Here’s a hypothetical: What’s the distinction between a smartphone, a MID, a PND, and a UMPC?
Posing this type of question invites a geek fight. We who follow the smartphone space and remember the personal digital assistant (PDA) know it’s a fool’s errand to apply strict categorical definitions to a device segment that morphs in near real-time.
Here’s another, far geekier hypothetical: Will RISC giant ARM or x86 giant Intel’s processors dominate a burgeoning universe of mobile broadband connected “stuff”? Said differently, if the smartphone is a computer, what’s the right processing engine for the job?
Who cares? The answer to this has massive implications for the global development, silicon, software, content, and applications communities, and will determine how far we as a mobile species can evolve. Ironically, the question is likely to be settled in the U.S. market first, for a change.
What do we know?
There is much that we know about handheld devices in the U.S. market. We expect just over 160 million handsets to have been sold in 2007. Around 10 million will be “smartphones,” and around 90% will support relatively advanced multimedia feature sets. Almost 100% are ARM-based. They run a variety of “open” and proprietary operating systems, browsers, and applications on chipsets from Qualcomm, Texas Instruments, Freescale, EMP, NXP, Infineon, Marvell, and other “usual suspects.” Virtually none run Intel or AMD architectures.
But handsets in the U.S. are “supplied.” Only recently has Apple demonstrated that people will “demand” ultra cool, ultra overpriced phones.
Among devices that are clearly “demanded,” we expect 25 million to 30 million portable media players to sell in the U.S. in 2007. Add to these 30 million portable gaming devices, and a few million portable navigation devices . and 14 UMPCs (ha ha). Most, but not all of these are also ARM-based. Higher-end portable devices and the laptop space belong to Intel and AMD (and ATI and NVidia), and of course run Microsoft’s Windows XP and Vista, Apple’s OSX, and occasionally Linux.
Uncertainty from Internet Insurgents
Apple, Google, and Amazon have given us an indication of how far more heterogeneous devices will be brought to market. The same can be said of H3G UK’s Skype phone from Amoi. But as networks “open,” the nature of connected computing and communications devices becomes very difficult to characterize and harder to quantify.
Who could have predicted the Kindle? Reviews and volumes aside, the device is symbolic of the way “Internet insurgents” will play in the U.S. 3G market with alternate business models and device categories. There will be more. Some strategies will be application centric, riding on incumbent CE or cellular devices. There will be killer Kindles or Kindle killers for which we have no industry parlance.
Root your analysis in assumptions around business models, and not widget counts. Projecting the size of this yet-unknown market is a complex function of network, processing, use case, utility, pricing, business model and other factors. But first and fundamentally, it’s premised on an analysis of business models, motivations and likely outcomes. With the Kindle, Amazon puts a check in the device and distribution boxes. It already has the content. A limited number of market participants can do this. The ones with the advantage have cloud-based business models and are not incumbent CE or cellular names.
So what’s this about ARM and Intel?
Networks will open to alternate devices. Many yet-unknown gaming, Internet, portable entertainment, computing, and mobile productivity devices aspire to greater functionality than today’s ARM-based phones at price points lower than today’s UMPCs can achieve on Intel Architectures.
So there’s a cost, performance, and incumbency void between ARM vs e.g. Intel Architectures. This implies a void between what Symbian, Windows Mobile, and RIM, can achieve and what Vista and OSX can achieve. Product roadmaps reflect steps toward capturing this space. This is the most immediate opportunity for mobile Linux-based offerings as developer platforms in the mobile space. Intel’s Moblin project and Google’s Android are bets on this opportunity.
The hardware/software architectural strategies are evident. As usual, it’s not zero-sum, but we look for Linux-based open software stacks to play significantly on all processing architectures in the new generation of connected mobile computing devices.
The MID is dead. Long live the MID.
John Jackson is the VP of Yankee Group’s Enabling Technologies group. He supervises Yankee Group’s global handheld and handset forecasting. For more information, visit www.yankeegroup.com
