|
Atom could be Intel's little engine that could
Commentary: But chip maker's bet on low-cost 'NetBooks' may prove costly
By Therese Poletti, MarketWatch
Last Update: 12:01 AM ET Aug 21, 2008
SAN FRANCISCO (MarketWatch) - At Intel Corp.'s big developer conference this week, the chip giant was extolling the virtues of its newest little chip called the Atom.
The Atom has surprised both company executives and analysts with its popularity among hardware makers. The chip was introduced in March and is aimed at an emerging market of very low-cost mobile devices, especially in developing countries.
However, for a company like Intel INTC, the world's largest chipmaker with a stake of more than 80% of the personal computer market, the question that comes to mind is how much the Atom could eat into sales of the company's other, more profitable products.
Nathan Brookwood, an analyst with Insight 64, voiced the question on the minds of many who follow the company, including some on Wall Street. "If Atom is as good as you guys say it is, isn't it in danger of cannibalizing?" Brookwood asked executives on Tuesday.
The Atom is helping foster this new category of portable devices, which Intel calls mobile Internet devices, or the lovely acronym MIDs. One product area is NetBooks, which are essentially smaller laptop computers with fewer capabilities than a full machine but offer increased functionality than a smart phone.
These devices have a similar ring to the idea of an "Internet appliance" - an idea from Silicon Valley's more storied past. They are designed mostly for one function: browsing the Internet.
Executives said Tuesday that Intel has 700 "customer engagements" with hardware makers who are looking at developing devices based on the Atom. "We are pleasantly surprised," said Intel senior vice president Pat Gelsinger, in an interview. Computer makers like ASUS Computer International, which specializes in low-cost notebooks, have been the initial buyers of the chip, but bigger players such as Dell Inc. DELL are also believed to be looking at the Atom.
Big factors driving interest in the chip is its low power consumption, lower-cost and ability to run software that is compatible with Intel's standard chip architecture, known as the x86. And even though the chip is inexpensive, it provides decent margins for Intel, and customers can use it to create Internet surfing devices that sell for as low as about $200. With the efficiencies resulting from its latest manufacturing process, Intel can yield about 2,500 Atom chips from one silicon wafer, compared with a relative yield of only several hundred units of its larger chips.
But the Atom, as Intel is quick to point out, is not as powerful a processor as its other chip families. The capabilities of its higher-end chips based on its Core architecture run the gamut from enabling life-like video games to powering supercomputers to basic office multi-tasking, such as watching a YouTube video and running several other applications at once.
"If you buy a $10,000 car and you expect it to behave like a BMW, it's not," said Dadi Perlmutter, executive vice president and general manager of Intel's mobility group.
But the risk for Intel is that a growing slice of the market may find that a $10,000 car is all they need. The bulk of growth for the PC sector today is coming from developing countries, which are passing over desktop computers in favor of smaller, cheaper laptops.
Indeed Intel executives admitted that they would not be working with 80% of their Atom customers if they did not have their cool little chip.
George Shiffler, an analyst with Gartner, questions whether there is a real market to be had. "They are kind of a tweener," he said. "They are a bit bulkier than a smart phone, but they are not quite a low-end laptop. The question is how are they going to be perceived (by consumers)...Lots of these tweeners have been tried before, history is littered with them."
Intel clearly has big plans for the Atom. It hopes the chip will seat itself next in cable set-top top boxes, and eventually smart phones, although Gelsinger said it's not quite there yet.
"We need to take the thermal envelope down," he said, referring to the heat dissipated by the chip. In one or two more generations, the Atom could be smart-phone ready.
If Intel succeeds with the Atom, it could make an interesting business school study on a company dealing with the classic "innovator's dilemma," a conundrum faced by many technology companies, as well as other industries. Leaders in their market often get caught off-guard by new disruptive technologies that are typically cheaper and "good enough."
The concept, described by Harvard Business School professor Clayton Christensen in his classic text "The Innovators Dilemma," refers to when a "lower-end disruptive innovation is aimed at mainstream customers who were ignored by established companies." One rare example of a company that managed to stave off a serious competitive threat was IBM Corp. IBM and its successful development of the personal computer, when upstarts like Apple Inc. AAPL and other hobbyists were inventing computers for everyone.
Intel has gone about the Atom in a classic innovator manner. It created a smaller design team in 2004 as part of a larger design team working in Austin, Texas, far from its usual development teams in Israel, Oregon and Santa Clara, Calif. The team designed the Atom chip architecture from the ground up to be a low-cost, low-power device, not a typically powerful chip that was scaled down to do less.
The company has been around this block before as well, with its XScale architecture based on Arm's design, which it tried to market as a cell phone processor. The company ended up selling the XScale business to Marvell Technology Group Ltd. MRVL for $600 million in 2006.
It will be interesting to see what Intel has learned this time around, and if the Atom will be the little engine that could.
|