If Windows 10 on ARM Works Properly, It Will Be Boring

Posted on June 2, 2017 by Paul Thurrott in Windows 10 with 121 Comments

If Windows 10 on ARM Works Properly, It Will Be Boring

While there’s understandably been a lot of excitement around Microsoft bringing full Windows 10 to the ARM platform, there’s also not much to say about this project. If it works properly, it will just be Windows 10.

As you may recall, Microsoft announced Windows 10 on ARM last December at WinHEC, but it’s been pretty quiet it ever since.

Windows 10 on ARM was not featured at all during Build 2017, for example, though Microsoft released a quickie pre-recorded video to silence the questions. And then this past week, Microsoft revealed an “Always-Connected PC” initiative, which includes ARM systems as well as those based on the traditional Intel x86 platform.

This week, Qualcomm piped up and discussed Windows 10 on ARM during Computex. The company literally provided no new information, but that’s because there is nothing to say: Windows 10 on ARM is just Windows 10. There will be Home, Pro, and Enterprise versions, just like on Intel.

Microsoft’s Terry Myerson did discuss how Always-Connected PCs would interface with the built-in eSIM in such systems at Computex too. But again, that isn’t just for ARM, it’s coming to x86 PCs too.

If you’re really bored, you can watch an extended demo of Windows 10 running on ARM. That, again, shows off nothing new at all.

And that, folks, is the point.

If Windows 10 on ARM works, and I think it will, the average user won’t see any difference between that product and Windows on x86. The systems could potentially be smaller and lighter, and get great battery life, and they will include integrated cellular data capabilities. But that’s why someone might choose such a system: For the benefits, not for the underlying architecture.

Windows 10 on ARM probably won’t work well for high-end tasks that require powerful Intel hardware like video editing, game playing, and the like. But then, neither will most thin and light PCs, regardless of the architecture.

Put simply, Windows 10 on ARM will be successful if it’s boring. If there is literally nothing to say, or at least anything unique. If this thing arrives and just works, Qualcomm and Microsoft win. That’s everything.

So here’s to boring. I hope it just works.

Tagged with ,

Join the discussion!

BECOME A THURROTT MEMBER:

Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Register
Comments (121)

121 responses to “If Windows 10 on ARM Works Properly, It Will Be Boring”

  1. Polycrastinator

    Well, if connected standby works right and the integrated cellular connection is good, I'll be excited for that. But I see what you mean.

  2. Waethorn

    I'd rather have a pocket mobile router than have to pay for internet access on every device separately, thanks.

  3. Waethorn

    So here's a summary of their demo:


    Office for desktop, which ran on a Tegra 3, 5 years ago.


    Websites, which every computing device can do.


    One utility application, 7Zip.


    And 4K video, which even a lowly Chromecast Ultra can do.


    You're right, Paul. This is boring. What will be interesting it to see how the WOAOW32 or whatever the emulation layer is called affects battery life and performance over comparably-priced x86 machines. I can't imagine that flagship ARM chips are much less expensive on mediocre x86 chips that outpower them.


    • SvenJ

      In reply to Waethorn: The argument with that list is that they have demo'd more than Office at other venues. Photoshop and other x86 applications did not run on Tegra 3.5 years ago. Not even 7Zip. That, and the state of the Windows app store at the time, was the issue with RT. The app store has improved some, or a lot, depending on who you are, but that's not a factor anymore, now is it.


      • Waethorn

        In reply to SvenJ:

        The Photoshop demo was fake. You can put Photoshop on a Core i7 with an m.2 SSD and it won't load up that fast. Office loaded fast in that demo too, and it's now running slower in this one, so the original video was staged.

        • skane2600

          In reply to Waethorn:

          Yes, there's a long history of fake demos. It's only when you have a device in hand and you use it awhile do you find out what its capability and performance is. The fact that the demos where 7 months ago and there's still no specific products announced suggests that they aren't even done yet. They might be taking a WINE-like approach: finding out which major applications don't work and kludging in work-arounds (Just speculation and I know WINE uses a different technical approach).

  4. nbplopes

    Hi Paul, I don't like the point of view of Boring vs non Boring. I think it is obvious that MS tech marketing along years have been anything but Boring. Furthermore I think that point of view is what in part lead to the position that MS is now. The idea that Apple was all about Marketing, the rest was fluff. The idea that Google is all about stealing peoples information to feed marketeers, the rest was fluff. So on and so forth. This was very much engrained in the Windows culture.

    Excitement can only last so far. The more you steer it up, things might look good in the outset, but also caves a bigger hole if one does not deliver. At one point either one delivers something close to the excitement or the thing starts cracking down. Furthermore if you keep on pumping up to cover the cracks, meeting expectation becomes even harder and a point of rupture will happen.

    Bringing Windows to ARM looks to be much just a continuation of Microsoft strategy, Windows Everywhere from back in 1998. Windows Universal Platform only bring homogenous development artifacts on what what otherwise required disparate if not independent software development efforts. 1998, almost 20 years ago!!!

    What for me is quite clear is that Windows Everywhere failed. It worked for 10 years, under the premiss of lowering user training costs, after all if everything works the same or extremely similarly the less it takes to learn. The downside is such that if working similarly is at the core independently regardless of the fundamental value proposition of using the tool, than its value is gone. Why use it, just because it looks and feels similar to something else? To achieve similarity one needs to put constrains in the device regardless if it is the best option in context, regardless if its is different approach provides a better end-to-end experience. This bit by bit chips the value and hurts the solution to the point that it does not solve much really.

