Dual-OS Update: Intel, Nokia, American Megatrends and Why Google Might Not be Needed.

Posted on 01 March 2014 by

TD300_Detached_Dual-OS (2)There’s another BIOS producer getting into the Operating System Vendor space. Why? Because there’s a potential market for Windows 8 devices to offer Android as a live-switch alternative and the BIOS producers are well positioned to enable this. What are the options and issues, drivers and players and is Google at risk of being left out of this potentially big shift to high-power Android in the consumer space.

At MWC last week I spoke to a number of people about Dual-OS systems and there appears to be quite a positive note on it due to slow progress being made by Windows stores. It seems developers are unimpressed by the numbers available for the Microsoft Stores but excited about the prospect of Windows 8 Store unification. They won’t put any major effort into Windows apps in the meantime though and that could leave an 18-month gap. What do Windows laptop and tablets manufacturers do in the meantime? The idea is to go beyond the Smartbook (cheap Android-only) laptop concept and offer dual-OS on high-power platforms which could work not only to solve the application problem but could offer some very high power FullHD Android gaming experiences. The latter idea is one worth thinking about as gaming developers huddle around Android on tablets and smartphones that are already getting FullHD displays.  Think about the developments in 2-in-1 hardware (fanless Core-based systems should be common later this year) and it fits nicely as a potential solution for the laptop manufacturers. If it gains traction in the next 18 months then how about another stab at the Android-powered laptop? Consumers could be better served with a laptop-focused Android build now that the architecture has a) advanced greatly since the last attempt at smartbooks. b) hardware for always-on high-power and high-quality 2-in-1 solutions is now available. c) Consumers love Android d) Gaming on Android is BIG.


Background: 6 Dual-OS Architectures.

There are many ways to achieve a Dual-OS solution. Here are 6 potential solutions.

  • Classic dual-boot. Requires reboot to change systems generally not consumer-friendly.
  • S3 sleep+Memory juggling for 4-second (or less, according to my contacts) switch-over.  One engineer I spoke to told me they can do it without partitioning memory and in sub 4-seconds using SSD and some tricks.
  • Dalvik Runtime solution (running concurrently on top of Windows) e.g. Bluestacks, Myriad Alien (Myriad is the solution used in the Sailfish OS on the Jolla Linux smartphone.)
  • A ‘virtual box’ software hypervisor on an existing OS. E.g. Linux and Virtualbox with Android and Windows on top of that. Somewhat inefficient and, given the amount of lines of OS code required, probably not the most secure solution..
  • A fully virtualized dual-OS system running concurrently on a Hypervisor which has direct access to CPU features such as VT-x. (E.g. DuOS from AMI. See below.)
  • Dual-CPU system (a-la Transformer Book Trio.) Expensive but possibly the neatest solution. (Imagine a Core CPU with Atom or ARM CPU on one system.)


Intel, ASUS, Samsung Dual-OS solutions and the Google Problem

At CES we heard Intel announce their Dual-OS platform. They’re working on it but there’s next to no information available. My research indicates that they’re looking at a platform that switches from one to the other in sub 4-seconds but one that also puts the alternative OS into sleep. There might be a few partner issues on the horizon though.

ASUS have their ‘proprietary’ system too. I can confirm this is not the same project as Intel’s and my sources tell me that Google weren’t happy about it, just as they weren’t happy with the Samsung ATIV Q which appears to have been denied access to Google Play and taken back to the developers for a re-think.

Is Google Play critical though?

Think about this; If Google won’t allow their services to be used on top of Android Open Source then how about Nokia services, Microsoft Services or Amazon services? Take a look at the Nokia X project to see how many Nokia and Microsoft services have been integrated into Android open-source. Think about the success of the Amazon Kindle, the apps that have been ported over and the Amazon services that have been integrated into that product. 3 years ago we didn’t have these alternative service layers but now we do and Google doesn’t have the monopoly any more. If (and that’s a big ‘if’) Dual-OS leads to Android consumer laptops and gaming rigs it’s this high-powered i86 architecture that has the lead and if Google doesn’t want to cross this bridge, Nokia (and by definition through the take-over and Nokia-X project) Microsoft and Amazon have a good opportunity.


DuOS Available for Download Soon

Back to American Megatrends: There’s an interview video with President & CEO of AMI, Subramonian Shankar available below (via IT-Techblog.) In the video you’ll hear about the hypervisor architecture and a public download of the system. Through some digging I’ve found a site called AMIDUOS which looks like the place you’re going to find the software soon. It was tipped by an engineer working at American Megatrends so it looks likely! I also have a request in for early access.

