Apple developing new chip for Macintosh in test of Intel independence

“Apple Inc. is designing a new chip for future Mac laptops that would take on more of the functionality currently handled by Intel Corp. processors, according to people familiar with the matter,” Mark Gurman and Ian King report for Bloomberg. “The chip, which went into development last year, is similar to one already used in the latest MacBook Pro to power the keyboard’s Touch Bar feature, the people said. The updated part, internally codenamed T310, would handle some of the computer’s low-power mode functionality, they said. The people asked not to be identified talking about private product development. It’s built using ARM Holdings Plc. technology and will work alongside an Intel processor.”

“The development of a more advanced Apple-designed chipset for use within Mac laptops is another step in the company’s long-term exploration of becoming independent of Intel for its Mac processors,” Gurman and King report. “Apple’s first ARM-based Mac chip, known as T1, was introduced as part of the MacBook Pro update last October. This chip powers the Touch Bar — a screen built into the keyboard that lets users access app shortcuts, system settings and emojis — and some security features such as the enclave that stores payment and biometric data tied to the computer’s fingerprint reader.”

“Apple engineers are planning to offload the Mac’s low-power mode, a feature marketed as ‘Power Nap,’ to the next-generation ARM-based chip. This function allows Mac laptops to retrieve e-mails, install software updates, and synchronize calendar appointments with the display shut and not in use,” Gurman and King report. “The current ARM-based chip for Macs is independent from the computer’s other components, focusing on the Touch Bar’s functionality itself. The new version in development would go further by connecting to other parts of a Mac’s system, including storage and wireless components, in order to take on the additional responsibilities.”

Read more in the full article here.

MacDailyNews Take: Apple’s reliance on Intel – a smaller company worth significantly less ($173.07B) than Apple’s cash on hand ($246.09B) – to power their industry-leading Macs brings to mind these two quotes:

I’ve always wanted to own and control the primary technology in everything we do.Steve Jobs, October 12, 2004

• In order to build the best products, you have to own the primary technologies. Steve felt that if Apple could do that — make great products and great tools for people — they in turn would do great things. He felt strongly that this would be his contribution to the world at large. We still very much believe that. That’s still the core of this company.Apple CEO Tim Cook, March 18, 2015

SEE ALSO:
Apple’s A10 Fusion chip ‘blows away the competition,’ could easily power MacBook Air – Linley Group – October 21, 2016
Ming-Chi Kuo: Apple to unveil new 13-inch MacBook, 13-inch and 15-inch MacBook Pros at ‘hello again’ special event – October 22, 2016
What to expect from Apple’s ‘hello again’ special Mac event – October 21, 2016
What Apple’s new MacBook Pro might have learned from iPhones and iPads – October 21, 2016
It’s official: Apple sends invitations for ‘hello again’ event on October 27th – October 19, 2016
Get ready, Apple’s new Macs are finally set to arrive! – October 19, 2016
All-new MacBook Pro, refreshed MacBook Air and iMac, and more coming at Apple’s October 27th special event – October 19, 2016
Apple plans to launch new Macs at special event on October 27th – October 18, 2016
Apple’s A10 Fusion chip miracle – September 20, 2016
The iPhone’s new A10 Fusion chip should worry Intel – September 16, 2016
Apple’s remarkable new A10, S2, W1 chips alter the semiconductor landscape – September 15, 2016

58 Comments

    1. Steve Jobs didn’t like being jacked around by IBM, so he threw in with Intel. Tim Cook doesn’t like being jacked around by Intel, so he leveraged ARM. There’s much more to the story, but it’s essentially about who’s dragging whose ass.

      1. This is the problem I have with many posters here. They blame the slow Mac role out on Intel rather than Apple.

        In the last few years it has been extremely rare that Apple has shipped Intel based products as soon as the Intel chips were available in mass production. Sometimes Apple has been announcing then shipping in limited quantities Intel based products as much as 9-10 months after Intel has been shipping those chips in full production quantities.

        Besides, the biggest break on the PowerPC came with Motorola’s G5 chip. It was a screamer, those of us that could get our hands on the preproduction version loved it. Apple planned to use it even seeding prototype hardware and software to us. Then we all discovered a significant bug in the chip (verified by virtually everyone). Motorola unilaterally decided that the redesign of that chip would be too expensive so they killed the entire development and production of it. This left Apple hanging with an inferior next gen chip from IBM. IBM’s subsequent slow improvements in the PowerPC over the next couple of generations just compounded the problem prompting the eventual move to Intel chips.