    If Windows had human threats it would be very egocentric. So much so that even Office 365 business was getting hurt when it was used as a lever to push Windows users do adopt Windows Phone.

    But the world as changed. Has humans got more used to computing systems the benefits of Windows Everywhere from the point of usability faded to the point that actually started creating more problems than solving. That is, the Windows Everywhere bubble was burst from a usability perspective and people got contact with a different sky, a sky that they prefer and whose benefits in comparison are very much tangible (pardon me the metaphors). And guess what, people seam to be able to deal with multiple OS's offering a different look and feel quite well.

    People no more look at a device and think, Does it run Windows? People look at theses devices and things, what does this thing solve, is it better from what we have or not? If the thing solves challenges in a very convincing way that otherwise do not really work well, not even close, wether if it is Windows, Android, iOS, Chrome, OS X, Linux or something else it does not really matter as much.

    From an outsider it looks that Microsoft never allowed the engineers to forget about Windows and totally focus on the problem to solve. Even in situations that it might have happened, it looks like that if it was found interesting either it was abandoned for later Windows integration in some other way, or would pass through the Microsoft Windows Machine to plug in Windows Patterns to make it look and feel like Windows, a self convincing exercise voiding the initial benefits. Just speculating don't know.

    The core question that I ask with Windows on ARM for consumers is ... "What is the core problem does it solve for the user?" I mean even if Windows 10 on ARM works flawlessly since inception, which is quite a long stretch considering that statistically we would consider otherwise, still this question is not answered at all?

    Compared with Windows 10 on Intel will Windows 10 on ARM:

    • Have better battery longevity?
    • Have better graphic power in a slimmer form factor?
    • Will it get rid of the fans and still offer powerful computing capabilities close or better than anything else on ARM?
    • Will user be able to do more in some scenarios, what scenarios? .
    • Something else?


    This is a question that I believe that MS needs to provide an unequivocal and convincing answer, no "imagine". Otherwise it will be just another technical drift, I'm afraid. I'm also afraid that the benefits Windows Everywhere boat as sailed in the consumer space. Microsoft failed to prove the advantages of it to the consumer. Worst, as things move forward across the big 4, MS is failing to show synergies between Windows 10 devices beyond "cuteness". For instance, XBOX One streaming to Windows 10 is nice, but no one really cares. I'm sure some people might career, but its not really much of value. The same thing between Windows 10 Phone and Windows 10. People see that they can have an Android, iOS or even Playstation along with a Windows 10 Laptop or Desktop and actually do more, much more than simply having Windows Everywhere as a tool.

  5. ncwmail

    I predict Windows 10 on ARM will be almost as exciting as Windows RT.

  6. Vuppe

    Yup. The software should fade away. This is exactly what it should be.

  7. JoshWright


    I know I will get shouted down for saying this but I think the development is both exciting and important. And before anybody asks, or infers, YES, I have been drinking.

     

    My understanding is that this will enable laptops with exceptionally long battery lives. On media releases and other sites, I am reading 'beyond all day' and '50% increases' of battery life as compared to current devices. I have a Surface Book and a Surface Pro. They are both very powerful computers in their own right and I will do my medium and heavy-duty work on those machine. I would be keen on a SD-835 powered tablet/ laptop/ phone not for the computational prowess but for the battery longevity and potentially small/ light form factor.

     

    Also, if eSim means that I can data share between my phone plan and a PC data plan, well that is a great thing too. Again, my understanding is that not only will eSim allow me to data-share but will actually clone the my phone number to multiple devices. That's pretty good if you want different phones around the house or different phones on different days. In future, it means that the phone that would otherwise sit in a top draw now acts as a second, or third, phone.

     

    So what else might this mean? I can't see a downside. It may mean migrating SD-835 such as Samsung S8 users to dual-boot or from Android to Win 10. It may mean attracting developers to the platform. It will mean a different class or device. This are all benefits to my (inebriated) mind.

     

    I also think that this is a positive step towards the holy grail of a 'single device', be it a surface phone, cellular PC or whatever you want to call it. I would actually love to see a version of Windows 10 that could be installed on existing SD-835 phones, such as a Galaxy S8. At least give people an option to switch over or better still an option of dual boot.

     

    I am constantly amazed by Paul’s understanding and analysis of Microsoft. I agree with almost everything he says or writes. I know that this is wishful thinking but I want a Surface Phone or Cellular PC.

    I am a long-suffering Windows Phone/Mobile user and am more than happy with my choice of the Lumia 950XL, warts and all. I am also resolved to the fact that they will take a Windows device from my cold dead hand.  Until this announcement, I was sitting, in despair, contemplating the death of the Windows Mobile platform. I am not enough of a pompous elitist moron to buy an iPhone but nor was I keen on an Android but I would wait to see what the Samsung Note 8 had to offer. This Windows on ARM development will breath new life in to the platform. The two-year contract on my current Lumia runs out in November this year. I will now wait and see what replaces the HP Elite x3, Lumia or what any other provider has on offer.

    I know that the Announcement is aimed at laptops and tablets but there are obvious benefits for the Windows Mobile.


    So, to me, this is both exciting and important but maybe that is just the beer talking.

    • lvthunder

      In reply to JoshWright:

      You must be drinking if you think the cell phone carriers aren't going to charge you extra for the device. That's why most people will prefer to hotspot their phone to get data to a PC if there is no Wifi.

