Another reason Apple may make its own Mac chips

“Here’s another reason why Apple may consider making its own chips: as noted by MIT Technology Review, Intel has disclosed in a regulatory filing last month that it’s slowing the pace with which it launches new chip-making technology, potentially abandoning its “tick-tock” model,” Dennis Sellers writes for Apple World Today.

“Currently, Macs already go two to three years between major updates. This could stretch out even longer now,” Sellers reports. “If Apple isn’t happy with that, it may decide to make its own chips (which I think it will anyway).”

“The company describes the A9X chip in the iPad Pro faster than 80% of the portable PCs shipped over the last 12 months in CPU tasks and 90 percent of portable PCs in graphics performance. You can be sure that A10X and A11X processors are in the works,” Sellers reports. “Apple certainly has the money to make its own chips [for Macs].”

Read more in the full article here.

MacDailyNews Take: It’s only a matter of time.

[Thanks to MacDailyNews Reader “Lynn Weiler” for the heads up.]

36 Comments

      1. I wish I could still run WriteNow. Loved that program.

        I hope if Apple plans this that they give developers a very advanced roadmap. Nothing running on X86 is going to run an A chip variation. I really don’t relish a multi-year transition where half the software I need is either unusable or running in some form of emulation.

        1. Spark, you must be new to Apple. Apple has gone through this transition before. They know how to make it smooth and nearly seamless.

          However, I completely disagree with the MDN take. It’s only a matter of Windows in the enterprise. IMO. Macs will continue to use Intel chips as long as enterprise continues to be mostly Windows. While the Mac is making great inroads, enterprise and government is still run on a majority of Windows machines, thus Macs in the enterprise need to be able to run Windows, virtual or natively.

          Apple will switch Macs over to their own processors when Windows is no longer a factor in enterprise and government computing. Nothing sooner. iCal me. If they wanted to they could make the switch today by creating a massively parallel A9 Mac with 6 or more of cores. But this accomplishes nothing while Windows is still a factor, because Windows will not be able to run on such a machine with acceptable performance.

          1. This might speed up if all email, databases, business systems go to web based solutions, so that the client OS is irrelevant, but that is a big IF.

          2. Ahh… Owned about 200 Macs and other Apple devices since 1984. I’ve been through it all. That’s why I am familiar with WriteNow!, an early and very elegant word processor for Macintosh. I also remember the switch from Power PC to Intel and running a bunch of apps under Rosetta during the transition to OS X. Companies I rely on, especially Adobe, are not quick to update their software. Apple has done a great job in mitigating the transition problems, but that doesn’t mean that there are no problems that users have to live through.

    1. And you think that Apple will be able to get around the reason that Intel is going to a three step process rather than the historical two step (“tic-tock”) process?

      Everyone (yes, literally everyone) is having difficulties at 14 nm no matter what anyone is saying. (Yes, it is a lot of marketing hype by Samsung, TSMC, Global Foundries, IBM, and the rest that they are not having significant problems at 14/16 nm. And their claims of having already ironed out all the bugs at 10 nm is pure BS.) ALL the chip fab houses are going to have even more problems moving to 10 nm (which is why Intel is going to a three step process). ALL of the chip fab houses are going to have even more problems going to the 7 nm step. (I actually expect it to take 4 years to go from full production on 10 nm to full production at 7 nm — not the three years Intel is currently predicting.)

      Apple’s Axx chips are going to be no different. Apple cannot get the chips built at smaller feature size than Intel can do no matter who they try to get to do it.

      Yes, the current A9x chip might be equivalent to Intel’s i3 chips with an internal Intel graphics capability, and in some limited cases even break even with Intel’s i5 chips. But no one in their right mind is going to claim that the A9x chip is equivalent to a current i7 chip coupled with a state of the art AMD or Nvidia GPU. Anyone making such a statement just proves they don’t know what they are proclaiming.

      So, bringing Mac chips in house by using an AXx variant is just not realistic for a few more generations. You’ll either need emulation (as was done in the 68k to PPC transition — which only worked because the PPCs were so much faster than the legacy 68k chips) or through dual binaries (which only worked for the PPC to Intel x86 transition because bloatware was not as rampant as it is today). Trying either today will be much more difficult than it was for either of those transitions. To make people do it again, the AXx chips will need to be head and shoulders above even the i7 (or the future equivalent) chips.

      1. Good points. It’s what those folks who’ve been talking about Apple transitioning to Ax CPUs keep forgetting.

        Sure, which is not to say Apple WON’T do it… but weighing the cost/benefit of doing so doesn’t make sense for Apple right now.

      2. I totally understand where you are coming from, but I have to believe Apple isn’t going to try and support 5+ or so year old software. I think they will be designing these chips looking forward. No legacy type stuff to support that may hinder them. I also have to believe their chips will have the benefit of being specifically designed for MacBooks just as the Ax chips have been for iOS devices. I don’t claim to be an expert on this, but I just think if Apple does this, we’ll definitely see actual usage improvements even if the spec sheet doesn’t live up to intel’s. Just like all these android devices cramming in overpowered hardware to compensate for the crappy android software….

