ARM takes on Intel in supercomputing

“The designer of the chips that run most of the world’s mobile devices has announced its first dedicated processor for use in supercomputers,” Jamie Condliffe reports for MIOT Technology Review.

“The British company ARM Holdings, which was recently acquired by the Japanese telecom and Internet company SoftBank, has announced a new kind of chip architecture dedicated to high-performance computing,” Condliffe reports. The new designs use what’s known as vector processing to work with large quantities of data simultaneously, making them well suited to applications such as financial and scientific computing.””

“Intel will be worried by the purchase. The once-dominant chipmaker missed the boat on chips for mobile devices, allowing ARM to dominate the sector. But until recently it’s always been a leading player in the supercomputer arena,” Condliffe reports. “Now the world’s fastest supercomputer is built using Chinese-made chips, and clearly ARM plans to give it a run for its money, too.”

Read more in the full article here.

MacDailyNews Take: Intel’s getting slammed on all sides.

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

19 Comments

  1. This little spark may be the beginning of the Big Bang – transition from energy hogging multicore x86 workstations to massive parallel Ax/ARM based Apple’s Computing Centers.

    Your next Mac Pro will be your iPhone / iPad + account Apple’s cloud based supercomputer.

    You can get as many cores + RAM + GPU power as much you want, from every Apple device. You just send your data and instantly get your results, no matter how strong is your computer at home.

    And no more bloody mess with new / incompatible processors, sockets, connectors, motherboards, RAM etc.

    Apple transformed communication, video and music industries.
    Now its time for personal computing industry.

    I’m pretty sure that future “sans Intel” plan is planned but I’m afraid Tim “No revolutions, please!” is afraid to make it happen.

    He is still the HP minded guy.

    1. I was with you up until the part about Tim not wanted it so every complains about Macs and the fact is Intel is largely to blame. Look at the fiasco surrounding Surface with power management etc. Intel is starting to hold Apple back.
      The planning to switch to ARM for laptop is well underway I suspect and as the A series gets deeper into desktop territory, I think Apple will show their cards for future desktop computing.

    2. Cloud computing is not the cure for all Ills. I do not want to push 12-15 GB of MPEG 2 data up to a remote server to transcode it into h.264 MPEG 4. That is something a Mac Pro handles very well without swamping my internet connection.

      1. Agree with this.
        I suspect that Apple is as prepared as anyone else to shift to cloud computing, but won’t push for that until the infrastructure bandwidth is much, much higher.

        Besides, don’t the power users complain about the Mac Pros because they want everything in one box, instead of connected by a (Lightning) cable?

        1. Yep it is, on paper, a wonderful concept which is why it has been championed for such a long time. However there is a long way to go till it reaches the light of day on a large scale. But both Apple and Google are clearly aware that is an amazing opportunity to displace the embedded companies that tie us to the PC world, even now so strongly. I am sure ARM is well aware of that too and while this move isn’t directly related, it will be the perfect showcase for them to demonstrate how their architecture(s) can be scaled to take advantage of mainstream computing as and when freeing itself from the Intel hegemony through cloud computing where being dominant in mobile can only be to its advantage once the everything in the box concept is made redundant. At least for the overwhelming majority of computing. Ironically the one place that would resist the most would be at the Super Computer end, so doubly smart move if it comes off.

        2. My MacPro is a real Mac Pro Tower. I consider the Trashcan a Mac Mini Pro HTPC. It is interesting as an engineering exercise, but is the answer to a question nobody was asking.

          Apple needs to either make a real Mac Pro again (internal storage with user upgradeable components) or license mac OS to third parties and get out of the Hardware business at the high end.

          At some point in the not too distant future I will have to retire my Mac Pro and buy a Windows Machine because Tim Cook is more interested in lobbying, taxes and social issues than making computer hardware.

          I have been using Macs since 1984 and it would be sad if I have to buy a Windows tower because Apple doesn’t give a shit about the Mac anymore.

    3. I agree with GeoX. Apple is planning some kind of transition away from the typical desktop architecture. Either that or they are just lazy. And all evidence suggests Apple is NOT lazy. Could they be looking to transition to an all-memory based architecture utilizing GBs ot memristor or Micron 3D Xpoint memory and abandon the hard disk/SSD entirely?