      • JoshWright

        In reply to lvthunder:

        I think it will also be helpful when 'connected cars' become ubiquitous.

      • JoshWright

        In reply to lvthunder: It depends on where you live in the World. I am in Australia and have a shared data plan over multiple devices. I doesn't matter if I am using a shared Android tablet, my Lumia 950XL or my wife's iPhone, the data is pooled and charged to a single account.
        An eSim would mean that they are still tied to a single account but I could 'log-in' to an old Lumia 920, which is in a top draw and 'log-in' to an old Lumia 1020, which is in the glove box of the car. I don't expect that eSim will be backwards compatible and cover those specific models but people buying phones on contract generally accumulate them at a rate of one per two years.
        The eSim concept will beneficial, but not essential, for the uptake of a Snapdragon run laptop. Many people would have just tethered if from the phone to get the same result in data usage and linking it to the same phone bill. The phone carriers are not worse off by making this seamless. If carriers do not offer this on unlimited basis, they will actually be better off for selling bigger data plans.


  8. chrisrut

    It's a very important next step - these always connected plus great-battery-life devices. Count me in. But one thing: I will be very curious to see what "always connected" is going to cost...



  9. dhallman

    Do you know what's not boring? This could run on a phone! **Cues heavy bass dance music for happy dance ♪♫♬♪♫♬♪♫♬ But don't call it a phone. Microsoft can't sell those.

    • jimchamplin

      In reply to dhallman:

      But why would you? Due to the thermal constraints, it would have to be clocked down considerably. Due to physical limitations the battery wouldn't be very large, and runtime would suffer. Running legacy code will require CPU overhead and you'll lose more speed since due to the thermal issues you can't run as hard and get as hot. Even if that wasn't an issue you still have to remember that trying to use old-ass apps on a tiny screen will make you want to put your own face into a wood chipper.

  10. hrlngrv

    I'm odd, and I know it, but my main concerns are floating point math and database processing throughput. I figure ARM should manage the latter about as well as Intel/AMD CPUs. It'll be interesting to see how well ARM handles floating point. A few decades ago, that wasn't a RISC strength, so separate FPUs were popular/necessary.

    The other thing I'm very curios about is how well Windows for ARM handles emulating Intel/AMD CPUs. I wonder how it'd handle MSFT's own Systinternals tools.

  11. edboyhan

    I think it will be boring, and the performance degradation (if any) will have more to do with the relative performance differences of the underlying H/W ISA's (Intel Core vs SnapDragon 835) than anything else. After all Photoshop isn't likely to run very well on a Core M or Core I3 either :-) .

    The issue of emulation IMO is somewhat overblown. After the desktop Bridge session at Build I got to talking with a couple of the MS guys. Emulation usage will actually be pretty rare. All of the OS already runs native on ARM due to the WinRt OS development done previously. Managed C# apps run through a JIT interpreter which also runs native. The issue with win32 apps on ARM has always been with the API's. Back in the day the winrt API (the underpinning of most of UWP) was pretty sparse. Today it is vastly greater, and with the Desktop Bridge and Visual Studio enhancements, many apps will run native with little or no effort.

    Emulation will tend to be rare and involve edge cases which don't happen that often -- at least that's the theory. We shall see.

  12. Jorge Garcia

    Google just needs to hurry up with their alternative to a Windows, so the world can start moving away from the past. ChromeOS isn't it, because it will never properly run android APK's. A near-windows-desktop-level version of Android would be a good interim product while project fuchsia is fully hashed out and developed. How come Samsung can see this need (DeX), but Google cannot?

    • skane2600

      In reply to Jorge Garcia:

      Project fuchsia's purpose is still a bit vague and it might eventually find itself joining a long line of Google abandonware. Wikipedia describes it as RTOS, which is an OS type not typically associated with a general purpose computer.


      Then again Wikipedia could be wrong. It describes two kinds of RTOS, hard and soft. In the old days we'd call what is now described as hard Real-Time-Operating Systems just Real-Time-Operating Systems and what is now called soft Real-Time-Operating Systems as just Non-Real-Time-Operating Systems. But I guess being an RTOS sounds cool so everybody wants the label. It's the OS equivalent to "All hat, no cattle".

  13. Locust Infested Orchard Inc.

    To be so dismissive of Windows-on-ARM, as so many have done in the comments below as has Paul Thurrott himself (who certainly ought to know better, and in most likelihood does, but has been fiendishly courted by Google by way of its Nexus/Pixel XL and Android), is staggeringly contemptuous and not being mindful that to be able for Microsoft to execute x86 applications on an always-connected small factor device with cellular capabilities and twelve-hour plus battery life, is the holy grail of productive computing on-the-go.

    The video that is hyperlinked in Paul's piece shows Windows along with some elementary applications (7-Zip and Microsoft Office, and in a previously released video back in December 2016, Adobe Photoshop CC 2017 too) executing without issue on an ARM chip. For those who are so engrained with their enduring blinkered love for iOS and Android, one cannot see anything to get the pulse racing, hence the reference to "boring" in the article's headline.

    If and when a 6.2" device running Windows-on-ARM, e.g., Samsung Galaxy 8 Plus is made available (which most certainly is not beyond the realm of impossibility as it possesses the critical Qualcomm Snapdragon 835 SoC), that would most certainly get eyeballs rolling and ruffle much more than a few feathers in both the Android and iOS camp.