The hypervisor solution is not without its issues. How do you turn off an alarm on the ‘other’ running OS? How do you control notifications, access to USB hardware, GPS and other hardware? Can you leave Android running and audio stream while switching to Windows to do some video and audio editing?

Dual-OS is a huge topic. it’s a bridge to Android-only laptop and gaming solutions. It’s a potential win for i86 and, in a bizarre twist, it’s a solution that Microsoft now has in its pocket. It’s not without issues though and if early solutions are messy it will die just as the smartbook did. Given the potentially huge gaming laptop/console feature though, the better platforms, a move to 2-in-1 hardware and the marketing budget that major OSVs have it could get the traction it requires. Are we seeing the first steps towards a shift in laptop OS for consumers and a potentially new path for the Wintel alliance?

Latest products in our mobile pc database.

10 Comments For This Post

  1. earl cameron says:

    What are the options and issues, drivers and players and is Google at risk of being left out of this potentially big “shit” to high-power Android in the consumer space.

    I think you mean shift.

  2. Chippy says:

    Shit is now shift!

  3. ArchiMark says:

    Kinda liked it the other way….. ;-)

  4. Tim says:

    i86 should be x86 as well.

  5. Chippy says:

    I honestly don’t believe that a generic Android x86 build has any sort of chance but yes, others may see an opportunity.

  6. Gerri says:

    By dual OS, I’d rather see Windows 8 and a desktop Linux distro especially for 10+” screens.

    “They won’t put any major effort into Windows apps in the meantime though”
    Gee, I wonder why devs aren’t making money on Windows. Maybe it has to do with them not making any Windows software.

    “dual-OS on high-power platforms which could work not only to solve the application problem but could offer some very high power FullHD Android gaming experiences”
    The only apps I use on Android are gimped ones that I just put up with until I get to a PC to use the much better desktop software (both free and paid ones too). So the only apps I see being useful on Windows 8 are those casual games. For “high power” gaming, there are PC games already. I doubt gaming companies will ever make a high end game comparably to their desktop ones when they’re likely targeting the lowest common denominator in the phone category limited by both processing power and screen size. Maybe they can make games competitive with mobile console systems which I see is somewhat happening. Making a high end Android game that only works well on a PC and maybe only ones with a dGPU doesn’t make sense when they can just use the existing system.

    Also, if you keep up with Android, Google has been actively gimping it. External keyboard support has been getting worse like the removal of keyboard shortcuts. App devs have followed suit in the removal of special keyboard features. Also, with KitKat, external storage support has been limited further in the name of “security.” The kind of people likely to even bother with with a dual OS system wouldn’t like the limitations of Android on a desktop device and only put up with it on their phones.

  7. ArchiMark says:

    +1 for Win8 and Linux…..

  8. Tim says:

    If performance and experience isn’t reduced, Android and Windows 8 might be fine on dual OS UMPCs (8″ and smaller for me) especially ones with no mouse pointer but on these “large” devices being shown, I doubt they’re going to do well. Probably just as well as the modern UI on desktop PCs.

    As said, Google is actively limiting Android with each version depending on your point of view. Maybe Google really doesn’t want to see Android on notebooks and desktops and want to push Chrome OS in that space by removing features that desktop tasks would usually benefit from. Also, as you’ve said, possible no Google Play access would almost defeat the whole thing. Sideloading apps would be too much for many people and 3rd party app stores aren’t likely to popular. Apart from Kindles, the Amazon App store doesn’t seem to be used mich on Android devices due to performance issues, app license checking problems and delayed app updates.

  9. D T says:

    I wonder about one thing though. Microsoft has always had special deals on licensing OS to OEMs, but with conditions. For example, 1/4 a century ago, they forced the OEMs not to make computers with other OSes on, in order to get the special pricing. So for the next decade plus, it killed off the potential for OS/2 and other possible OSes, and in 90’s, GNU/Linux based desktops. Would the current MS be so generous that they will turn a blind eye today, and yet provide the $15/licence to OEMs?

    And then it’ll be so ironic that for the dual OS OEMs who have already been paying MS the patent licensing fee (which is per unit based, I think, may be $5? I am not sure), and then pay again for licensing the MS OS (Windows), hahaha.

    As a matter of fact, I have rarely seen any factory made dual booting home computers. Sure there have been the odd dual boot Android and Windows here and there, usually by some minor player OEMs, and then the functionality on Android was very bad (at the time) due to lack of x86 compatible Android apps.