        1. Yes that’s mostly correct, and the blame on Intel is slightly misplaced. However, Intel has been basically treading water with performance since the haswell chips, and there wasn’t a massive gain to implemting them right away. So it is a similar issue to Ibm over a decade ago, and since the A series chips are with the A10 already at the level of ivy bridge processors, the next generation could easily double performance again. They use less power, and performance per watt is always the best metric, as described by Steve in 2006. Essentially apple wants to control the whole widget, and f they can do something better in house than a partner can, they do so. So the slow rollout of New Intel chips in macs (except iMacs which seem to use the newest desktop chips right away, it got skylake at the end of 2015) can be blamed on both companies goals and performance with regards to generational improvements in chips.

          In short, if apple’s A series chips continue to increase in performance exponentially for the next several generations, there is no reason to use Intel processors if they continue to have modest performance gains of 1-4% every generation.

        2. It could also be argued, though, that since Apple has to wait for those Intel chips to come to market before they can put them in their machines, they are having to wait for it longer than if the machine was being designed side-by-side with the chip.

          Your PPC knowledge is greater than mine, but my understanding was that it was practically co-developed with Apple, and so there was that side-by-side development that worked more quickly.

          (None of this excuses the fact that other hardware developers get their Intel products out much more quickly. That isn’t my point, and it is a good one in your argument. But Apple makes sure it plays nicely with the whole system – all hardware and software environments imaginable. At least that’s the reason we like to think it takes them a bit longer to get to market.)

      2. I have to correct your cause-effect timeline, Herself. The iPhone was developed under Steve Jobs’ leadership and it used ARM from the start. The A-series processors went through a couple of generations prior to Steve Jobs’ death. So Apple has simply continued ARM development under Tim Cook.

        In addition, the original AIM alliance included Motorola. So Jobs was irritated at both IBM and Motorola for the slow pace of PPC development.

        I believe that this is simply a case in which Apple finally has the opportunity to take full control of its destiny with respect to CPUs. This required the convergence of several factors: (1) Independent foundries (TSMC, for instance) capable of producing microchips in large volume using the latest processes, (2) Viable, licensable CPU architecture (ARM), (3) Acquisition/development of microprocessor design group within Apple to push advancement and, (4) Product shipping in sufficient quantities to justify the effort (iPhone).

        When you add these pieces together, it seems likely to me that the iPhone will end up being the key to a whole new generation of ARM-based desktop and laptop Macs that combine tremendous performance and power/watt efficiencies with low cost CPUs achieve through the large unit sales of the iPhone and iPad. iPhone unit sales are roughly 15x Mac sales, and iPad unit sales are roughly 2.5x Mac sales. That is a lot of leverage for ARM CPU development.

        1. Thank you for supplying more details. I admitted up front that I was oversimplifying, did I not? The point I wanted to bring out — an emotional pivot point that drives risky business decisions — is that both men were ready to play hardball if their partners became lackadaisical. Apple doesn’t twist in the wind, but escapes with bold action. Changing processors can upset the apple cart in any number of ways, but they did it with ingenuity and confidence. And with mastery.

        2. Agreed. Apple as a company has successfully executed multiple transitions of CPUs and operating systems. No other company can claim the same.

          However, I must contest your assertion regarding “oversimplification.” It is one thing to lose the nuances of a situation via simplification. It is another to alter the facts. You asserted that Tim Cook got tired of being jacked around and left Intel, and that is not true. Not only did Apple not leave Intel from a Mac standpoint, but it was Jobs that drove the expansion into ARM for iOS devices.

          There is simplification…and then there is fiction. And we have enough fiction already on this forum and in our country.

        3. Correction: you did not say left Intel, you said leveraged ARM. Big difference. But the other error still applies. We need more facts, more open debate, admission of errors when they exist, and more harmonious discourse. I would suggest that this thread is what MDN is about, and the threads hijacked by the political ranters are not.

        4. Thanks, KingMel. I’ll try to do better. I agree that this thread is what we’d prefer that MDN be about. Still, one oughtn’t be quick to dismiss the investor viewpoint that powers this website and invites politics in as an inescapable adjunct to commerce. If Apple is one thing today, it’s one of Capitalism’s better success stories, no longer a club for tech enthusiasts.