    With various patents Microsoft has been granted in relation to foldable mobile-like devices, the family of Surface devices appear to have several new little siblings in its wake not before long.

    With Microsoft working on a unified adaptive shell to span across all Windows devices of varying sizes, Composable Shell or "CSHELL" as it is referred to internally, it will be able to scale in real-time between all types of devices. A combination of CSHELL, the glorious live tiles, the newly introduced Fluent Design System, and not forgetting Continuum, will together set to astound, mesmerize, and silence all critics of both Microsoft and Windows.

    For Microsoft enduring endeavors, finally, it will be the case of, he who laughs last, laughs longest.

    Computex 2017 - Reference device with Qualcomm Snapdragon 835 executing Windows 10 Pro x64

    Source: https://www.engadget.com/2017/05/31/qualcomm-snapdragon-windows-always-connected-pc/

    Computex 2017 - Qualcomm Snapdragon 835 executing Windows 10 Pro x64 Demonstrations

    (1) https://www.youtube.com/embed/oSXUDKpkbx4

    (2) https://www.youtube.com/embed/YR39vRMmelo


    Microsoft Build 2017 & National Retail Federation (NRF) 2017

    Windows 10 IOT Enterprise (formerly known as Windows Embedded) running on a 6.3" Intel Cherry Trail Atom CPU device that notably has 4G LTE. This device is a mobile point-of-sale (PoS) device, proving Intel can, if they really desire, stuff smartphones with their chips.

    Source: https://www.windowscentral.com/checout-m-runs-windows-10-iot/

    (3) https://www.youtube.com/embed/7oYehR3yIgQ


    • siko

      In reply to Locust Infested Orchard Inc.:

      Thank you for your excellent post!

      I still see people not getting this.

      I don't care anymore.

      For me windows rulez.

      Thanks again!

    • Jorge Garcia

      In reply to Locust Infested Orchard Inc.:

      OQO all over again. Windows just doesn't work below say 9-10" display. Windows only shines when there is a large-ish screen and a mouse, and a SEAT present. For every other task, there is iOS and Android.

      • skane2600

        In reply to Jorge Garcia:

        iOS and Android would have the same problem if they actually had serious productivity applications. It's not the OS that drives what is a viable size, it's the ergonomics.

        • Jorge Garcia

          In reply to skane2600:

          That is why there needs to be a dynamic version of Android that behaves "much like" Windows, only when certain criteria are met, but can still open Android Apps for those occasions when no fully "desktopped" version of an app exists. I assure you Apple is working on EXACTLY this for their iPad Pro line. They are going to let MacOS wither on the vine.

          • skane2600

            In reply to Jorge Garcia:

            If the "criteria" is being connected to keyboard, mouse, and monitor through a docking station, it's likely going to be DOA. As far as the iPad Pro is concerned, it already represents a retreat from being a tablet toward being more like a laptop. It would be silly for Apple to "MacOSize" iOS to be a full featured OS since it won't run legacy programs. Apple's laptop and iMac form-factors are never going to be replaced by iPads simply because the former form-factors are just too useful. Outside of consuming content, social networking etc, tablets are niche devices and always will be.

    • skane2600

      In reply to Locust Infested Orchard Inc.:

      One should never evaluate a product simply based on a few demos, they are all too easy to manipulate. At best Windows on ARM would increase battery life although we have no proof that it will since there are no devices yet to evaluate. If it does, it will hardly elevate what we already have to "holy grail of productive computing on-the-go". A 6.2" phone device wouldn't even be in the running for such a title given its inappropriate ergonomics for running legacy apps.


      I think the enthusiasm for Windows on ARM is left over from a time when some people believed (perhaps naively) that if only full Windows could be ported to ARM somehow Windows Phones would succeed. But the problems aren't purely technical but also based on the configuration of the human body (The size of our hands, the constraints on what we can comfortably carry, the limits on how small text can be and still be read by our eyes, etc)

  14. Daekar

    Lots of ignorant armchair quarterbacking in the comments for this one, I see. There are a lot of people make judgements with zero information, especially since they don't have a clue what kind of performance the new ARM chips are going to have. If they can improve on Atom performance and increase battery life, that will be enough, really. I believe that such a thing is totally doable since so much of that would be ARM native anyway. The next two years are going to be very interesting.

    • skane2600

      In reply to Daekar:

      Well, anyone who comments on the merits of on an unreleased product is just speculating. It's no more "ignorant" to say that the product won't succeed as it is to say it's "doable".

    • hrlngrv

      In reply to Daekar:

      Indeed, no one outside MSFT knows anything yet. I'm very curios whether ARM can run binaries meant for Intel/AMD via emulation using less battery power than running the same binaries on Intel/AMD. If so, Intel and AMD would be in a very, very bad place. I'm very skeptical about that. While much, perhaps most, software could be rebuilt for ARM and run more efficiently on ARM, I'm also skeptical that ARM can do everything more efficiently than Intel/AMD.

      • Locust Infested Orchard Inc.

        In reply to hrlngrv:
        Quote: "Im very curios whether ARM can run binaries meant for Intel/AMD via emulation using less battery power than running the same binaries on Intel/AMD. If so, Intel and AMD would be in a very, very bad place."

