Biz Briefs: A New Brain For Intel

The last thing you would expect to hear Paul Otellini praising is an Apple product. Otellini, 54, is the incoming CEO at Intel, the chipmaker that along with Microsoft has ruled the PC world for much of the past 20 years and has pushed the Macintosh platform to the fringes of market share. Yet when Otellini outlines his company’s new strategy, the first product

he mentions is Steve Jobs’ best-selling MP3 player. “What is the iPod?” Otellini asks, and his answer sounds strange from the mouth of a man with the well-manicured looks of a successful accountant. “It’s my music machine, man. That’s what you want. This,” and here he gestures to a laptop across the conference room at Intel headquarters, “is my content machine. That [desktop] PC is my productivity machine. You have to start by thinking about the things people want to do with computers and work backward.”

That may sound like a simple enough statement, but it represents a profound revolution in the way the Santa Clara, Calif., chipmaker–long the powerhouse of Silicon Valley–does business. Forty years ago this April, Intel co-founder Gordon Moore predicted that given advances in transistor miniaturization, computer processors should double in speed every 18 months. Not only did Moore’s law become the most trustworthy truism in technology, it was also the rock on which all Intel marketing was founded. Why did you need a PC with an Intel Pentium II processor? Because it was four times as fast as your poor outmoded Pentium I. And so the product cycle continued.

Until last year, that is, when Intel delayed production of its latest Pentium 4 chip and scaled back its proposed speed from 4 GHz to 3.8 GHz. That was partly owing to technical complications; pack too many transistors onto a microchip, and you have magnetic resistance and overheating issues that require bulkier fans and suck up more battery life in your laptop. But the bigger problem is simply that most of us no longer have such a need for speed, at least when it comes to everyday applications such as e-mail, Web browsers and spreadsheets, which work just as well at 1 GHz as at 2 GHz.

For Otellini, who will become CEO in May, that reality has the makings of a crisis. And in fact the Pentium 4 issue was only one of a whole host of mishaps and missteps that Intel found itself confronted with in 2004. The LCOS (liquid crystal on silicon) chip for high-definition TVs, a pet project of Otellini’s that (as president and COO) he had announced with much fanfare in January 2004, was abandoned in November when the cost of production became prohibitive. Waggish engineers made a disco ball out of defunct LCOS chips for Intel’s holiday party. The company also blundered on its pricing of flash-memory chips, the kind found in cameras and the smallest of MP3 players; in both flash memory and chips for servers, Intel found itself losing ground to its biggest rival, Advanced Micro Devices (AMD).

Perhaps more wounding was the fact that longtime allies seemed to be abandoning Intel. Microsoft announced that it would turn to IBM for the chips for its next video-games console, the Xbox 2, though it was Intel x86 chips that powered the original Xbox. Kevin Rollins, the new CEO of Dell–the world’s biggest manufacturer of Intel PCs–mused publicly about the possibility of switching to AMD chips. (Rollins has since decided to stick with Intel.) Craig Barrett, the current Intel CEO, who will step down in May, went into mea culpa mode. “This is not the Intel we know,” he wrote in an e-mail to employees. At an industry conference last year, he literally got down on his knees and begged forgiveness. “We don’t sugarcoat our issues,” says Barrett. “If you operate at the leading edge, you can’t be afraid of failure.”

Failure, of course, is not what Otellini has in mind. So how does he see Intel handling this raft of new challenges? For most companies, losing the CEO (Barrett must step down at age 65, according to Intel policy) would only add to the crisis. But Intel has a long history of smooth transitions from one leader to the next, and Otellini has been the heir apparent for more than two years. “Bob [Noyce] was the consummate entrepreneur,” says Otellini, describing the company’s founding chief. “Gordon [Moore] was the genius. Andy [Grove] was the management guru. Craig [Barrett]’s legacy was building our manufacturing facilities in the middle of a downturn.” Of himself, Otellini says, “I’m the product guy.” The implication: clever product design and planning can help guide Intel out of the wilderness.

