While reading a friend’s Windows 8/ARM article on Forbes, I had what is often described as a BFO (Blinding Flash of the Obvious).

For a year now, we’ve all been theorizing (err… guessing?) what Windows 8 on ARM would look like. The biggest question has been the will they/won’t they of the Win32 desktop on ARM. Early on I thought it’d be there. Then later last year I thought maybe not. Now, I think, I’m all but certain it won’t be. This is unfortunately all still supposition, since we have no more info to go on yet – but regardless, let me explain.

On my blog yesterday, a reader named Sebastien pointed out that Gartner has expectations for over 404M PC’s to be sold this year, and that ARM tablets would be additive to this. I was doubtful, and I even commented as such. Rather than Windows 8 ARM tablets taking sales from iPads, I felt that they would nibble at the Windows 8 x86 desktop/laptop market instead. Depending on the price that the ARM tablets land at, the price x86 tablets land at, and how far WinRT apps go in meeting consumer’s day-to-day needs, this could still be the case. However unless they offer distinct value above the iPad (not counting screen size or other “specs”), they need to be price competitive to the iPad (, without carrier subsidies) in order to steal sales from it.

However, I don’t believe that Microsoft is going into this game without a game plan. I think they have a strategy to try and ensure that there is still a PC market and a tablet market for Windows, and that these markets behave distinctly, ideally with a minimal amount of “cross-cannibalism”.

Android tablets have been criticized for many things, but like the HP TouchPad, a chief complaint has been price. Windows OEMs have complained before about the cost of both Windows licenses and Intel processors. By supporting Windows on ARM, Microsoft made some of that expense disappear (the cost of the Intel processor and any additional required components), and provided a platform that focuses on delivering battery life over performance, likely saving money on cooling and battery components as well.

But what about the Windows desktop? I think it may not be on ARM tablets. I probably sound conflicted here, as less than two months ago I said:

 I strongly believe the Windows desktop (or as errantly referred to, “the desktop app”) will remain.

The Windows desktop is the Windows experience. It has been for over 20 years, and though it is “over there” (imagine me pointing to your right, where the desktop now lives – stage left from the Start page) it isn’t going away on x86/x64 desktops (or possibly x86 tablets) anytime soon.

However, if Microsoft were to hide not include the Windows desktop on Windows 8 for ARM (Win32 is still semantically there, but minimalist), it could offer OEM partners a way to cut costs there as well, by charging less for the license (which would in turn reinforce my belief that Windows 8 for ARM will be OEM only).

However, by not including the desktop, Microsoft does something further. It creates a psychological boundary. It creates the boundary that Windows 8 ARM slates are, in fact, lesser cousins to Windows 8 x86/x64 PCs. While you will be able to run your WinRT apps on either platform, and even seamlessly authenticate on both, and sync documents and settings over Windows Live between them, the ARM slate clearly becomes something you need in addition to a Windows PC if you need to continue to run Windows desktop apps. You’ll need a Windows 8 ARM slate plus a PC. All of a sudden that whole PC Plus* mantra makes sense.

Another interesting angle is that it actually doesn’t pit ARM against x86, at least not directly. ARM systems will likely wind up at the low end of the cost spectrum, and Win32 compatible and performance-focused x86/x64 desktops and laptops at the opposite end. Depending on the price points that Intel (and perhaps AMD) are able to hit with certain tablets, it could still pit x86 against ARM, but the latter needs to prove that it has the energy efficiency to go head-to-head against ARM tablets on very low-cost, lightweight devices.

Whether most customers will take Microsoft up on purchasing both an ARM tablet and an x86 PC in the short term until the WinRT platform is proven is a question, as well as whether most “touchless” desktop users will truly take advantage of Metro apps and the Start page. Regardless, if executed as such, it could come back to agree with Sebastien’s (and Gartner’s) belief that the Win8 ARM sales would be additive to the existing PC sales projections – unless too many consumers find that WinRT-based apps are good enough for all of their needs, which is a possibility, in which case ARM-based Windows 8 tablets could nibble away at the bottom of the PC market, as netbooks (and possibly the iPad itself) have done in the recent past.

This also means that I also believe the next version of Office will continue to be Win32 – needed for desktop/mobile power users, but we may well see a Metro-based version of Office that, like Windows on ARM, could cost less, be touch-focused and, like Microsoft Works used to be, much more constrained to casual productivity tasks instead of trying to shift (and shove) all of Office over to the Metro user interface, which I believe would be a mistake.

*Intriguingly, while searching for references, I ran across this article in which Bill actually coins the term “PC-Plus” back in 1999 during a similar era when analysts were questioning the PC’s future.

This shouldn’t need to be said – but everything in this post is pure conjecture on my part, and it is my personal opinion. I do not have enough information to form the factual finding that I do at my normal job in my daily writing. But I believe it is important to get this thesis out there for discussion. I hope you found it interesting.