        With the full breath of Sky Lake CPUs released (HEDT LCC [from low core count dies] due towards the end of June, but HCC [from high core count dies] some time in 2018, according to latest reports, amid some disappointment), and with the consumer chips of Kaby Lake released, the next installment, the eight generation Core, Coffee Lake, is expected to improve performance up to 30% for the Y series mobile chips, slated for release in December.‡

        With two major announcements in as many months by Intel's competitors, Threadripper and EPYC from AMD, and ARM chips having demonstrated executing Windows 10 Pro x64, Intel has suddenly been jolted with 240V up their hind quarters.

        Intel are rightly feeling the heat brought upon by their arrogant complacency. Intel are sure to make Coffee Lake a substantial upgrade in performance to ensure it stays relevant and ahead of the other silicon chip foundries (e.g., TSMC, Samsung, etc) for its own survival.

        Based on the encroaching competition, I envisage Intel will be able to challenge ARM and the Qualcomm Snapdragon 8xx in both performance and critically power consumption, at the earliest, towards late 2018 and 2019.

        [With the forthcoming release of Coffee Lake, I suspect Microsoft will refresh its Surface Pro line, in or around Q3-Q4 2018. The just released Surface Pro 5 (without the moniker 5) was therefore in my view a stop-gap measure, a year and a bit prior to the real worthy successor of the Surface Pro 4.]

        This would allow for Microsoft's vision of an always-connected PC with telephony/cellular features in-built, in a form factor not currently addressed by today's run-of-the-mill cellular phones/phablets, becoming available on both architectures, ARM and x86-64.

        So it shall become apparent that the fourth wave of the cellular/mobile revolution shall be instigated by Microsoft. Time to bid farewell to Google's Lagdroid and Apple's iOS.

        [The first wave was dominated by PDAs, notably executing Palm OS, Symbian, and WinCE. The second wave was dominated by Pocket PC/Windows Mobile (up to 6.5), Blackberry OS. The third wave was dominated by iOS and Android].

        Source: http://www.tomshardware.com/news/intel-eighth-generation-core-i7-coffee-lake,34578.html

  15. Althalus

    could this possibly allow for a portable gaming device? With the Play Anywhere initiative, they could push a mobile device similar to Nintendo Switch?

  16. harmjr

    Wait Paul this is the most significant advancement to Windows in recent years. Ok so I cant run photo shop and AutoCad fine. I didnt want to but I would like a device that runned the Desktop Versions of OneNote, Outlook and other software like CHROME. There are so many cheap tablets and well how many Tablets are produced by the big computer companies HP, Dell, Lenovo. They just dont make any and guess what kind of device I will buy my child to break over and over a $50 kindle! Oh and what processor powers that Google Pixel you keep praising.... Or that Samsung Galaxy its not a Intel Chip! So if its boring thats a good thing that means Windows has a future. Windows just Works...

    • Narg

      In reply to harmjr:

      Just a slight fix to your statement: No, you can't run Photo Shop x86. So why can't Adobe write a Photo Shop ARM? As if anyone would want that anyway... The PC world has always had HEPC (high end PC) and lesser machines. Nobody in their right mind would ever run high end software on a low end PC, even if they could. The ARM will not be anywhere near or expected to be a high end machine. It's kind of like asking someone to merge a Smart car onto a busy highway. Sure it can be done, but only the crazy attempt such...

  17. shanneykar

    For someone who covers Microsoft exclusively [ I guess ], you are missing the point of Windows 10 on ARM. Here is my 2-cents:


    Windows 10 on ARM working flawlessly [ not driving powerful graphic apps but for average Joe ] is the MINIMUM requirement. Yes as you said nothing interesting there.


    But what's exciting and why MS spends all this effort is entirely another story. Intel failed with Atom. Qualcomm ARM is in almost every smartphone. So what can QUALCOMM give MS that Intel is not yet there to provide:


    1. Form factor
    2. Its mobile form factor is going to make MS dream finally come true. Putting PC on the pocket. Imagine a galaxy s8. or a surface hinged mobile.
    3. C_SHELL
    4. Most IMPORTANT and EXCITING. Why? So now I have the best OS in mobile format. And MS knows and working towards future where content move with user and not stuck to a device. So with another few iterations Windows 10, I am REALLY looking forward to a future where I have just screens everywhere [ Mirror, Closet, Refrigerator, Car, TV, Projector, Monitors etc ] and ONE DEVICE with CPU identify each screen, adapts to its size, and drive them. Even opens the default app for that device.
    5. Fanless Design
    6. Yes, Surface does that even with i5 but I bet it is an engineering marvel.
    7. Battery
    8. Now you have to boast 13.5 hours. With ARM, a whole day will become normal.
    9. Instant ON
    10. With everyone used to Instant-On, you get on a PC and when you have to wait seconds to minutes, annoying but we accept since its PC. That will change with ARM and super exciting
    11. APPS
    12. When you get the powerful OS in mobile form, I will definitely come back to windows platform for mobile phone [ even though initially I may keep android for some apps like robinhood ] and app developers will come back to windows.


    Always connected is a feature but intel has that as well.


    So bottom line - Windows 10 on ARM if works flawlessly and run win32 apps as well, it is DEFINITELY boring but INTERESTING/EXCITING

    • hrlngrv

      In reply to shanneykar:

      Re #4, how many automobile, TV or refrigerator makers would standardize on Windows given where Windows is today on phones?

    • skane2600

      In reply to shanneykar:

      There's a difference between a dream and a fantasy. There's never going to be a "PC in your pocket". A real PC in your pocket would mean you could take the device out of your pocket and somehow you'd have a full size keyboard, reasonable size monitor and perhaps a mouse running a full desktop OS. If a PC in your pocket is just a device that you can hook up to peripherals when you are tethered at home or the office, we've had those for years. Laptops are the true PC in your briefcase and that's about as small as you can go without significantly compromising the experience.