        5. Now that is well said!

          Snowflakes like KingMel, Predrag, Breeze, et al don’t get what you just described.

          Take a bow for your laser sharp insight … 🌹🌹🌹🌹🌹🌹🌹🌹🌹🌹🌹🌹

  1. It’s really an exciting story. I know how the tech industry hates the Touch Bar on the MBP. They claim it’s not innovative enough. However, if Apple is simply using the TB as a test-bed for a specialized chip that’s great news. I personally like the idea of a Touch Bar but everyone says a touch-screen is better. If Apple can use an ARM chip to take on extra low-power duties, that’s better value than a touch-screen for me.

    Oh, and even if Apple wanted to buy Intel, I’m sure Intel wouldn’t sell the company. There’d also probably be some monopoly issues. Apple buys Intel and sells no more Intel processors to any company selling consumer desktops. Apple would corner the home computer market.

    1. Good post, but Intel is a publicly traded company. Apple can buy Intel stock on the open market, if it so chooses, although I believe that Apple would have to disclose when it reached 5% of ownership. That approach would also be viewed as a hostile takeover, and the stock price would escalate as shareholders attempted to make Apple pay.

      1. Intel has failed in the low-power mobile CPU space. Billions of portable computers modeled after the iPhone run on ARM or Samsung Exynos or Snapdragon CPUs. That dwarfs the number of Intel-based PCs in the world, and the refresh cycle is much higher.

        The Celeron was not the answer. The Atom was not the answer, except for the netbook niche, while is also largely a failure. Even major server farms may soon go with ARM and other processors over Intel CPUs. China has the fastest supercomputers running on its own processors. So, yes, Intel is scared and it very well may be in trouble over the next decade or two. Because new fab facilities cost billions and Intel is hampered in R&D because of its x86 legacy that must be maintained. Meanwhile, Apple has churned out ten plus generations of its A-series CPUs and they keep improving rapidly.

        The A11 and A11X will be the turning point for the next generation of Macs.

      2. There’s a lot more to Intel’s troubles than the A series. Heard of AMD? You might want to give it a look – they are anticipating announcements in the next few weeks and months that are expected to blow away Intel’s work over the last few years.

    1. Absolutely. Intel knows the x86 is dead (they publicly said so) Microsoft k owe the x86 is dead and are devoloping both arm based windows (again) and cloud based windows, and Apple knows the X86 is dead.

      Google could care less as it doesn’t need them.

  2. Apple looks for options & options it controls.

    It also looks for situations where it can pressure suppliers into better delivery, terms, etc.

    My guess is that this “project” was leaked deliberately by Cook.

      1. you’ve got nothing but a losing lifestyle.

        This had nothing to do with you and everything to do with Tim’s ventures away from running a company to running his mouth.

        racist?
        You belong to the party of the KLAN you idiot.

        And I seldom if ever post anymore.

        1. TT, you are getting worse every year. I hardly recognize you anymore. We used to have some common ground. But all I hear now are the ranting of a deluded partisan extremist with serious emotional issues.

          You are not aging well…

  3. Apple plays the long game. I smiled when Apple bought chip design capability many years ago. Apple is like a chess master playing against kids who are playing checkers — PC makers, mobile phone industry, Microsoft, Samsung, even Google. Perhaps _now_ some people (analysts?) will begin to see it. (But I doubt it!) Anyway, Apple put this into play many years ago.

  4. If Lapple switch to the RMA chip 1- it won’t be able to ran windows. 3- the ARM chips is slower 😧 B- they will have the monopoly because they are going to be soul supliers of Spple components. Remember when Microsoft got in trouble when they pre-packaged internets explore with Window and cut out the competition? Does Apple need to be the next Microsoft.

  5. It is inevitable that Apple will release an ARMs based laptop that can run a variant of MacOS. Don’t expect this to be an instant switch. There are plenty of users that still need to run windoze. My guess it that there will be a Air-type machine that has 40 hours battery life and can run both MacOS and iOS apps.

  6. …is another step in the company’s long-term exploration of becoming independent of Intel for its Mac processors,” said Mark Gurman and Ian King who report for Bloomberg.

    NO it’s not.

    I wish tech journalice and analcysts would get computer science degrees before blethering on in public about utter nonsense.

