Beyond form-factor differenciation, challenges of hybrid devices

This post comes from a question that Jolla asked me on Twitter, after I twitted them about “Mobile computers”

Before Nokia ditched all their Linux smartphone branch (Maemo, Meego, Meltemi etc.), this branch was refered, first as Nokia internet tablets (or NIT), and after, when the N900 came out, as “mobile computers”. The N900 was indeed a mobile computer, that was capable of running multiple tasks, like browsing the web while receiving chat messages, or skyping while playing angry-birds. Obviously, it is not a full blown computer, due to the size of the screen, that is too small to display many windows at the same time.

For geeks, having a computer in the pocket was the nearly a dream, and sadly, the N9 did not continue that dream. Being way more beautiful, and with a better UI, it was mostly designed for end-users (non geeks). Geeks still loves some of the powerusers features (like terminal and super multitasking capabilities), but the N9 is not a mobile computer anymore.

Nowadays, nobody talks about mobile computers, although it is still an interesting concept, because of quick changes in the mobile world.

Blurry categories

Some years ago, products were easily defined.

  • A screen, a keyboard, a mouse ? That’s a computer
  • A T9 keypad, an antenna ? That’s a phone

That’s it ! But quite recently, and thanks to Moore’s law, phones became more powerful. The raise of mobile internet also created some computers that are less powerful in order to be more autonomous, so instead of being strictly categorized, devices now sometimes belong to several categories. In an attempt of classifying them, I will still build a list

  • Desktop, are non-autonomous, but powerful
  • Laptop are basically an autonomous desktop, with some tradeoff on performances
  • Netbook are super autonomous laptops, that are really not powerful
  • “New generation Tablet PC”, that are for example MS surface are both a computer and a mobile device. They are sometimes more powerful or autonomous than a netbook, but might not be able to compete with them in every front (like using specific software like CAD)
  • Tablets are very autonomous devices, that are most of the time touchscreen, but have strong limitations (like lack or limited multitasking, no window management, impossibility to run specific software like CAD etc)
  • Phablets are a category of quite autonomous phones with a huge screen. Not a tablet, but not a phone either.
  • Superphones are powerful smartphones that are more limited in autonomy
  • Smartphones are very autonomous devices that can do many interesting tasks (mails, internet)
  • Dumbphones are super autonomous devices that are made to send SMS and make calls (and maybe more, see Ashas)

Well, this is only an attempt, as you see that the difference between laptop and netbook is more something smooth and continuous than something discrete. For example, how do you classify my Lenovo X200 ? There are also not so much differences between superphones and phablets nowardays, and the Asha line can be considered as both dumb and smart-phones etc.

What I see in the previous classification is that instead of having two devices, one with heavy computation power, and one with heavy mobility, the trend is to build different devices that fill the gaps in a continuous way. Some devices are 75% computation + 25% mobility; some are 25% computation + 75% mobility etc.

We can also see a strong correlation between computation power and screen size. A desktop can affort having huge screens (22′), a tablet and a laptop have decent size (7′ to 15′) and smartphones or dumbphones have less (1’5 to 4’5). It is because displaying the number of pixels associated to the size of the screen takes computation power, and uses battery faster.

Well, filling those gaps is nice, it is a way to differenciate and create new products and markets, but one problem arises. We can now build devices that are both powerful and efficient (iPad, Galaxy SIII, Lenovo X200), do we really need 100% computation power, or 100% autonomy ? Do those products still make sense ?

The answer is yes, obviously, but not for everybody. Scientists still uses supercomputers, engineers still uses heavy SW that needs a good workstation, and explorers or adventurers still need autonomous devices. But for the average joe, what do he / she need ?

The road to the universal hybrid device

What surprises me, or makes me afraid, is that the Galaxy SIII is more powerful than my Lenovo X200. Well, strictly speaking it is not true, the Lenovo crushes the GSIII in the Linpack test for example. But still, that’s an interesting fact. I can imagine that, for the future, I would have an ARM (mobile) processor (that powers for example an hypotetical Galaxy S V) that will be exactly the same than the Centrino powering my X200, so the need of having the X200 become void, since one mobile device will be able to be as fast as it.

But what I want is a device that is design for working on, for coding on it, for being efficient. I want a keyboard, a trackpoint or a mouse. So I want my Galaxy S V to become a computer as well. Or I may want it to be a tablet, since the touch screen might facilitate some tasks.

