Can Marvell Smash the Atom?

Marvell Technology Group (MRVL) stock is way up today partly because investors had been overly pessimistic about its results for the October quarter (its fiscal third quarter of 2009). But the big news is that Marvel is going into the Netbook market. Which means competing with the Intel Atom processor. This is really cool. This is really exciting for investors, consumers, and gadget freaks.

Despite revenues that dropped to $791.0 million from $842.6 million in the 2nd quarter, Marvel kept its Net income high at $71 million and its free cash flow amazingly high at $246 million. For a in-depth summary of the results and what management said see my Marvell Analyst Conference Summary for December 2, 2008.

The Netbook market could be called the sub-notebook PC market. Netbooks don't have the full-powered Intel or AMD Turion processors or graphics capabilities of notebook computers, but they are cheaper. They are designed mainly to run a Web browser and do e-mail. I've seen a lot of false advertising recently, with one ad describing the Intel Atom processor in a netbook as "powerful."

But when you think about building a good Netbook from the ground up, any smart engineer would say, why Microsoft Windows for the operating system, and why Intel Atom for the processor. The only answer is that they are what people are used to. On the other hand consider a netbook not as a smaller PC, but as a larger smart phone. Suddenly there are many hardware and software choices, some of them found in devices like the Apple iPhone or the Blackberry Bold. And what you will find in many of these devices is Marvell semiconductor chips. And non-Microsoft operating systems like Linux.

Normally going up against Microsoft and Intel, even if you have better products, is not a very good business strategy. Intel in particular, in its competition with AMD, has shown a willingness to use heavy-handed tactics to maintain and extend its market share. Even its ability to subsidize advertisements for its products when people open catalogs or Web merchant sites insures that the playing field is not even. Should investors cheer Marvel tilting up against Intel?

The outcome is in doubt, so hedge your bets. But there is good reason to believe Marvell will at least carve out some good profitable market share, aside from their history of consistently entering new fields and turning the fields.

Marvell already has most of the intellectual property in place to build a single chip that can run the entire netbook. Netbook manufacturers will supply the display, case, and connectors; one chip shall rule them all. The cost will be far lower than Intel can compete with profitably. Intel will probably try to differentiate its products by saying they run Windows, but the reality is they don't really run Windows in the sense that a dual-core notebook processor runs Windows.

If you have not been following Marvell, you may not know what they already do. They specialize in combining digital and analog circuitry in a single chip. They got started by supplying chips that make hard drives capable of storing more data faster. They now dominate the drive chip industry. They branched out to a number of areas, including chips for printers. They also bought Intel's old XScale cell phone processor division, which I believe the Intel guys are going to be kicking themselves over in about October, 2009. Intel was losing money on the division, Marvell is making money.

Marvell also makes superfast chips for physical and wireless internet connections. They make video chips for DVD players and large screen TVs. Their Sheeva processor, combining technologies from ARM, XScale, and in house, runs super fast on super low amounts of power.

By building a netbook from the bottom up, based on a processor that is ultra-green because it was meant to run cell phones, incorporating all the necessary circuits on a single chip, Marvell will enable netbook manufacturers to sell their products for well under $300. Marvell says $200, then $100 is possible.

Will people mind not being able to run Windows? Some people will, but the Netbooks can't run the really cool Windows stuff, like games and Adobe design products. What Linux enables for netbooks is the ability to do email and Web browsing, plus low-overhead tasks designed for the netbook. The competitive advantage of open-source software running on Linux will really shine on a netbook.

With a Sheeva processor from Marvell running applications much faster than an Atom processor, and consuming far less power, it may be hard to keep the Win-tel team alive in this particular market. AMD's decision to stick to notebooks, where its graphics superiority over Intel gives it an advantage, may be just the right thing to do.

Of course, this is speculation about the future. Marvell is the David in this battle. But I have seen Marvell's slingshot, and I would not want to bet on the Goliath, Intel, in this market, even thought it dominates the field at the moment.

It should be noted that Marvell believes the netbooks based on the Marvell chip will open up the Internet to a new global market of between 1 and 2 billion people. Based on what I know about global income demographics, that sounds about right.

As always, keep diversified.