Network: The CEO without an ego

Intel's Craig Barrett considers himself first and foremost an engineer. He talks to Stephen Pritchard

THE COMPUTER industry has its share of larger-than-life personalities and fragile corporate egos. It is a business in which companies with strong figureheads attract attention out of all proportion to their market shares or bottom lines. It is a business where hype is something of a common currency.

Dr Craig Barrett is a man who dislikes hype. As chief executive of Intel, Dr Barrett heads one of the most powerful companies in a powerful industry. Information technology accounts for about half the growth of the entire US economy. But the management style of this 59-year-old engineer is in sharp contrast to the flamboyance of some of his peers and, indeed, to that of his immediate predecessor at Intel, Andy Grove.

This is not something that worries Dr Barrett. His 25-year career at Intel came after 10 years' teaching engineering at Stanford University. A textbook he wrote then is still in widespread use in American universities. Dr Barrett still identifies himself, first and foremost, as an engineer, and Intel as an engineering-led company.

"I don't consider myself to be a professional businessman but an engineer who taught engineering," he explains. "I enjoy the technology, and moving the technology out into the marketplace. I enjoy competition, and I enjoy the pragmatic parts of the business. Our industry has an extraordinary amount of hype: talk about developments which are going to replace the PC tomorrow, or outlandish claims about benchmarks. It sets false expectations and confuses the consumer."

Intel, though, has a far higher profile now than in any time in its history. The "Intel Inside" campaign has been an enormous success and played a significant part in the dominance of the Intel-Windows standard in the desktop computer marketplace. Dr Barrett, then chief operations officer at Intel, played a big part in that success.

"We are still an engineering company, but we have partially transformed ourselves into a marketing company," he says. "The Intel Inside campaign was one of the largest in the universe; that was not an engineering company- type programme. But we are still grounded in the technology."

The technology, and the possibilities it brings, accounts for much of Dr Barrett's personal motivation. "I love the technology, and I get to work with a great bunch of people. It is fun. If it wasn't fun, I would be fly-fishing in Montana," he admits.

Seeking refuge on his Montana ranch would certainly have been the easier option. Craig Barrett has taken on the chief executive's job at a tough time for Intel. Revenues are flat and the computer chip market is far from healthy. "Growth is our biggest challenge, both in existing and new markets," Dr Barrett admits. "We have had 18 months of flat revenues. We need the PC market to grow." New applications, such as digital imaging or electronic shopping, are areas Dr Barrett expects will drive people to buy their first PCs, and persuade existing users that they need to upgrade.

As a company, Intel invests millions of dollars in research into new markets for computers and computer chip technology. Moores Law - the idea that chip performance doubles every 18 months - is holding true more than 30 years after Gordon Moore, one of Intel's founders, first conceived it. But faster chips come at a price.

A microchip factory is more than 10 times as expensive than a decade ago. Intel estimates the plants to build its next generation of chips will cost more than $10bn. The company will have to sell huge numbers of chips to recoup that cost.

Between the wars, US car manufacturers reacted to flat demand for their wares by introducing new models far more frequently. New designs were calculated to make older cars look dated, and spur motorists to buy a replacement, even if their old cars still ran well. The idea was successful, if cynical. Computer users could be forgiven for thinking that the PC industry is using similar tactics today.

Ever-more-complex software demands increasingly powerful computers; users are forced to upgrade their programs to stay compatible, even if their old packages still work. They then need faster computers to run the new software, but gain little from the process themselves.

This is not an argument that finds favour with Dr Barrett. Last year, Intel invested some $300m in new ventures; around half the money was spent outside the US. Some of this goes into new technologies which Intel can incorporate into its chips, motherboards or networking hardware. Some goes to find and develop new uses for computers.

Intel, for example, has invested significant sums in video conferencing, and it is one of the key partners in Blue Tooth, an industry group developing wireless networking for devices from mobile phones to laptop PCs. "We try to move the technology forward," Dr Barrett says.

"To a certain extent, the challenge is to find new uses for the technology, to ensure it is absorbed into the marketplace. We eat our products as fast as we make them: we spend a lot of time looking for new uses for our technology, to keep our internal growth growing."

Electronic commerce is one area Dr Barrett has high hopes for. Faster Internet connections and better graphics will make online shopping a far more attractive prospect for consumers. It also soaks up processing power. Nor are new applications the only motor for growth. We are working our computers harder, and using them for more tasks at once. This is for both visible applications such as word-processing or surfing the Web, and for background tasks such as scanning for viruses, compressing images, or running software that detects a fault in a PC before it happens. Then there are demanding applications such as games.

"If you are over 15 years old, you may think what am I going to do with that performance?" Dr Barrett says. "If you are under 15, you never ask that. If you play interactive games, you will always want more power."

Interactive devices are another area where Intel expects to see growth. Digital TV sets, handheld computers and smart mobile phones are volume businesses that could be far bigger than the market for conventional PCs. Intel now has a foothold in that business: it owns the StrongARM family of low-cost, low-power processors originally developed by Advanced RISC Machines in the UK.

The market for consumer interactive devices is important, but Dr Barrett does not subscribe to the view that it will replace the PC. He expects to see one billion connected PCs within a few years, and many will be in homes.

Many of those same homes might buy a set-top box, too.

"I am firmly of the belief that when you sit down to be entertained, you sit down in front of the TV set. You either want to be entertained, or catch the news. You may not want to be terribly interactive," he says. "I think the PC and the TV will co-exist. If you want to be entertained, you sit in front of the TV set. If you want to be interactive, you will do that with a PC."

Intel's record in the low-cost computing business, though, has not been good. The market for PCs costing less than $1,000 is one of the fastest growing, yet last year Intel found itself caught without a low-cost chip, and rivals, particularly AMD, closing the technology gap. Intel faced a fall in its market share.

The company reacted with the Celeron microprocessor. Celeron is now in its second, more powerful incarnation, which has been well-received by PC makers. Intel is making up for lost ground. Dr Barrett believes advanced technology and good engineering will keep his company ahead - and extend the power of the computers on our desks.

"Ultimately, the only way to stay ahead is to invest in the technology and bring it into the marketplace," says Dr Barrett. "We started out being lucky: our architecture was chosen by IBM [for its first PC]. But since then we have invested our R&D dollars. If you stop investing you'll become one of the crowd."