    And would users want dual boot? You see, the audience on UMPCPortal is mostly power users and enthusiasts. But we are only the minority in the market. Most people are very happy with their iPhones and their iPads. So for the readers here, sure many would want dual boot, as that is simply “cool”. But for those who don’t, the added cost would not be welcomed.

    Android and Linux have come a long way. Recently I read about the Android x86 KitKat on liliputing blog, and so I installed it on my Acer Aspire One and it runs amazingly well. Some apps won’t work though, like some w/ video playback feature like Netflix. So Android on x86 is real, and Intel has a special group that supports that.

    I have also tried Bluestacks and GenyMotion, each has some pros and cons, but requires quite a bit of power to run, so under emulation, the battery life of such device (like a Bay Trail tablet), will go down faster.

    Therefore, the best solution is the BIOS supported memory swapping kind, if they can pull that off without serious bugs. It has been announced for quite a while and I am wondering why it still hasn’t been out of vapourware yet. So I guess they prob run into some bugs or problems.

    Traditional dual booting is not really that bad, but only booting is quick on Windows 8 today, not Android. Android still takes very long to boot, but it’s expected not to be shutdown most of the time. Unless this can be resolved, traditional dual boot will still not be very good for impatient users (which means most people :-) ).

    And IMHO and I have said that before, it’s always about applications / software / programs or whatever title you want to call them. Windows has been great not because of Windows, but because it can do most of the peripherals and most of the software titles. Then suddenly iOS has created a very rapid rise marketplace, and then Android follows suit, so suddenly, Windows has been met with competitions.

    Here is a funny story: I was at the local Microsoft Store, and I had a very nice chat w/ a rep there. However, I still had to mention that the lack of apps on the Microsoft app store. Guess what the rep replied with? He said that Apple’s Mac app store has even less apps. Suddenly I don’t know what to say. First, I doubt if that is a lie. But second, does that mean Mac has less software than the Modern UI apps? I doubted. But as for Mac, there are no two separate kinds of apps, like MS’ “Modern” and “non Modern” (what else do you call that?). But I can’t seem to believe that is true but I think that since Mac users, just like Windows (pre Modern) users, get their wares from other methods: shrink wrap retail boxes, download from the internet, or whatever. I don’t think that is a fair comparison because are there more Modern UI apps out there that are not on Microsoft’s app store? I don’t think so. So I think that is just a trick answer. MS’ Modern UI is built to compete with iOS iPad directly (since they don’t have Windows Phone for tablet), so why didn’t the guy put the iOS for iPad list of titles to compare with the Modern UI ones?

    IMHO, there will be far and few in between for dual boot Android and Windows systems made for the consumer market; or even the commercial market or vertical market. I could be wrong of course.

  10. dotpeople says:

    Dual OS can also mean dual-Android, dual-Sailfish, dual-Windows, i.e dual persona.

    Since many “apps” are pretty-but-dumb terminals for cloud services, multiple OSes provide buyers with management of cloud service risk. Why should the buyer’s menu of cloud services be limited by the vagaries of device, carrier and cloud service bundles?

    Type-1 fully virtualized hypervisors are good for security, e.g. baseband processor isolation, but requires the fanless Intel CPU to support vPro, e.g. the i5-4302Y CPU in this premium Panasonic tablet: http://www.pcpro.co.uk/blogs/2014/01/10/panasonic-toughpad-fz-m1-review-hands-on/

    Nokia TLK and SVS (Type-1 hypervisor) on ARM:

    Type-1 Xen on clients (Intel/x86 and Samsung/ARM):

    OK Labs dual-persona:

    Green Hills dual-persona:

    AMD and BlueStacks bring Android to PCs in Europe:

Recommended Reading

Top Ultra Mobile PCs

Intel NUC Bean Canyon
0.0" Intel Core i7-8559U
Intel Hades Canyon NUC
0.0" Intel Core i7-8809G
Intel NUC June Canyon
0.0" Intel Celeron J4005
Acer Aspire Switch 10
10.1" Intel Atom Z3745
HP Chromebook 11 G3
11.6" Intel Celeron N2830
Acer Aspire E11 ES1
11.6" Intel Celeron N2840
Microsoft Surface Book 2
0.0" Intel Core i5 7300U
Acer Chromebook 11 CB3-131
11.6" Intel Celeron N2807
Lenovo Ideapad Flex 10
10.1" Intel Celeron N2806
10.0" Intel Atom Z3740