the latest i7 must be something completely different, i3 performance is still good enough for running windows (don't know about osx).

      I also don't see Apple dumping Intel for ARM. Android and chromebooks desktop PCs, are a much bigger threat to Intel if they don't react fast enough - I believe Intel is not sleeping and they are very capable. Also it's better for Intel to be an important part in android than be 100% (or at all) in Apple PCs.
    • Here's a comparison.

      Here is a review of the new Intel Baytrail Z 3770. The link takes you to the 3rd page of the review where they are doing benchmark comparisons between platforms.

      The SunSpider score when the Baytrail is running on Android is 531.6. The same test with the same processor on Windows 8.1 (X86) is 335.5. (Lower score is faster). Mind you, Android is a stripped down operating system, as is its web-browser, while Windows 8 is a full desktop operating system and IE 11 is a full desktop web-browser. Yet, the Baytrail is much faster on the more demanding platform: Windows 8.1 w/IE 11.

      Same thing for the IceStorm (3D Mark) test done on Android versus Windows 8.1. On Android, the Baytrail scores 8,409; while Windows 8 it scores 12,239 (higher is better). Once again, Baytrial is much faster on the more demanding platform. How can that be?

      That's because X86 is what Atom was designed to run on and the X86 platform (WIndows 8.1) truely utilizes the Baytrails architecture/language/capabilities.
      Hope this helps show what I was trying to say up above.
      Don't fear the future
  • I hope they don't take your advice

    I mean, you talk about how good the ARM Chromebook was. And it's decent if you just want to look at a few webpages. And it worked best when none of them had flash, because if they did you had a lot of page refreshes on your hands.

    People use their MacBooks for a bit more than that, though.

    And even if they did... why plastic? I like my aluminum MBA partially because it's that darn sturdy.
    Michael Alan Goff
  • It seems easy, but no

    an ARM macbook will face a lot of problems,
    I know that apple can port OSX easily to the ARM architect, but many problems are there :
    1. ARM processors can't virtualize intel architect in any level, it can't even virtualize itself. So you're stucked for what you have.
    2- Since you can't virtualize any other machine, and you can't install windows dual boot (maybe Windows RT but it's not for selling), you can't run windows. Only MAC OSX ported to ARM programs, and IOS programs.
    3- Most MAC users dual boot windows or run windows in a virtual machine to do common tasks, let's face-it, windows is more rich and powerfull than MAC OS, and have more quality software , without the ability of running windows, MAC is a very limited machine.
    • Well

      You might be able to use some Linux Distros...
      Michael Alan Goff
    • have any proof?

      have any proof on your point 3? I think many Mac owners may use Windows on their machines in some way... but I seriously doubt its most. You assertions about Windows being so much better than OS X is just plain false anyways.
  • benefits

    It's without question OS X can be ported to 64 bit Arm processor as UNIX is traditional a very portable OS with most codes in C language only those with low computer knowledge doubt so

    The question is more likely should they port it. What's the benefits? More power? More power savings? More software?

    Right now the advantage is for the lower cost and power savings
    • Power savings?

      Let me just say that there is virtually no reason to worry about battery life. The practically paper thin Macbook Air that came out a few months ago has at least 12 hours of battery life and most people are saying that's minimum, and they are easily getting up to 16 hours of battery life.

      Now OSX Mavericks which is coming out in a matter of weeks has like 35 power savings optimizations to achieve 20% power savings or something in these Haswells.

      Then Apple is switching to to igzo displays probably by the next Macbook Pro/Air refreshes, which use significantly less power. And on top of all that, there is billions of dollars being dumped into improving battery life all around the world right now so while they're not improving by leaps and bounds, it's pretty certain batteries will continue their march toward improved energy density. I'd say by next refresh, a Macbook Air will be running 20+ hours of battery life on normal use.

      I have 3 hours on my 3 year old Macbook Pro 15" and I almost never run out.

      Now, debating power of A7 vs i3... I have a Galaxy S4 with a quad core ARM chip and I guarantee my 3 year old laptop has at least 5x the power of my phone.