So far, at least, that emphasis has worked wonders. Otellini, the first nonengineer to helm Intel, has been stressing consumer-friendly products over speedier chips in his speeches for the past four years (he calls the strategy by the awkward name “platformization”). He put the plan to work in 2003 with another of his pet projects–the Centrino–a set of chips specifically designed for wi-fi-enabled laptops. For wi-fi capability, all you really need is the Pentium M, the chip at the heart of Centrino, but Otellini wanted to sell a bundle of chips along with it that would help maintain a laptop’s battery life (not to mention Intel’s bottom line). Against the wishes of his engineers, Otellini held back the launch of Centrino until the full chip set was ready. The plan worked. Intel has sold more than $5 billion worth of Centrino chips around the world, helping the company hit a record $34.2 billion in revenue by the end of 2004, missteps notwithstanding.

Today, according to Mercury Research, Intel chips are inside 87% of laptop PCs. And in February 2005, Centrino got an upgrade to help it run music and graphics better–stepping onto graphics chipmaker NVidia’s turf. With successes like that, it’s no accident that Otellini is respected by Intel insiders as a steady hand–a welcome change in a company famous for its bitter boardroom battles. “In the Andy Grove era, it was very raucous,” says Andy Bryant, Intel’s CFO. “It was not unusual to have loud arguments in public places. Paul is a firm believer in not winning arguments by yelling or insulting.”

Otellini (who once remarked that “a waking hour is a working hour”) has wasted no time putting that political capital to work, announcing a wholesale reorganization of the company–splitting it into five groups with such names as “mobility” and “digital home,” each of which will focus on different ways consumers use technology. Yet analysts have not stopped being skittish about the company’s future. “They need to get their road map stabilized,” says Kevin Krewell, editor in chief of Microprocessor Report,a trade magazine. “People in the industry count on Intel to know where it’s going. If it swings wildly around, the market gets nervous.”

For one thing, Otellini’s “platformization” still has to prove itself in areas other than Centrino. Later this year Intel will announce an as-yet-unbranded platform for desktop computers (dubbed Desktrino by company insiders) that takes advantage of Intel’s dual-core technology. Dual-core puts two brains on the same chip, which makes it easier to run multiple applications at the same time. But Desktrino may be outmoded before it is born. IBM, Sony and Toshiba took a surprise lead recently when they announced production of their Cell processor, which has eight brains to Desktrino’s two.

Intel watchers are also nervous about how much the company is dependent on the rapidly maturing computer market. “They have to look beyond the PC,” says Apjit Walia, an analyst with RBC Capital Markets, who points out that Intel has fallen far behind its rivals in putting chips in cell phones and other wireless devices. “They have been talking the talk in communications; they haven’t walked the walk.” Indeed, Intel’s communications division is still losing money, despite a $10 billion investment since 2001. Barrett and Otellini say the division is a work in progress.

Doomsday predictions, of course, have surrounded Intel for years: that the PC market is maturing, that the competition is cutting into Intel’s lead. Yet the company has managed to keep growing. Despite AMD’s recent resurgence, Intel’s position in the PC-chip business remains unchallenged, with a market share of nearly 90%. It is also a step ahead thanks to Barrett’s farsighted investments in manufacturing. While AMD has one major semiconductor plant, Intel has four placed strategically around the world, churning out chips 24 hours a day. And unlike PC manufacturers and retailers, who have to deal with wafer-thin margins, Intel–thanks to its dominant position–enjoys a 55% profit margin on every $300 PC chip. “Most companies would kill to get margins like that,” says Nathan Brockwood, senior analyst for Insight 64, a Silicon Valley research firm.

The real trouble is that it is nearly impossible to find new businesses that offer that kind of enviable profitability. So where will Intel’s future growth come from? Some Wall Street analysts argue that Intel should be spurring growth by acquisition, perhaps buying a major communications company. But Otellini remains committed (at least publicly) to growth from within, continuing to see opportunity in consumer-focused chip development. He is particularly bullish on WiMax, a new long-distance, wireless Internet technology that could fuel the next generation of Centrino laptops–and a lot more besides. For example, Intel demonstrated WiMax at the latest Sundance Film Festival by streaming an entire movie wirelessly, bit by digital bit, from a distance of more than six miles. One director said it was the best-looking film screening he had ever attended. That, like the iPod, is music to Otellini’s ears.