    Dear Disagreeable People: Rather than rant your ignorance at me, please read and learn, on your own, DIY, that sort of thing, for a change, exactly WHY Apple will stick with Intel for Mac and NOT go ARM. This will get you started:

    https://en.wikipedia.org/wiki/Complex_instruction_set_computing

    https://en.wikipedia.org/wiki/Reduced_instruction_set_computing

    You’re welcome.

      1. A combination of Intel CISC with ARM RISC?

        No. Try again. Think of all the proprietary code that’s specific to Intel’s CISC platform. Now think of all the developer code that is directly dependent upon that code.

        Excellent example, one that people like me who are forced to work cross platform think about every day: Virtualization. That’s GONE on ARM. That’s bad. That’s very bad. That’s a deal breaker.

        1. The lines have become somewhat blurred between RISC and CISC architectures in the last decade or so. Intel’s CISC chip architecture has incorporated some RISC elements for at least 15 years. Even Apple’s A series architecture is not 100% purely RISC. In the coming years it maybe be possible to write hardware abstraction layers (the metal code) for each chip architecture that the execution code can sit on top of. I wouldn’t be so quick to rule out this possibility. Certainly, your battery mileage and speed would vary depending on which chip you would use.

          The other option, which I’ll grant is more likely, is that Apple’s consumer line of Macs migrates completely over to the A series and loses the ability to boot into Windows. Apple’s Mac Pro line would get both chips with an Apple designed controller that is able to seamlessly switch use of processors depending on the needs of the software without the need for a reboot.

        2. Intel’s CISC chip architecture has incorporated some RISC elements for at least 15 years.

          Yes, of course. But the RISC hasn’t by any means replaced the CISC. That means the drawbacks of dumping Intel x86 architecture chips remain.

          Now, could Apple gradually wean developers off the CISC and direct them to use RISC architecture that’s also included in future Macs? Yes! That would mean a more friendly, gradual transition. That could happen. That would be innovative. I’ve never considered that possibility until now, with Apple adding their own chips into Intel Macs. Hmm!

        3. Honestly they don’t care Derek. Virtualization is like any other high end professional requirement in the Mac space and Apple has demonstrated quite effectively that if this is your need, go PC. This includes gaming. Apple is a high end consumer information appliance and services company, not an IT company per say. The days of wanting PC parity are over.

          Now. Care to wander over to the Razor web site with me and the tiny few who need more than just appliances?

        4. That hurt my feelings a little bit, Thelonious.

          But I can’t disagree. 🙁

          I want a good, powerful Mac. I don’t care if it runs Windoze software. I have Windoze machines for that.

          There. I’ve said it. You’re right. I don’t care. But I recognize and appreciate that some of you do! I still believe that moving away from Intel makes a lot of sense, but not at the expense of those of you at the very top end of the market and in development of those things that make the Mac great.

        5. Do you mean razerzone.com? I haven’t kept up on who’s making the killer gaming PCs.

          I’ve been using Windows software as long as I’ve had Macs, initially via emulators and as of 2006, Intel’s 64-bit CPUs with hardware enabled virtualization. I can use Boot Camp if I need something more than virtualization offers.

          I wouldn’t exactly say the days of wanting PC parity are over. There continue to be inroads to enabling enterprise grade software on Macs. But clearly Apple focuses on iOS in the enterprise at this point.

        1. this is why i and i suspect many others have long read mdn, for contributions like these, from people who know their onions about computers.

          and even though some of the issues are out of my league, i still benefit.

          now contrast this sort of informational exchange with the political rants of any number of other contributors that pollute the waters here and act as partial deterrent to reading these threads.

          time for men to exercise some degree of discretion and start filtering out the blatantly political, and often quite crude ravings of a vocal minority.

          please.

    1. Derek,

      There are more fundamental reasons for the foreseeable future (3-5 years at least).

      Think of the I/O if nothing else? Compared to the Intel CPUs how many PCIe lanes does the Ax series support? Compared to the Intel CPUs what kind of QPI does the Ax series support? Compared to Kaby Lake and currently in development Cannon Lake and successors how many ThunderBolt 3 interconnects does the Ax directly support? Compared to the Intel CPU what kind of vector math does the Ax directly support? The list goes on and on.

      Could Apple bake all those things into the “A13m” or “A14m” (“m” for Mac variant of the A series CPUs)? Yes. However, it would make those chips as different from the current Ax series chips as those are different from the current Intel CPUs.