    4. Tim Cook worked for IBM and Compaq, not HP. But I think I know what you mean…that he is timid where Steve Jobs was bold, or aomething like that. You may be right, but I think Tim is committed to the vision and the plan laid out by Steve, and isn’t really dragging his feet so much as that some of his suppliers are dragging theirs.

      1. BS. Apple isn’t even innovating in simple ways that their competitors have been. Macs are 2-3 generations behind in CPU and more importantly GPU. Apple still sells 5400rpm hard drives, 2-core processors, and insanely overcharges for RAM and SSDs.

        Okay, so the narcicisstic Social Media generation doesn’t care.

        Fine, let’s forget about actual processing speed and look at user convenience. Is Apple cutting edge? Hell no. Several laptops I know have fingerprint sensors for unlocking. iPhones all rely on them too. Touch ID still hasn’t arrived for Macs.

        Why? Because Apple thinks people are going to pay >$300 for an Apple Watch to unlock their Macs for them? Which of course relies on iCloud, which is Apple’s implementation of subscription computing, which is reprehensible to anyone who even occasionally needs to work offline to do their jobs. Believe it or not, most of the world is not covered in WiFi.

        1. yup.. i give apple another 10 years in the personal computer market before they final just give up and make hand held toys for tots and nothing else.

          in 10 years they’ll be even more behind the curve…

            1. I remember the argument shifting from “don’t use Macs because they are toys” to “don’t use Macs because they don’t have as many games”.

        2. I agree with all of your points. Apple’s roadmap, based on a futuristic vision, is not one any of us would have designed, because we have to live and work in this world, never mind the next. We are pissed because the Mac has become an obvious afterthought to them. But it is what it is, and I gave up trying to prod them. They definitely seem unfocused with respect to the products we care more about, specifically the Mac workhorse. Yet they are, really, keenly focussed — on other special points, circled in red on their secret roadmap, that we could care less about. The crucial clue to their real focus is in their fanatical pursuit of thinness, of miniaturisation, of simplifying away component structures, of reducing user interface elements to invisibility – taking Think Simple to its logical extreme. All of these point to a belief in the ascendance of the internet of things – embedding smart chips in every object from cars to broom handles – making everything a sealed appliance. Very like the original Macintosh computer.

          1. IoT (Internet of things) requires a method or methods of communication between all the bits and pieces that is secure, reliable, functional and user-friendly. We have neither at this point. Apple is clearly striving at both. If they can pull it off first AND get broad adoption, that will give them one excellent avenue into the future. Their initial experiment is the Watch, which qualifies as a sophisticated IoT device as well as a hub for other IoT devices.

            What continues to hold back IoT at this moment is the proliferation of interesting hardware with NO security. The result is a shameful pile of dangerous junk. The priority of security has yet to penetrate the skulls of most IoT developers.

            Then there’s the curse of Bluetooth. We’ll see what Bluetooth 5 provides.

  2. No ARM powered laptops/desktops until Swift gets broader adoption in Windows programming. Once you have broader cross-platform support for major apps coded using Swift, only then would Apple make the jump for its laptop/desktop processors. And everyone else follows. There is an opportunity here for a computer maker to champion a Swift based OS and its associated App Store.

    1. I don’t understand your point about ‘a Swift based OS’. Why? A programming language does not an operating system make.

      I haven’t read any recent (2016) news about Windows compilers for Swift code, but they exist. Compilers allow Swift code to run on various operating systems.

      One of the difficult parts is integrating OS and CISC CPU specific APIs into one’s Swift code running on ARM-based RISC CPUs. This is probably the biggest reason we’re not going to see Macs run on ARM RISC chips. Then there’s the fact that ARM-based RISC chips don’t (yet) compete in speed with Intel’s CISC chips.

      In any case, Swift is still in development, isn’t finished. Even version 3 arrives, it will have a long list of further features to develop as a language as well as further OS and hardware compatibility features. The Swift project has several email lists relevant to different aspects of Swift development.

      https://lists.swift.org/mailman/listinfo

Reader Feedback

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