    • Jeremy Petzold

      In reply to shanneykar:

      I think MS's dream for phone to desktop docking can only be realized in an Android ecosystem for mobile.


      128-256 gig model phones. Windows 10 installed and android installed. you drop it in the doc and it switches to windows 10. you pick it up, it switches to android.


      Shared storage and beefier Microsoft apps for office to integrate well with desktop office and you have a winner.

    • obarthelemy

      In reply to shanneykar:

      The concept of "best" anything, OS in this case, always seems to me immature and tunnel-vision. Win10 certainly is the best OS for some users. And for some uses. Probably for a lot of users/uses even. But not in the absolute, for all users in all situations.

      For example, my elderly parents are struggling with its power, richness, and... paroquialness. They've got Android on their phone and tablet, the Windows' way to do things doesn't come naturally. They miss their usual UI, apps, capabilites; they get numerous attempted hacks so I'm remoting to their PC about once a week.

      My own tablets are dual-boot, and spend 95% of their time in Android.

      I like Windows. I don't think it's "the best" except on my desktop, maybe on a laptop, certainly not in a tablet nor in a phone. And certainly not for simple tasks and unsophisticated users. It might get there, say when I no longer need a registry hack to put it in tablet mode and lock down the homescreen ^^

      • siko

        In reply to obarthelemy:

        Depending on how you look at things.... best can also be defined as.... more mature, stable, versatile, more efficient, better manageable, better documented. With other words... proven!

        An OS takes decades to mature. Take Unix, Linux, Windows.... all took decades (and windows being relatively young). Android, as far as my information goes, is a rip off Linux distro. Even so, it is only starting to mature. It has to prove a lot of things before most people agreeing to this way of evaluating will consider it to be 'good'. YMMV


      • Jeremy Petzold

        In reply to obarthelemy:

        You should have your parents on a chrome OS device.

  18. skane2600

    I think boring is the best case scenario for Windows on ARM. Will it run all legacy applications with reasonable performance? We don't know. Will the devices be smaller, lighter, with better battery life? We don't know.


    Having been involved with some sketchy demos in the electronic toy industry years ago, I don't take such things at face value.

  19. Ekim

    Would sure be nice if I could load it on the Surface RT.

  20. SvenJ

    I think it will only be boring based on the hardware we see it on. On a laptop or 2-1, yea, so what? There will be those who embrace the constant connection, maybe better battery life, but those are fairly niche requirements when there are i5s getting all day life already (8 hours or more) with 'normal' use.

    I would be excited by this showing up in a 'phone' form factor. No reason it couldn't. I imagine using my 950XL like I currently do, but having a brand new experience when I dock it. I imagine popping up the start menu and not having half the apps grayed out because they don't support 'Continuum'. This is different. This is just a second monitor on a Windows PC. Been doing that forever. I imagine clicking an app in the start menu and having it come up in a Window, not full screen like they do now. I imagine opening a second app and having a new Window appear, not taking over the whole screen again. I can run multiple apps now, and switch between them, but I can't even split-screen them today. Again this isn't 'Continuum' it is Windows with a second screen, and we've been running multiple floating windows on a second screen forever.

    Full Windows, running on multiple form factors, and taking advantage of the full capability of Windows, attached monitors, windowed multi-tasking, and all, with the 'Continuum'-ish dynamic understanding of the form factor on which it is running is exciting, and new.

    Some say this has been tried before, Motorola Atrix, now DEX, even the current Continuum, but those were all phone OSs trying to scale to a big screen. This is a full PC OS running on a phone sized device. As long as we can get the phone functionality to be usable, not desktop windows on a 6" screen, it might be interesting to a lot of people. Hopefully it gets there. Still a niche, but potentially a big one.

    • Luka Pribanić

      In reply to SvenJ:

      I have to say I think the same way. This will never be what I think of as a "primary" device, but an ARM phone with full W10 OS could very well be one that I would spend most time on. Add capabilities like tablet-shell, laptop-shell, desktop dock, and it would cover 90% of my computing time, and I'd keep using same core device. This is the ultimate Continuum. I can imagine other nice possibilities as well, but let's keep it at that.


      But I do hope other ARM licensees get the right to run Win10, not just Qualcomm. Now THAT would open up the market and competition. Don't forget what MediaTek did for smartphones, and that AMD and Nvidia both have ARM devices..

    • skane2600

      In reply to SvenJ:

      The problem with all these schemes IMO, is that they are strictly tethered experiences, rather than mobile ones. You really need a mouse and keyboard, not just a monitor to duplicate the desktop experience. If you already own a PC you don't need your phone to be a CPU unit and if you don't, you have to go out and buy peripherals. A portable device always has additional constraints over a desktop device: size, weight, power. If such a phone were under $200, then maybe it could compete, but I doubt that we'd see such a low price.

      • Jeremy Petzold

        In reply to skane2600: You would have a "duel boot" device. A mobile oriented Android system that shared storage with Windows 10. when you doc it you get windows 10 and its full glory, when you have the phone, you get Android.


        • skane2600

          In reply to Jeremy Petzold:

          So what? Assuming the mess of combining storage between Android and Windows could be handled, you'd still be unable to run Windows in a mobile fashion. It would be just like running a compact desktop PC at home except a little smaller, a little less powerful, and you'd have to hook it up every time you came home.