You see my point, I want a device that can “transform” to other devices. And this concept of hybrid devices already exists.

The Asus padfone, an attempt of making a hybrid device

A very good idea on the paper, the Asus padfone is a phone that can be plugged inside a tablet, that is basically a screen + battery, the phone provides the OS, and the tablet can be docked in a station, that is a battery and a keyboard. So, in theory, the padfone is both a phone, a tablet and a laptop.

In practice however, the experiment is a fail, as the padfone was not known for it’s groundbreaking success. Why that ?

First, the user experience on a phone, a tablet and a laptop are very different. A laptop often have a point and click user interface (mouse) that is heavily supported by the keyboard (keyboard shortcuts like copy / paste and direction key navigation).  Whereas a tablet is closer to a smartphone. It is a fully tactile device, that can run only one task at a time (at least on the iPad, the reference tablet). Although creating an hybrid tablet / smartphone makes sense, creating an hybrid laptop-tablet don not makes a lot.

Another problem is also computation power. As I said before, ARM CPU are being more and more powerful, but  they are not ready yet. There is no way that the Galaxy SIII, one of the most powerful smartphone out there currently can compete with my Lenovo, although the Exynos 4 powering the GSIII is running at 1.5 Ghz x 4 whereas the one in my Lenovo is running at 2.26 x 2.
This is simply because of the type of the CPU. ARM CPU are not meant to run computation, they don’t have technologies to do that (like multiple ALU, pipelining capabilities etc.) And sadly, on our laptops we are still running “heavy” computation. Just check the number of tabs in your internet browser, and the fact that you are playing a flash game while listening to music etc. Because on a laptop, multiple tasks are being run simultaneously, all the time, ARM processors might not be enough to do all that work.

I won’t talk about GPU since it is not my domain, but what I said about CPU should be also very valid for GPU.

Given the previous points, currently, it is impossible to merge both smartphones / tablets and laptops, because they are too far away in the autonomy-computation power scale. A laptop might be 15% autonomous / 85% computation whereas a tablet is more 65% autonomous / 35% computation, and a smartphone 85% autonomous / 15% computation power. Technology is not ready yet.

The Motorola Atrix and the Lapdock. A first attempt at merging smartphone and laptop.

But today, technology is indeed ready to do the first steps merging, or hybrid devices. It should not be a radical merge, like the Atrix, that was a failure because of the CPU unable to handle the Linux system for Laptop. Instead, it should start by merging devices that are close.

The iPad and iPhone often shares similar CPU and GPU. What is strange is that Apple did not find useful to be able to merge these two devices into an unique one, because it would be the easiest merge.  Since nowardays, mobile CPU / GPU are able to handle both smartphones and tablets, and phablets, people might want to be able to have a portative phone for mobility use (in the crouded train, or at least for calling people), while docking it in a bigger case for other usages (when a big screen is needed, for example for presentations, or for reading a PDF etc). The tablet and the phone are then the same device, with no redundancy of stored information etc.

Challenges of hybrid devices

Small list of challenges for hybrid devices

  • UI that might be different between a laptop, a smartphone and a tablet
  • CPU power, that should be enough for the smartphone, but also the tablet and the laptop (beware of difference of usage)
  • Autonomy, that should be enough to power a powerful CPU in highly mobile mode (phone)

Answer to Jolla

I have done a small state of the art of today’s hybrid devices. Now, what I forsee in a not-so-distant future is that hybrid devices will start to pop out, first merging devices that are close (phone – tablet), (laptop – netbook), and soon, creating one unique device, that is universal. It will have some docking capabilities in order to make it a phone, a tablet or a computer.

People will ask for unification of their devices, as they want their mobile devices (phone / tablets) to do more and more tasks that their less mobile devices (computer / laptop) do. People want to do text processing on a tablet (it can be done, but not as good as with Word for example), they also want to be able to do that on the go.

The answer to that need should have been the mobile computer, some years ago. A mobile computer for the mass, running some Linux variant, shipping adapted versions of popular open-source software. But now, it is no more the case. Thanks to tablets, the need shifted from a small device able to do anything to a larger one, that sits between laptop and smartphone.

But tablets are inherently limited, especially the iPad that can only run one task at a time, despite its large size. Providing a better tablet, aggregating features from both the phone and the laptop, seems to be the future. However, it is impossible to have both a large and small devices (advantage of the tablet and the smartphone), so I think that hybrid devices are the answer for that problem.

