The hardware that will define Microsoft Windows 10 isn't even out yet

Every major version of Windows had its iconic piece of hardware. Windows 95 had ever more creatively designed mini-towers. In Windows XP, the Windows Tablet became a focus though it likely was the ever thinner clamshell notebooks that really shined. Windows Vista was all over the map but you could argue that the all-in-one was its iconic piece of hardware. Windows 7 had the Intel-branded Ultrabook; a Windows spin on the MacBook Air. Windows 8 was supposed to showcase hybrids and convertibles but some really screwy marketing decisions would have you wonder if Microsoft loved or hated those form factors.

Windows 10 fixes the design problems of Windows 8 and should shine nicely with those hybrid and convertible products. But given that it is designed to span cell phones to desktop PCs, as well, I think it is hardware with that cell phone to PC range that will eventually be the showcase of this product. Intel and Microsoft may have a surprise up their sleeves because both have systems in the works that could define this product.


Microsoft’s Hololens likely represents the ultimate end for what eventually Windows will become over the next decade or so. It is a head-mounted computer with a 3D camera which can place a display on anything you see and, in its evolved form, eventually turn most anything you touch into a touchpad, touchscreen, or keyboard.

The thing is that this product is still in prototype form and likely won’t reach its stride for 2 to 3 generations suggesting that while it may eventually define a subsequent version of Windows it doesn't have an initial market that is broad enough to define Windows 10. For the most part Windows 10 will have to reside on something much less advanced until Hololens more fully matures.

Modular PC

While this is on old IBM (now Lenovo) PC concept, it does appear to play to Windows 10 strengths in being able to scale from small hardware to large screens. Initially, this concept had a core module that contained the processor (CPU), graphics chip (GPU), system memory, storage, and chipset. This module could then be plugged into a variety of devices which, at the time, included PDAs, Tablets, Laptops, and a desktop dock connected to a monitor.

As it evolved, the PDA component became the core module because that made it smaller and more acceptable as a hand held device. The closest thing we have to this under development, which we know about, right now is Google’s project Ara which likely will not be running Windows 10.

Cloud PC

Now, you could host Windows 10, and put it behind something like Dell’s Cloud Connect. This is an Android based HDMI stick that Dell is positioning as its near-perfect, forward-looking thin client offering. It is full Android but while it does scale up nicely, it really isn't designed to scale down. Plus, placing Android under Windows would likely not float Microsoft’s boat any more than running Windows 10 on Project Ara when it comes to an ideal solution. I imagine every time the tag line “the best way to run Windows is on Dell’s Android based Cloud Connect” was used, a Microsoft executive would “get his wings” (likely because he was shot by their new CEO).

However I do think that by this time next decade most of us will be largely computing via the cloud, unless you are one of the aforementioned executives. In that case, you’ll be living on one.

Intel Compute Stick

The Intel Compute Stick is very similar to Dell’s Cloud Connect but it runs full Windows natively. The price delta between the two products in a quantity of one is about $30 (£19) with the Intel part actually being less expensive.

(I started writing this thinking that the delta was $30 (£19) the other way but I just looked on Amazon and the Intel MeeGoPad Stick is as low as $122.99 (£80) while the Dell Cloud Connect stick is $151.57 (£99). And in configuration the Dell has 1GB of RAM and 8 GB of storage while the Intel part has 2 GB of RAM and 32 GB of storage. Damn…)

Anyway because the Intel device can run Windows natively, it can better scale small to large but finding a hand held display or a battery powered touch display at laptop size has proven problematic. You could scale this but, at least for now, I haven’t been able to find the parts in market that would allow you to. A best solution would be to take one of these and put it in a touch desktop display to create a low cost, upgradable, All-In-One.


Using and updating what we learned from the Modular computer it might be better to start with a Smartphone (ironically this appeared to be what ARM was suggesting when they had their analyst briefing a few days ago and spoke about PC level performance). The idea of a phone that could be docked or connected to create a scalable solution has been tossed around for a while now.

It isn’t uncommon for a Smartphone technology vendor like Qualcomm or a phone builder like Samsung to showcase that their current range of phones are capable of sending an Ultra HD signal to a TV and they certainly could then integrate with monitors at high resolutions of nearly all sizes.

Intel brought a product to market a few months ago called the PadFone X Mini that they created with Asus. This is a Smartphone that can be docked to become a small tablet. More accessories could expand this to laptop and even desktop form factors particularly with wireless connectivity built in. Intel has a technology called WiDi designed specifically to support this and there is a competing technology in use called Miracast.

With these technologies you should be able to revisit the idea of a modular computer and create what should become the iconic solution for Windows 10.

Wrapping Up:  Could ARM Get There First?

While Intel has the inside lane currently, Windows Phone (which would have to be the foundation for the solution) doesn’t run on Intel but ARM. An ARM-based solution tied to Microsoft’s various cloud services could also create this outcome and this might actually be more stable given it will be the legacy apps that will have the greatest trouble operating with this kind of size diversity. The newer apps, which generally run on both ARM and Intel should scale more easily and that suggests there will be a race to see if Intel can convince Microsoft to allow an Intel-based Windows phone before an ARM vendor fleshes out a solution.

The vendor most likely to figure this out is Lenovo because they have the old modular PC DNA, a full phone line, full PC and monitor lines, and know how to make accessories that scale. But Asus, who is often seen as more aggressive design-wise could steal into the lead. My best guess though is we likely won’t see this product emerge until Windows 10.1 because I doubt anyone is willing to take the initial risk on a product like this. Then again, Intel or Qualcomm or NVIDIA and Microsoft could pull off the upset of the century. We’ll see.