      • SvenJ

        In reply to skane2600: I do already own a PC, but it sits on a desk at home. I do own a Surface, and I carry it....along with my phone already. I also own a very portable mouse and keyboard, so all I really need with the phone I'm suggesting is a screen. Some very lightweight portable ones already exist that I could throw into checked luggage. I could just grab a Miracast Dongle to stick on a hotel or conference room display. A lapdock like HP makes might be more convenient in the hotel. OK, it is tethered, but I'd have my Surface and my phone plugged in anyway, so no big difference. The phone doesn't need to be $200. I paid a bit more than that for the phone I'm currently carrying. Now if the lapdock, or other peripherals, cost a lot less than the Surface, I'm ahead, in cost, and complexity. This is not a solution for everyone and every situation, but certainly has a place.


        • skane2600

          In reply to SvenJ:

          So your bring along all these separate pieces and put them in checked baggage and assemble them later so you can avoid using an integrated device like a laptop that you could actually use while you are commuting? We all have our own idiosyncratic preferences but I don't think this scenario makes any sense for most people.

    • Jeremy Petzold

      In reply to SvenJ: 8 hours.....yeah, nice...I think we will see 18 hours out of an ARM based device.


  21. dstrauss

    Paul, sometimes for the life of me I just don't get your point:

    "If Windows 10 on ARM works, and I think it will, the average user won’t see any difference between that product and Windows on x86. The systems could potentially be smaller and lighter, and get great battery life, and they will include integrated cellular data capabilities. But that’s why someone might choose such a system: For the benefits, not for the underlying architecture."

    Why else does ANYONE buy any OS - for the benefits, not the architecture. Maybe there are still evangelical soldiers out there in the trenches of Windows vs Mac Os vs Android vs Linux vs Amiga but I doubt it. If WoA systems "could potentially be smaller and lighter, and get great battery life, and they will include integrated cellular data capabilities" then that is the defining feature set for the user to get excited about.

    There is NOTHING exciting about Windows itself anymore, or ANY OS for that matter...

  22. creugea7

    The question I have is about security and virus protection. Is virus protection needed on a system like this?? Since it will be running win32 apps in a container, in emulation, what are the chances of getting and spreading something?? If that part works as I think it would this would be huge for us hear (and enterprises) where I work as it would be less stressful about such virus scenarios and such. Also less chances for a user to royally mess up windows.

  23. hometoy

    The excitement isn't in the Windows or Microsoft's part in this. The only excitement to be generated will be

    1. Price, if it significantly reduces the cost of systems with and without LTE. If it shaves off enough to make a Surface < $500 then THAT would the exciting part (and, arguably , do more for competing against Chromebooks than Windows 10 S).
    2. Performance. It doesn't have to be a power-house if the tradeoff is in significant battery life improvements for about the "same computing power". Or if it includes LTE while maintaining the lower-than-Intel-based-chip systems.


    People running Photoshop, Visual Studio and 3D-intensive games are not going to find much use for this probably, but for everyday users if it does what regular Windows 10 does AND improves battery life AND significant price reduction AND LTE connectivity then it may be more appealing.

    • skane2600

      In reply to hometoy:

      Most Chromebooks are under $300 and the primary market for them is education where cellular connectivity doesn't add value. So I don't really see always-connected ARM PCs being a direct competitor to Chromebooks.

      • Jeremy Petzold

        In reply to skane2600: it isn't about being a direct competitor in the market but giving hardware makers the ability to use Windows in their system designs for the ARM based product lines. Look at Samsung. their core deign competency is ARM.

        They buy hardware to make chromebooks...now they can easily pivot to Windows devices too.


        • skane2600

          In reply to Jeremy Petzold:

          Samsung already makes Intel based Windows laptops so apparently they've achieved Intel design competency too. If we imagine some other company jumping into PC making for the first time using ARM, should we really expect an increase in Windows sales? Is there currently a dearth of PC options that an ARM-based PC would remedy?

      • normcf

        In reply to skane2600:

        As well, stateless Chromebooks will always be more secure than any windows, including S, because a reboot always brings you back to a virgin copy of the OS. On any windows OS you can never be sure there is no malware after it's been on the web or had a used USB drive plugged in. This is something that is extremely valuable to business for certain roles (no, I didn't say for every role). Some businesses may determine that Windows S is good enough, but others are already seeing the benefits of of ChromeOS.

  24. VancouverNinja

    The main goal of being on arm is simply to allow Microsoft (Windows) to be on the same playing field as Apple and Android. Period. Its main benefit is all hardware based, as you pointed out. Windows 10 is clearly the superior OS but could not reach the pricing levels to compete for the consumer at the entry pricing level. It was also stopping their vendor partners from being able to fill these entry level price points with Windows offerings.


    Windows 10 S on ARM calls into question any benefit of Chrome books to anyone - that is the story. It is the part of the story that makes Windows 10 on ARM exciting. Windows 10 on its own, and where future releases are headed, is exciting enough.



  25. rameshthanikodi

    "this is useless because it's not as fast as my core i7" - karma777police

  26. BoItmanLives

    Bloated x86 Windows 10 - that MS can't even really get right on the most powerful Intel CPUs now 2 years in - only it's going to be emulated on a peashooter mobile SoC and have no 64 bit program support.