What I want Jolla to study is not something new, it is the concept of these hybrid tablet-phone devices. Many attempts have been done, but most of them failed. Maybe because Android is not suited for those devices, and running multiple tasks on Android (like on the Samsung Note) is mostly a hack. I hope that Mer (Linux) and Qt are a better basis to build a good user experience on those hybrid devices.

I mean, Jolla, gives us an awesome smartphone first, and then, release a dock that transforms the smartphone into a tablet !


9 thoughts on “Beyond form-factor differenciation, challenges of hybrid devices

  1. Love the idea of a smartphone base. The dock should have its own cpu / gpu capabilities, kicking in when needed. Smartphone as conductor, dock the orchestra.

    • Hardware wise, it would be hard, as it will need a partial / complete hardware reorganization. And software wise it might be even worse (need to load in an hotplug way a GPU driver). But yes, indeed, your proposition is interesting. Let’s see what future bring us.

  2. maybe you can do it the other way around. dock and phone with different cpu / gpu. the OS skill would be to keep data sync between devices and completely change the interface in accordance with medium. the ability to show and work in different ways with the same data according to the device.

    An example with the CAD. there is already CAD for cellphones like autocad WS. you can’t do everything in Autocad WS. but you can do something, see CAD files quickly on building site. take a quick note. then get in the office sit on your desktop and work further with that information imputed via the cell phone.

    mobile apps for reading scientific papers are another example.

    • Hi Miguel.

      Well, I don’t understand the benefit of having different CPU / GPU inside the dock and phone, this basically means that you have different devices … But indeed, different UI for different apps (like the Desktop + Metro system) is a good idea.

      However, I believe that you can create both touch and keyboard/mouse friendly UI. Just need to do a bit of research and design (as I’m actually doing now).

      The problem about CAD software was their need of computation power. Did you used Catia ? The best mobile phone over there would not be able to run the complex models that Catia can load (10000 + pieces for example) so we will still need to wait to get a perfect hybrid device that is able to do that.

      • I used the example of Autocad WS precisely because of CAD programs have high processor needs. Autocad WS is a product of Autodesk, the bigest world cad software house. A company that sells autocad, 3d max studio (rendering program), Revit (BIM abled program), Navis (4D production planning), Inventor (mechanical engineering CNC machines), etc, still launched Autocad WS, a cad application for smartphones and tablets.

        Why this? Because sometimes, you are away from the office, and all you have is a smartphone or a tablet. So you pull that out and show a detail to a client or debate something at the building site, note it perhaps crudely correct it. Then you go back to the office at your workstation and correct for good.

        This data UX swapping would like to see working better. Programs would change its interface according to the medium they are running (processor and UX input output) on to make the most of each.

        For instants, Nokia city lens+ smart phone, it may be better to find a specific place when you are already at the street address but can’t find the spot. For other purposes I would use nokia maps and a large screen.

        So there is the issue of processor but also of screen area and how to interact with data. In both cases I am using a maps application, Only one device has a camera and I can carry around. Another does not have a camera put has a large screen. A system that understands and takes advantage of the medium while transfering the information back and forth different devices according to production flows.

  3. Why not have a laptop with a small square phone device in it (rather like an iPod mini). Powered by the laptop when docked, working independently when apart. When not docked, the phone is just a phone and mini addressbook! When docked it also acts as the modem…

    • Interesting concept. A company might pick this idea up to create something good to use. And this subject, not about merging devices, but about devices that cooperates might be a subject for a next blog post 🙂

  4. I Disagree. Just give me a N900 with 2012 tech/speed and usb to plug a normal keyboard in and a big screen. Currently I lug a Atom laptop around and plug it into big peripherals at the office and home. We cant be far from replacing that can we? It doesnt need to run cad or video compression. Just a great browser and as much openness as possible.

    • Well, my article is not really about telling what device should Jolla make etc. It was more a state of the art of hybrid devices.
      You are clearly right on the non-need to run CAD of video compression SW, but there is a need of those. Thinking about photoshop, or processing the film that was shoot during the vacation ? Sometimes, you need a “powerful” device like a laptop.
      And I’m sure that a Quadcore 1.5 Ghz ARM device plugged in a big screen + keyboard + mouse will be as smooth as a PC, even for web browsing, simply because PC CPUS (Intel, AMD) have capabilities that mobile devices don’t have.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s