While Intel (NASDAQ:INTC) has had its share of struggles breaking into the mobile market, subsidizing its tablet chips to gain market share, and recently announcing its first mobile chip with an integrated modem, the company remains dominant in the PC chip market. During 2014, Intel's PC client group recorded $34.7 billion in revenue, with a stratospheric operating margin of 42.2%.
A major reason for this success is that Microsoft's (NASDAQ:MSFT) Windows and Apple's OS X only run on chips built on the x86 architecture, creating a duopoly in the PC chip space, with Intel controlling most of the market and Advanced Micro Devices claiming the rest.
While inexpensive Chromebooks, which can run on both x86 chips and ARM-based chips, licensed by ARM Holdings (NASDAQ:ARMH), have won a small portion of the PC market, the real threat to Intel's PC dominance comes directly from Redmond, Wash.
Will Microsoft look beyond x86 chips?
Microsoft has been pushing for Windows, particularly the upcoming Windows 10, to run on a wide variety of devices, from ARM smartphones to powerful x86 PCs.
The problem, as evidenced by the failed Windows RT (Microsoft's ARM version of Windows 8), is that applications built for the Windows desktop only work on x86 processors. The strength of Windows as a platform is derived from its vast library of desktop applications, and without support for these applications, an ARM version of Windows for PCs was destined to fail.
Windows 10 could change things, though. Microsoft is betting on universal apps, which allow developers to target any Windows 10 device, from smartphones to PCs, without having to write different versions of the application, and with the user interface adapting automatically, minimizing device-specific code.
One goal for Microsoft is to push its Windows Phone platform, which has a small user base today, by essentially creating a single, much larger Windows 10 user base across all Windows 10 devices. But another effect could be a viable ARM version of Windows for the PC, running only universal apps. In Windows 10, these universal apps would behave just like desktop apps on the PC, unlike the bifurcated experience on Windows 8, so the only challenge is building a sufficiently large library of universal apps to make up for the lack of legacy desktop applications.
This, I think, is the biggest long-term threat to Intel's dominance of the PC market. Given Microsoft's incentive to allow Windows to run on as wide a variety of devices as possible, I suspect the x86 monopoly of the PC market will eventually come to an end, although it might take many years.
If ARM PCs seem far-fetched, Microsoft has already announced that Windows 10 will support the Raspberry Pi 2, a tiny $35 ARM single-board computer. The fact that Windows can run on such a device is amazing enough, but the real story is that the Raspberry Pi 2 is essentially a $35 Windows PC. It won't support legacy desktop applications, of course, but putting Windows on devices such as the Pi should help sell developers on universal apps.
Even if universal apps prove wildly successful, support for legacy desktop applications still gives Intel a major advantage. But over time, as the universal app ecosystem grows, this advantage will disappear.
Over the long term
This isn't a near-term threat to Intel, and it depends on Microsoft convincing developers to create universal apps, something that is far from guaranteed. But Microsoft has no good reason to allow the x86 monopoly to continue, and in the long term Intel might not be able to maintain its outsized PC margins if universal apps take off.
We'll have to wait for the launch of Windows 10 later this year to see whether Microsoft's write once, run everywhere universal apps live up to expectations. If Microsoft can get developers on board, ARM-based Windows PCs will eventually become a viable option. That would be bad news for Intel.