OEMs have the power to make or break smaller Windows tablets

Consumers seem to prefer smaller tablets, and its down to the OEMs to try and turn out Windows tablets that work well at this smaller scale. But OEMs could muck it up for everyone.

Two ARM-based tablets — an iPad mini and a Nexus 7. Is the future of Windows tablets more this size?

When I first heard that Microsoft was planning to put their weight behind smaller tablets, I was initially happy. Having owned a few full-sized iPads, an iPad mini and a Nexus 7, the smaller tablets definitely make more sense than the larger ones.

It had even struck me as unlucky that Microsoft had launched the 10 inch Surfaces just as Apple seemed to be  validating the idea of a smaller tablet .

And then the horror set in.

OEMs, with their propensity to just throw rubbish at the market without any proper thought, could build some of the worst tablet hardware known to mankind. A small-scale Windows tablet just increases the chances of us all being lumbered with terrible products.

Let's look at some mistakes that they can avoid.

Windows RT

The better operating system for a tablet — by a country mile — is Windows RT . This is for no other reason than because Windows RT was specifically designed to run on low-power, high-portability devices, which is exactly what a small-scale Windows tablet is.

But there's a problem here. Whereas any Tom, Dick, or Harry can build an x86-based Windows box, Windows RT devices can only be built by very special partners of Microsoft. So if you're Dell, Lenovo, Samsung, or Asus, you can build them. But if you're a couple of lads working out of an industrial estate in the middle of nowhere — no Windows RT for you.

We know that Windows RT devices haven't been selling, and that Microsoft's special Windows RT partners have been complaining. Dell has been saying that demand for Windows RT has been weaker than had been hoped; the same low demand has kept Samsung from releasing Windows RT in Germany.

OEMs that have rights to build Windows RT devices might not fancy the risk of further development with questionable rewards. Therefore, both those OEMs and ones without rights to build Windows RT devices might feel that an x86-based device might be less risky.

And this is where they could really screw things up.


If you are an individual responsible for releasing onto the market a small Windows tablet with active cooling (ie, a fan), then — well, I've failed to come up with a pithy insult that's fit to print.

An actively-cooled, small-scale Windows tablet would have to be the most heinous computing product — no, the most heinous product of any kind — ever to be foisted upon an unsuspected market.

I can just imagine sitting in bed unable to hold the tablet whilst watching the latest Breaking Bad because it was so hot, and also being unable to hear the dialogue because the fan was going. Or having to have it plugged into the mains because it drew so much power.

Tablets have to have passive cooling — no arguments, no question. A tablet with a fan and heat vents will be an utterly rubbish abomination of a thing.

(Yes, I know Surface Pro is a "tablet" that has passive cooling. But that's not a tablet, that's a hybrid PC, and we'll get to that.)

If you're anything like me, you find keeping up with processor product roadmaps somewhat mind-numbing. However, we do need to talk about them.

ARM processors are much preferred by manufacturers of smartphones and tablets because they were designed specifically to consume small amounts of power and give off small amounts of heat. You can keep your smartphone in your pocket without a) your leg melting, or b) your trousers catching fire. (Or handbag/purse, or whatever.)

The x86 chips are designed for a world where you usually have mains power available. Therefore, you get loads of processor oomph, but you also soak up loads of energy and give out loads of heat. The x86 makes sense if you're building desktops, laptops, and servers.

What's wrong-footed Intel is that x86 was looking great until the dominant form of computing needed to be low-powered, low-energy/low-heat. Now Intel is trying to pivot madly to generate products that behave more like ARM chips, but whilst still being x86 chips. The existence of Windows RT at all demonstrates a lack of confidence from Redmond that Intel could turn their appraoch around. It now appears that Intel might be able to do this, but perhaps not for some time.

The problem that needs to be solved here is that x86-chips that can be passively cooled might be "jam tomorrow". Although Intel has proven it can create an Atom-based smartphone in the Lava Xolo x900 (by most accounts, a pretty decent smartphone), it's only in the next rumoured/leaked set of chips —  Haswell  — that a proper, passive-cooled tablet will be possible.

My concern is that OEMs might find themselves in a position where they're compelled to release tablets onto the market before passively-cooled x86 systems are easy to develop.  

No keyboard

Another way in which the OEMs have disappointed the markets over the past six months is in their insistence that "hybrid PCs are better than tablets".

Ignoring whether hybrids are better than tablets or not (spoiler: They're not), OEMs have been lucky so far in that when they started, the de facto size of a tablet was about 10 inch, and as a result, putting a keyboard on one wasn't that terrible an idea.

That same trick with a 7 inch tablet isn't going to work. A small-scale Windows has to be touch only. No keyboard, no mouse. Whilst you're at it, you can complete my pipedream and not put the Old Windows desktop or Office on there, too.

The only people who typically have hands the size of a five-year old human are five-year old humans. If you want that to be your target market, go right ahead.


What the OEMs generally managed to forgo with the transition to the re-imagined world of Windows 8 was any sort of re-imagining of what they were building. All they did was say to themselves: "Yay! Let's build hybrid PCs!"

What I don't understand about "hybrid" is why, as technologists, we hear this word and think "cool". We should hear this work and think "oh, yuck!"

Hybridity means one thing — specifically, it means "compromise". The idea of hybrid anything should bring us out in hives.

Tablets — like the iPad or the Nexus devices — work because they are not PCs. They have strengths compared to the PC, and weaknesses as well. Conversely, a PC isn't a tablet, so it has strengths compared to a tablet, and also weaknesses.

If you just take those two and mash them together — ie, you create a "hybrid PC" — you don't magically get a device that's all the good bits of both.

My concern is that the OEMs will hear "small-scale Windows tablet", take the products that they have and make them smaller. Specifically, what they'll make is smaller hybrid PCs.

That would be a terrible state of affairs. It really is time now for OEMs to start applying love and thought to what they do.

Mind you, it was also "that time" six months ago... 

What do you think? Post a comment, or talk to me on Twitter: @mbrit.