    1. Apple already uses AMD graphics chips and cards in a high percentage of its Macs. How is buying AMD going to change that? Not buying AMD gives Apple the flexibility to use either AMD or Nvidia. Buying AMD would virtually kill and chance of using Nvidia. (And there are certain market segments where Nvidia chips and cards are significantly better than AMD chips and cards.)

  1. Beginning to think that this is indeed inevitable though in what time scale or form I’m still not sure. It may simply be iOS devices becoming more pervasive into Mac territory at its simplest and slowest transformation. Otherwise how they deal with compatibility is the big question so won’t be any time soon if it goes beyond that, but sooner or later the disadvantages are going to be outwayed by the advantages.

    1. Apple should definitely become more vertically integrated.

      But chips are another matter. In order to be successful doing that, it has to have the knowledge, be super efficient, ultra responsive to the consumer, and superior in quality compared to the current supply base. Does Apple have enough chip fab experience to be competitive? Does current Apple management even listen to engineers anymore?

  2. “Intel has disclosed…that it’s slowing the pace with which it launches new chip-making technology…”
    That’s good for Apple. Apple can’t keep up with the Intel chip releases now.

    1. Hopefully that statement was sarcastic.

      The reality is that Apple CAN keep up with Intel’s chip schedule. Apple just *chooses* not to do so. Many of Apple’s current Mac products are a full generation behind what Intel is currently shipping.

      1. There is no way Apple can keep up with Intel’s tick-tock cycle. Intel themselves can’t keep up with the pace they set for themselves.

        Increased complexity means more time between releases. Look at OS X releases. Of course now Apple can throw more resources at it today, but Jobs set a torrid release pace that hasn’t been seen since. Apple had to back off before the programmers died of exhaustion.

        Now it looks like Apple is attempting annual freebie releases that have no relation to Intel’s processor schedule at all. Unfortunately, with all the gadgets and gimmicks Apple has thrown onto OS X, such a schedule has resulted in a string of amongst the three lowest-rated OS versions in Apple history.

        Thanks MacTracker!

        OS X public beta – Sept 2000
        10.0 Cheetah – Mar 2001, 6 months later
        10.1 Puma – Sept 2001, 6 months later
        10.2 Jaguar – Aug 2002, 11 months later
        10.3 Panther – Oct 2003, 14 months later
        10.4 Tiger – Apr 2005, 17 months later

        — Intel announces a tick/tock processor release schedule —

        10.5 Leopard – Oct 2007, 18 months later (tick)
        10.6 Snow Leopard – Aug 2009, 22 months later (tock)
        10.7 Lion – July 2011, 23 months later (tick)
        10.8 Mtn Lion – July 2012, 12 months later (tock)
        10.9 Mavericks – Oct 2013, 15 months later (combined iOS & OS X team – now freeware)
        10.10 Yosemite – Oct 2014, 12 months later
        10.11 El Cap – Sept 2015, 11 months later

        IMHO Jaguar, Tiger, and Snow Leopard were absolute standouts in quality. Why? Because Apple took the time to do them right. Today quality is shoddy because someone at Apple thinks that screwing three women is a great way to produce a healthy baby in 3 months. Every bastardized OS X since Snow Leopard has really sucked.

        Take your time and do it right, Apple. And make it lickable again.

        1. HP and Dell keep up with the Intel tick-tock.
          Apple charges a premium price for their computers. Historically, that was because they used leading-edge technology. Now, they are delivering 2 to 3-year old technology and still expecting us to pay the premium price tag.
          Apple is driving the value of its Macs downward. Sales are up but I suspect that is just momentum. The continued neglect of the Mac line will reverse that trend. Tim Cook is cashing in on the good thing that Apple had while he dismantles it.

          1. I agree absolutely. I have wanted to buy a Mac but won’t pay today’s prices for yesterday’s technology.

            The Mac seems to reverting to what it used to be – expensive, underpowered and old.

    2. George is right. Apple can’t or won’t keep up with Intel’s roadmap, which is well known by everyone.

      Intel has the best process technology in the world. Apple could never keep up.

  3. This is all really silly. Intel said nothing about slowing down its pace of development. All it said was that it’s moving away from the tick-tock model, specifically they stated, ” Intel announced that it had deprecated the Tick-Tock cycle in favor of a three-step “process-architecture-optimization” model, under which three generations of processors will be produced with a single manufacturing process, adding an extra phase for each with a focus on optimization.”

    By the way, this isn’t some decades-old model Intel is abandoning. They didn’t come up with this strategy until 2007. It served them well, but moving forward from here, there’s more that can be done optimizing at each die shrinkage level rather than continuously shrinking, and never fully optimizing at that point.

    Look at Intel’s roadmap and you’ll see they’ve got plenty coming up in the desktop and notebook space… notably Kaby Lake is due Q3 of this year.

    So many people confuse Intel’s missing out early on the move to mobile as a sign that those ahead in mobile can overtake in desktops and notebooks, but really, not only are they not even competing against Intel in that space, but Intel has jumped on mobile and is making huge progress in catching up and passing in some respects.

    Don’t be surprised to see Intel much more involved in Apple’s iOS devices, and don’t hold your breath for Apple to replace Intel in OS X devices.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.