      Besides, how many “CPU” or “CPU-like” or unique support processing chips (like the M series) do people think Apple can do effectively? Apple’s focus is NOT being a chip house. With the A series, the M series, the T series, the W series and the S series, Apple probably already has its hands full. Those former Powerficient guys are impressive, but you can spread them only so thin across just so many projects.

      Yes, Apple could make an “Intel CPU killer”, but will they? Possible, but unlikely for several more years to come.

      1. No Your over thinking all this .
        Apple are making a chip to do the everyday lifting stuff
        Apple will pair this with something like the AMD RYZEN chip to do the heavy lifting. There is a way for Apple to have both chips in different options as the transition continues going forward.
        The word is that RYZEN is a whole new game and AMD are Back with a vengence!

      2. Excellent points! It is so nice to have someone add to a subject around here, as opposed to ranting ignorance.

        I’m looking forward to how Apple augments their Intel boxes with their own chips. Seeing as Apple is already making their own GPU chips for iOS devices, wouldn’t it be incredible if they made their own GPU chips for Macs as well? How ambitious would they be with those chips? Would they dare dump crummy Intel graphics? I personally can’t figure out how Apple has been so incredibly cheap and lazy as to stick their customers with Intel graphics for all these years. That sort of crap should have died on Macs back when they all went 64-bit in 2006. – – But I rant.

        Conclusion: Potentially great stuff to come with Apple adding their own chips to Intel Mac boxes. I like it.

    2. OS X used to run on RISC chips in the early days.

      Yes, that has been many years ago, and hardware has shifted with the chip change, but the fundemntals were always there.

      It would take a lot of work to even get back to where it was then, but who is to say that they may not have already been working on it?

      Apple has made transitions smooth in the past. Remember System 7’s rollout? Then the PPC came shortly thereafter. Lots of concern, very few actual issues.

      OS X itself was a huge departure from the norm. Cocoa made it work smoothly. And the shift to Intel happened soon thereafter.

      We already have a new OS on a new chip that has worked flawlessly. (iOS and A) I don’t see why the folks at Apple wouldn’t be able to pull this off again.

      Sure, software would need modifications. But Apple knows that, and has done it before. And lots of mods have already been made for iOS – it isn’t like it would be a whole new process.

      1. There is no problem moving OS X itself to RISC chips. Apple did that with iOS, which is a subset of OS X. But there’s quite a lot more of concern than simply porting OS X back to RISC.

        I’ve done my part to point out these concerns. So I’m not going to go over them yet again. All I can keep saying is that there are critical barriers that aren’t going away. To ignore those barriers is, ahem, not advisable.

    3. Every time there’s an article related to this your on here with your rant, which is completely unrelated to modern times.

      Code compiled with Xcode is compiled down to LLVM bit code. It can be recompiled in real time for the appropriate processor – any processor, be it CPU, GPU, DSP, ISP, etc. This allows for maximum efficiency.

      So any “code” that Apple would want to hand over to an ARM CPU running in low power mode versus an Intel CPU, can simply be recompiled and targeted for that ARM CPU.

      This technique has been used by Apple for many years now, starting with GPU code. If a specific GPU did not support a certain function then the code was re-targeted for the CPU to handle the processing.

      If you want to learn something more about modern compiler techniques, there’s a great 2 hour chat with Chris Lattner – the guy who started the LLVM project and created the Swift language.

      1. I’ve had thoughtful conversations about this with other members here. I’m not going to get back into the ignorance warz again. You’ve got some information of interest but appear to be deliberately missing the core points of the problem. It’s not going to happen. Go ask Chris Lattner, who knows orders of magnitude more about the problems than I ever will.

  7. This is an interesting development, but not in itself a threat to Intel.

    It’s a second, ultra low power processor which comes into it’s own when the main processor is essentially turned off. Conceptually it’s rather like the motion coprocessor ( often referred to as M7, M8, M9 and M10 ) in iPhone 5S onwards and newer iPads too. It efficiently handles simple tasks ( accelerometer, gyro, compass and listening for Siri commands ) while the device is otherwise asleep.

    This new chip could sit alongside Intel processors indefinitely, but I have little doubt that Apple will be releasing ARM powered MacBooks sooner rather than later. However I don’t see it as an either/or choice, both can co-exist and find their own markets. An ARM powered MacBook could usefully employ the T310 chip too in exactly the same manner.

Reader Feedback

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