How To Avoid A Wintel Mobility Disaster Of Your Own

This week saw a dust-up between Microsoft and Intel about claims for application compatibility on smartphones running Windows on ARM processors versus Windows on Intel Atom processors. The dispute demonstrates just how late Intel is to the smartphone and tablet markets, and the ripple effect on Microsoft's already tenuous plans for smartphones and tablets.

Just like most everyone else in the industry, both Intel and Microsoft got the tablet market wrong. They saw tablets as little PCs rather than big smartphones. If you need an example of just how wrong a design can go, I offer the Avaya Flare, which for $1,500 weighs more than 3 pounds and looks more like a clunky Walmart LCD picture frame than anything else.

Even Apple didn't fully appreciate what the tablet market would become. It supposedly first developed the iPad and then shrunk it to be the iPhone. But the genius is that the two are much alike, sharing common design, operating system, and apps. They're intended to mostly retrieve information, with very little ability to input large amounts of complex data. The type of information retrieved, need for portability, and the duration of viewing tends to dictate the size.

It appears that both Intel and Microsoft followed their own desires rather than those of their customers. Corporate groupthink can be an incredibly hard thing to overcome, even in the face of the success of the iPad and the myriad failures of Windows-based tablets. It took threats from large OEM partners to convince Microsoft's top brass that Windows 7 wouldn't cut it as is on tablets. Intel's failure to grab onto the smartphone and tablet markets is even harder to fathom.

The enabling technology for smartphones is the system on chip (SOC), which attempts to combine as many aspects of a computer system as possible onto one piece of silicon. By combining cores, memory controllers, cache memory, graphics, and more, the chip allows for the vast majority of the device's work to be done in the chip's lower-power-consumption confines. The stuff off-chip includes radios, display drivers, audio amplifiers, GPS-everything you need to interface with the outside world. This is familiar ground for Intel, which understands these systems well and should never have ceded the market to the Qualcomms and others that have made their fortunes in dumb cell phones.

The Compatibility Problem

This brings us back to the scuffle this week, which centered on Windows 8. Microsoft initially said Windows 8 would only be a 64-bit operating system, and then when it looked around for 64-bit chips that could be the basis for power-sipping phones and tablets, it found none and changed its mind, saying Win8 would support 32-bit ARM chips on those portable devices.

At a recent Intel investor conference, a company exec apparently said more than he should have or more than he knew when he said that app compatibility would be had only with Win8 64 bit running on Atom chips. The exec went on to say there would be no compatibility possible with ARM chips.

The Intel exec's statement is probably accurate, depending on which set of apps you think should be able to run on a Win8 phone or tablet. Be that as it may, Microsoft said the statement is inaccurate.

It's not hard to see why Microsoft is getting comfortable with ARM. The Wintel beer and pizza party is no fun if Intel is two years late bringing the beer. Windows Phone 7 got a major update this week, adding some 500 features. It's a clear signal that Microsoft intends to continue to develop WP7 at least for phones if not for tablets.

But in so doing, it's also clear that there's a different code path for Windows 7 than the one for Windows Phone 7, and those will probably continue into their next versions. So when you talk about compatibility, you have to ask: Compatibility with what?

On phones, I think it's pretty clear-there's supposedly more than 10,000 WP7 apps now, and those will work on WP8 whenever it arrives. There are hundreds of thousands of Win7 (and previous) apps, and Win8 will probably run only 64-bit versions of those, making obsolete those apps that haven't been recompiled-including many apps IT shops developed themselves. On laptops and desktops, that means a long transition to Win8, just as it has been with any Microsoft operating system.

The place where it all goes to hell is on tablets. As I stated earlier, recent history tells us that users want tablets to act more like big phones than little laptops. Sure, Microsoft can create a version of Win8 that has the right user interface on Atom chips, but the compatibility users want is with apps that understand how to use the six-way gyroscope, GPS, touch screen, and other features that make smartphones smart. Those apps run on a WP7-ARM combo. Running big fat Wintel applications on those devices, at least at this point, doesn't appear to be all that important.

While there's no doubt that it would serve Microsoft and its customers well if Redmond laid out a plan for all of its client devices and corresponding app compatibility, the commotion should get enterprise IT architects thinking, too. Hopefully, as you've made your transition to Windows 7, you've taken the time to catalog the apps your IT team runs now and must run into the future. If you haven't decided to either Webify apps or bring them back into the data center to be accessed through some sort of virtual desktop, now's the time to do it.

No matter how you look at it, there will be no Windows XP-like monopoly in your future. Users will want to access apps from a variety of devices and operating systems. My mantra: Webify if you can, VDI if you must, and avoid platform-specific programming like the plague.

As for Microsoft and Intel, that they've given Google and Apple and ARM this big of an opening will be one of the biggest mistakes these two giants have ever made.

Art Wittmann is director of InformationWeek Analytics, a portfolio of decision-support tools and analyst reports. You can write to him at [email protected].

To find out more about Art Wittmann, please visit his page.

More than 100 major reports will be released this year. Sign up or upgrade your InformationWeek Analytics membership.

InformationWeek Analytics is conducting a survey to determine the satisfaction with mobile operating systems vendors. Respond to the survey and be eligible to win an iPad. Take the survey now. Survey ends May 27.