Why Microsoft Needs Intel's Haswell CPU
Mark is a member of The Motley Fool Blog Network -- entries represent the personal opinion of the blogger and are not formally edited.
Microsoft's (NASDAQ: MSFT) Surface Pro has arrived, to mixed reviews and criticisms of its weight and compromises. I expect this to change once Intel's (NASDAQ: INTC) next-generation processor, code-named Haswell, arrives later this year. Haswell will allow Surface Pro to be as light and compact as ARM processor tablets such as the Surface RT. It will also render Surface RT pointless.
A surprising lack of enthusiasm
I've actually been surprised by the critical response to the Surface Pro, which runs the Windows 8 Pro OS on the latest gen Intel Core processors (code named Ivy Bridge). This is the first genuine tablet to run a 64-bit OS, which means it's capable of just doing just about anything a desktop can do. Running Windows 8 also endows this Surface with the ability to run Windows 7 software, which Surface RT can't do.
Despite the critical reception, I expect Surface Pro to do far better than Surface RT, despite its higher price tag. According to IDC, only 900,000 Surface RT tablets were shipped in Q4 2012, giving it a worldwide tablet market share of 1.7%, behind the Barnes and Noble Nook, whose market share collapsed from 4.6% in Q4 2011 to 1.9% in Q4 2012. I believe the value proposition for the Surface Pro is much more compelling than for Surface RT. Surface Pro will appeal to the millions of Windows users who want a tablet, but not an iPad, which they consider merely a toy.
Surface Pro will join the growing ranks of Windows 8 tablets and Ultrabooks (all of which will be touchscreen-based from now on), which I expect to propel Windows 8 licensing in the coming quarters. Unlike RT, Windows 8 is off to a strong start, with over 60 million licenses sold as of Microsoft's earnings report on Jan. 24.
The multiplatform strategy
When Microsoft announced the Windows on ARM initiative at CES two years ago, it seemed like a good idea at the time: provide full Windows capability on devices as light and portable as iOS or Android tablets, which all use ARM processors. Here, I referring generically to any processor or SOC (System on Chip) that uses ARM technology. ARM Holdings (NASDAQ: ARMH) doesn't actually fabricate anything, but licenses low power processor designs to the likes of Samsung and Apple.
At the time of the CES announcement, Intel processors simply couldn't compete in this arena. Eventually, Windows on ARM was subsumed in Microsoft's overhaul of Windows, to make it a better-cross platform OS for tablets, phones and PCs. The company foresaw a future of ubiquitous touchscreens, and the result was three closely related operating systems: Windows 8 for Intel, Windows RT for ARM, and Windows Phone 8 for ARM-based smartphones. While the three provide much commonality of user experience, they couldn't be perfectly or completely converged, especially in the area of backward compatibility with existing Windows apps.
Overtaken by events
Often in the technology business, strategies that are based on an existing set of technologies become OBE'd (overtaken by events). Microsoft's platform strategy is a case in point. Here, Intel is the culprit.
Intel hasn't been idle these past two years, and has continued its strategy of shrinking chip sizes and power consumption. Intel's most advanced chip fabrication process is designated 22 nm, because that's the typical feature size in billionths of a meter. This is already smaller than the typical ARM processor, where the feature size is typically 28-32 nm, and Intel continues to refine their designs to incorporate more energy-saving abilities.
Haswell will be Intel's latest processor family to use the 22 nm process, but will be even more energy-efficient than previous generations, while at the same time being more computationally powerful. Most importantly, the Haswell core processor design is intended to be scalable for use in everything from handheld devices to servers.
With Haswell, Intel based tablets will be comparable to ARM tablets in power consumption and battery life, while being faster and more powerful. Imagine a Surface Pro with the Surface RT form factor, and you get the impact of Haswell. Intel is also widely expected to move the Atom embedded processor to the 22 nm process, which should enable even lower-power devices such as smartphones and “phablets.” These new 22 nm SOCs will certainly get the benefit of Haswell innovations and be competitive with ARM processors in energy efficiency and battery life. These developments have the potential to put a dent in the growth of ARM-based tablets and smartphones -- and potentially in ARM's revenue stream from licenses -- as device makers are attracted to back to Wintel.
This means that the fundamental raison d'etre for Microsoft's ARM strategy goes away. Lately, I've been splashing a lot of cold water on expectations for Microsoft's ARM initiatives, both for tablets and phones. Nothing that happened last quarter by way of Windows Phone or Windows RT sales persuades me that the initiatives will be successful; in fact, their success or failure is starting to look moot. The ARM initiatives are about to be OBE'd by Intel's low power processors and SOCs.
Why fight the future?
Once Haswell arrives and finds its way into the next generation of Surface Pro, one has to ask whether there's any point to Surface RT. Given Intel's history of commoditizing processors, there probably won't even by a price advantage for Surface RT. I expect the current generation of Surface Pro to cannibalize Surface RT sales. Even if Surface RT manages to survive this, Haswell will almost certainly finish it off.
More importantly, Haswell provides for a truly converged and consistent user experience across platforms that the current combination of Windows 8/RT can only partially deliver. Looking a little past Haswell into the future, the arrival of 22 nm Intel SOCs for smartphones will further expand the reach of Windows 8. Given the economies of scale attendant with developing and maintaining a single OS that can address the smartphone, tablet and PC markets, it's hard to see how Windows RT and Windows Phone 8 survive, even if they do achieve a measure of success this year.
Given this future, the allocation of resources, development and marketing, to ARM-based devices seems misguided, doubly so given the lack of meaningful progress to date. Instead of fighting the last war for ARM-based mobile devices, Microsoft needs to prepare for the next war for ubiquitous, touch enabled commodity Intel devices, for which Windows 8 is a good start.
MarkHibben has no position in any stocks mentioned. The Motley Fool recommends Intel. The Motley Fool owns shares of Intel and Microsoft. Try any of our Foolish newsletter services free for 30 days. We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy. Is this post wrong? Click here. Think you can do better? Join us and write your own!