    I fail to see who this would even be for or what user challenge this solves. Laptops have had cellular for a long time. Pocketable Windows devices have existed for a long time and never taken off.

    What new does this bring to the table?

  27. Narg

    iOS is kind of boring too, arguably so is basic Android. It's the software written for these devices and their OSes that makes them exciting. And, they both do a great job at video editing and game playing. So why couldn't Windows on ARM do that too? I think you are speculating incorrectly Paul.

  28. stevenmci

    Honestly, no one will care.

  29. Bill Russell

    "If Windows 10 on ARM works, and I think it will"

    depends what the definition of "works" is. As my, and most others' reason for using "Windows", it won't.

    For example, "Windows" is the only OS which I can run my Solidworks on among other things. If this is just dismissed as a niche "legacy need", they've got another thing coming. I can say that a power sippin' WOA machine ain't going to cut it.

    If I don't need to be using Solidworks on a room-warming Windows 7 tower desktop system, I am going to use my phone or chromebook (and now chromebox!).

  30. Maelstrom

    Windows 10 on ARM + CShell + eSIM + PWA = ultimate ultramobile device. So, there's nothing boring about that!

    • obarthelemy

      In reply to Maelstrom:

      For what % of the market, you think ?

      • Dont Fear the Future

        In reply to obarthelemy:
        More than Mac's or Linux's percentages in a year or so I would think.


        • skane2600

          In reply to Dont Fear the Future:

          The relevant benchmark for the success of a WoA device IMO would be how its sales compare with Windows on Intel. The combined percentages of the Mac and Linux (non-server) are already a lot smaller than Windows.

          • Dont Fear the Future

            In reply to skane2600:
            I agree. That would be the best benchmark. Intel is making strides with battery life, but to be thin and light while maintaining battery life and performance requires an expensive Intel chip. I can easily see ARM-based products meeting those needs at a much cheaper price point.
            For example, a Surface ARM tablet. It can have the same dimensions and batteries of the newest Surface Pro, but have even better battery life and at the cost of a Surface 3 Non-Pro Atom based tablet.
            I think of all the Atom based Windows 10 devices out there being replaced with ARM and being much more powerful for about the same cost.


            • skane2600

              In reply to Dont Fear the Future:

              We'll have to see when real devices are available but at present I note that entry level Windows laptops are no more expensive than entry level Chromebooks despite the latter using ARM chips instead of Intel.

              • Dont Fear the Future

                In reply to skane2600:
                The best evidence we have of ARM vs Intel was the "extended demo" linked in the article; where the press had the demonstrator open their websites up on like 9 tabs in Edge and it performed really well (with even a guy at the end saying, "This is doing better than my Mac can"). I'd bet my Surface Pro 4 couldn't do much better either.
                The x86 emulation stuff? Who knows. But the SD835's native performance (Edge browser) performed very well, and for most people that will be good enough for them. Especially if the price is right.
                P.S.: Office ran pretty good too, and that was the Win32 full desktop version of Office, not the mobile version.
                Edit 2: I just re-watched that video, and not only did they have many tabs open in Edge, you can also see that they still had File Explorer, Word, Excel, Powerpoint, Outlook, Excel, Task Manager and the Control Panel still opened and running in the background too. Pretty impressive; especially considering that the developer device was the size of a really thick smartphone. If heat / throttling was occurring, it is possible it would perform even better in a larger device like a laptop where the heat could be dissipated better due to the larger surface area.


                • skane2600

                  In reply to Dont Fear the Future:

                  It's the applications that MS doesn't control which will determine the viability of WoA. I was around during the "almost" compatible PC clones era and there's no longer any patience for less than 100% compatiblity. The ARM chips will have to be significantly faster than their Intel counterparts to overcome the inevitable reduction in speed that any emulation scheme will create. The ARM chips that MS has been talking about are high-end so WoA computers are likely to be medium to high cost devices.

                • Dont Fear the Future

                  In reply to skane2600:

                  True. We'll just have to wait and see.

              • Dont Fear the Future

                In reply to skane2600:
                Very true. We have to wait and see.
                But how well does the performance / battery life of those Intel Computers compare to those Chromebooks?
                However, that's an apples to oranges comparison anyways. But soon we'll be able to test both ARM vs Intel on Windows devices; and see if Geekbench scores actually mean anything or not.


    • BoItmanLives

      In reply to Maelstrom:

      Oh boy, is "eSim" the new fanboy buzzword ala "Hololens", "OneCore" and "Continuum" that gets spammed ad naseum until Microsoft quietly stops talking about and ultimately abandons it..

  31. bluvg

    What about x86 malware? Perhaps W10S blocks it on every platform and not just ARM, but is that also a factor with W10 on ARM?

  32. glenn8878

    Might be fun to have long detailed discussions about what this means for Windows devices. Almost as good as long walks on the beach. Let's talk about Windows' bright future, but the truth is it'll be a disappointment.

  33. ssimo

    > Windows 10 on ARM probably won’t work well for high-end tasks that require powerful Intel hardware like

    > video editing, game playing, and the like.


    Two points:

    (1) I would include browsing with multiple windows/tabs as a fairly common task that cannot be done well by low-end PCs

    (2) Just like you wouldn't buy a car which cannot meet all of your needs, it wouldn't make sense to buy low-end PCs even if there is a slight chance that you would need browsing with multiple windows/tabs, video editing, and gaming. I fear that folks who buy Windows 10 on ARM are going to be the second batch of disappointed users after Windows RT.

Leave a Reply