For many years, the market for personal computing equipment was dominated by the Wintel "duopoly." But no more.
In practice, Wintel wasn't so much two companies scheming together as it was a marriage of convenience pairing two dominant technologies: Microsoft's Windows operating system and Intel's x86 processor family.
It wasn't an exclusive relationship. Microsoft always had Advanced Micro Devices and flirted with chip designs such as MIPS, Alpha, and PowerPC in the early days of Windows NT. And Intel encouraged other operating systems, notably Linux. But the marriage stayed largely intact as the two product families yielded much of the PC industry's profit, and the companies' fortunes remain closely linked today.
But now the companies are eager to grow beyond Wintel. Two major conferences this week--Microsoft's Build for showcasing Windows 8 and the hardware-focused Intel Developer Forum--show just how important the effort is. Microsoft and Intel, which as allies haven't made so much as a blip in the mobile device market, are trying to become relevant there through partnerships with companies that already have a mobile presence.
Why are they moving to an open marriage? In a word, Apple.
First with its iPhone and then with its iPad, Apple is pioneering categories of mobile computing where Intel and Microsoft are weak. Consumers are buying millions of the products even with newer models on the way, and Apple is socking away gargantuan piles of cash.
On the operating system side of the equation, Google has built Android into a viable smartphone competitor to Apple's iOS. Meanwhile, Microsoft's Windows Phone still lags and Windows 8 for tablets won't arrive until 2012. On the processor side, virtually all mobile devices, including all of Apple's, use chips based on ARM's designs.
Seeing each other's competitive weaknesses, Intel and Microsoft are leaping into the arms of each other's rivals.
At the Intel Developer Forum in San Francisco, Intel Chief Executive Paul Otellini shared the stage with Google's Android chief, Andy Rubin, to announce a "continuation" of the two companies' "strategic alliance." As evidence of how swimmingly Intel and Google get along, Intel showed off an Android tablet running on a next-gen "Medfield" chip designed to compete with ARM in tablets and phones.
"We're going to collaborate very closely to make sure Android is optimized the best it possibly can be for the Intel architecture," Rubin, who is Google's senior vice president of mobile, said on stage. "Going forward, all future releases of Android will be optimized everywhere from the very low levels--at the kernel, taking advantage of memory management and all these great features of the low-power IA [Intel] architecture, all the way up to multimedia, 3D graphics--everything that's part of a system-on-a-chip today."
Added Otellini, "Every time we collaborate with Google, good things come out of it. I'm excited and have high expectations around this."
At the same time, Windows division president Steven Sinofsky spent a lot of time at the Build conference in Anaheim, Calif., touting Windows 8 running on ARM processors. The freebie tablets for Build attendees used Intel chips, to be sure, but Microsoft spent a lot of time talking about how new-era programming methods to build Windows 8 "Metro"-style apps will produce software that can run on either ARM or Intel systems.
"Everything we showcased today at Build also runs on the ARM-based Windows PCs being created by ARM partners and PC manufacturers," Sinofsky said in a blog post Tuesday. "Windows 8 running on ARM will ultimately be available with ARM-based hardware that you can purchase...The new development tools enable you to start today to build Metro-style applications that will seamlessly run on x86 (32 and 64 bit) or ARM architectures."
ARM isn't a single chip--companies including Qualcomm, Texas Instruments, Motorola, and Apple all make their own varieties--but Windows "allows this uniqueness to shine through," Sinofsky said. For easy cross-platform programming, Microsoft likes the two approaches--first, the combination of HTML5 and related Web technologies and second, its own Silverlight software. But new developer tools will make it possible even for programmers who use the lower-level C and C++ languages to build cross-platform Metro apps, he added.
Notice a trend here? The cross-platform world is a lot more complicated. PC makers will have more engineering choices to make and more component supply headaches. Software developers will have more building and testing to do.
Some of this complexity is the product of the horizontally integrated PC industry. There's nobody in charge, and multiple companies have influence over the processors, memory, storage hardware, graphics, batteries, screens, software, and overall design.
That can lead to a market fluidity that can be healthy, for example when it's time to alter course. But it also means a lot of potential confusion for customers.
In the Apple world, by comparison, there are plenty of component suppliers, but it's clear that one company is the boss. Apple has some complexities--for example, the split between Mac OS X running on Intel-based Macs and iOS running on mobile devices. But overall, the company's tightly integrated products simplify choices for programmers and customers. That means they won't get to buy, say, an iPhone with a hardware keyboard or a less expensive Mac Pro Lite, but it also means they won't be faced with a choice of 30 different laptops when they walk into Best Buy or 20 different Android phones at Carphone Warehouse.
Ideally, customers won't have to worry about which chip is in their electronics gadget du jour. But in practice, it will be a factor. Legacy Windows apps, for example, won't arrive on ARM, and it's not clear yet how much ARM's advantage with low power consumption will take a toll on processing power.
Complexity isn't all bad, though. When markets have vibrant competition and customers are willing to change course, then incumbent powers must genuinely improve products rather than complacently rely on momentum to prop up their businesses. (At least, until the incumbents squash the upstarts with patent-infringement suits.)
Wintel, WARM, and Andrintel
The Wintel alliance lives on, of course, even as partnerships open up. Even though Microsoft is serious about ARM and Intel is serious about Android, a large industry remains devoted to traditional PCs, and a large number of people will continue to buy and use them. When Intel wants to make USB 3.0 a reality, it seeks help from Microsoft. When Microsoft wants chip features for virtualization or multimedia, it talks to Intel. And from a logistical point of view, there's a mature network of partners that know how to work together to build a PC.
In short, while a lot of the new market growth may be with tablets and smartphones, it would be wrong to think of Wintel as headed for divorce. But if Intel and Microsoft are successful in their attempts to penetrate the new markets, Wintel will have to make room for WARM and Andrintel.