Saturday, April 26, 2008

The Coming Convergence

BURLINGAME , CALIF. -This week, Apple shocked the semiconductor industry by acquiring P.A. Semi, a 150-person chip design company, as first reported by my colleagues at (See: "Apple Buys Chip Designer.")

Opinions differ on why Steve Jobs chose to take this additional risk of vertical integration when every chip vendor on the planet is falling all over itself to sell its wares to Apple (nasdaq: AAPL - news - people ).

Semiconductor companies have figured out incredibly innovative ways to "hyper-integrate" their chips. This trend has given us products such as camera phones, music phones, video phones and phones with wireless local area networks. The most recent galvanizing event came when Apple’s iPhone hit the market, packing all those capabilities, along with Web browsing and e-mail, into a single, sleek device. Are we that far from a laptop replacement, demanded by consumers who are tired of carrying multiple gadgets?

Achieving this level of functionality in one miniaturized device means bringing together many sub-systems. Camera modules. MP3 players. Layers of memory. Baseband. GPS. And, of course, the microprocessor. The iPhone, for instance, uses a Samsung chip, which is a three-level stacked-die package containing an ARM processor and two 512-megabit mobile SDRAMs.

That chip alone accounts for approximately $76 or 30% of the component cost in the first-generation, eight-gigabyte iPhone, according to analyst firm iSuppli. (I've written about this component of the iPhone before, too. See "The iPhone's Component Ecosystem.")

Now it so happens that ARM-based processors have an exorbitant market share among mobile devices, and by building the iPhone on the ARM-core, Apple has had to work with the same quasi standard that almost every other phone maker uses.

Many other component vendors--Broadcom (nasdaq: BRCM - news - people ), Marvell, STMicroelectronics (nyse: STM - news - people ) (which just created a joint venture with NXP), Infineon, Texas Instruments (nyse: TXN - news - people ), and National Semiconductor (nyse: NSM - news - people )--also serve multiple handset vendors. The Samsung chip that Apple used also had power problems, and there was widespread speculation that Apple was planning to move to Intel (nasdaq: INTC - news - people ) in the next generation, banking on the veteran’s low-power-design excellence.

Against that backdrop comes this announcement, which most analysts read as another effort from Steve Jobs to control the differentiation elements of his product line. He already controls the operating system, which has enabled the creative, mercurial genius to offer an unprecedented user experience on both his handheld and desktop products.

Jobs’ unrelenting commitment to delighting the customer has continued to yield bounties as frustrated consumers flock from the PC to the Mac, and from Palm's (nasdaq: PALM - news - people ) Treo and Motorola's (nyse: MOT - news - people ) Razr to the iPhone.

For Jobs, therefore, it is worthwhile to take this additional risk with the hope that a successful chip breakthrough may put Apple in an unstoppable position, vis-à-vis its competitors Research in Motion (nasdaq: RIMM - news - people ), Palm, Nokia (nyse: NOK - news - people ), Samsung and Motorola.

Apple has been insanely profitable, and is now sitting on $19.4 billion in cash, having generated $4 billion in cash over the last two quarters. In fact, shareholders have expressed their discontent at Apple not doing anything significant with that stockpile.

Well, here comes the answer. An answer of Jobsian proportion.

Apple’s long-time competitors have learned that having your own captive chip division is not easy. Motorola spun off Freescale as a standalone semiconductor company. Not only has Nokia systematically moved away from developing its own chips, it doesn't want to rely on a single partner, such as Texas Instruments (TI). Instead, last year, Nokia started creating multi-sourcing relationships with ST, Infineon, Broadcom and others. Palm is gasping and Research in Motion doesn’t seem like a company that would make a move toward vertical integration.

Samsung is the only mobile handset vendor in this market that is also a significant semiconductor player. In fact, Samsung owns its own chip fabrication facilities and has considerable ability to differentiate via manufacturing excellence and cost. Having Samsung as a key vendor should make Apple nervous.

But even without Samsung, there are so many other players who would be delighted to join Apple's parade. Apple could play them against each other, then pick the one that delivers the best chips. Intel, IBM (nyse: IBM - news - people ), TI, Broadcom ... one nod from Apple, and they would all be on their knees, begging to serve the master, Jobs.

Why then, does Steve Jobs close these options and bring the chip design in-house?

The answer is, he doesn’t. He will continue to play all of them against each other, including this newly acquired unit. If the in-house team delivers a breakthrough, he wins. If they don’t, he can continue to buy his chips from the other vendors.

The $278 million is chump change compared to the option it creates for Apple. The option to differentiate not only at the design, operating system and software application levels--all of which the Macintosh, the iPhone, and the iPod already do--but at the fundamental electronics level.

The convergence device movement will march on over the next decade. The next billion consumers will likely do their "computing" through a convergence device. No, the device may not be an iPhone. In fact, most likely, a whole new product category is in Apple’s labs right now which would give us the option of leaving the laptop at home.

Dell (nasdaq: DELL - news - people ), Hewlett-Packard (nyse: HPQ - news - people ), Lenovo and the other laptop vendors will need to get their own strategies together to bring convergence devices to the market. Nokia, Research in Motion, Palm, and Samsung will not only have to respond to the iPhone, but also to this new, yet-to-be-created category.

Too much is at stake. Jobs took a risk. He created an option.

If Apple succeeds, expect to see many others in the industry scramble to recreate a vertically integrated industry, which is where it all started. WinTel will eventually be seen as a detour.


No comments: