CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

Barrett, Intel face stormy seas

The executive shuffle comes as the chip giant seems to have been caught unawares by sea changes in the PC industry.

Intel's leadership transition is taking place just as the company seems to have been caught unawares by sea changes in the computer industry.

The chipmaking giant today promoted president and chief operating officer Craig Barrett to the position of chief executive to oversee operations at a time when the company must learn to cope with lower margins and the irreversible trend toward low-cost computers--or else risk losing its seemingly unassailable position as a computer industry superpower.

Barrett, 58, succeeds Andy Grove, 61, who will remain as chairman.

In many respects, Barrett is the person who See special coverage: 
Intel's changing of the guard made Intel the world's dominant microprocessor maker, through a relentless drive that focused on the production process to ensure a constant supply of quality chips for the majority of the world's PC makers.

"Craig has been the architect of Intel's operations throughout the last decade...meticulously ramping [products] into high-volume production," Grove said in a prepared statement.

But this knack for deftly handling the nuts-and-bolts side of Intel's chip business begs the question: Is Barrett the man to forge a path in new markets?

One Intel insider familiar with his management style is circumspect. "If it's a matter of figuring out where the world is going, there's a legitimate question whether or not he can [do that]," said the source.

On the other hand, if it's a matter of taking existing markets and working the business model so that the company can be profitable, then Barrett "can make it happen. He's an incredibly strong manager. He loves figuring out how to do things," the source said.

The upshot is that he will have to be equally good at identifying new markets and executing strategy since he faces an industry that is far different from the one Grove has navigated over the last decade. And what's most unsettling to Intel is the fact it is not the driving force behind these changes. At least not yet.

The first evidence of change came last year with the phenomenal growth in sub-$1,000 computer sales. With Intel slow to address this market with a low-cost chip, PC giant Compaq Computer turned to Cyrix. The result was the first wave of sub-$1,000 PCs that had many of the bells and whistles of their over-$2,000, Intel-based cousins.

By year's end, these cheap PCs constituted almost 40 percent of the retail market, according to analyst estimates. When Advanced Micro Devices (AMD) chimed in as well with a low-cost chip offering, a reluctant Intel finally responded. (Intel is an investor in CNET: The Computer Network.)

At first, Intel simply discounted prices on older Pentium chips. Only in the last few months has the company developed with a chip designed specifically for this market. The new Celeron processor is expected to be officially introduced next month, more than a year after the first wave of sub-$1,000 PCs hit the market.

But even this appears to be a nothing more than a quick fix. The Celeron is little more than a Pentium II without the extra "cache" memory. This allows Intel to quickly offer a cheap Pentium II processor, but because of the lack of high-speed memory, it doesn't run any faster than older high-speed Pentium chips, according to a number of analysts.

The chip company has thus ignored what Grove, author of Only the Paranoid Survive, calls an "inflection point" in the computer market, according to a recent report from Forrester Research. Inflection points represent major industry trends that companies need to identify and jump on to thrive.

The sub-$1,000 PC sector "makes up almost half of the PC business," analyst Carl Howe said. "Even corporate buyers...are putting pressure on their suppliers for cheap PCs Despite a new squadron of engineers dedicated to Basic PCs, Intel's gold-plated business model based on exclusive chips sold at high margins will stall." ("Basic PC" is Intel's nomenclature for a sub-$1,000 system.)

Forrester's Howe also said that Intel's misstep is reminiscent of IBM in the late 1980s. "Intel's late arrival with chips targeted at sub-$1,000 PCs and its attempt to migrate computer builders to its proprietary architectures echo IBM?s failed attempt to move PC users to its patented Microchannel," he said.

An early warning of that stall came March 4 when the company announced that earnings would fall 10 percent short of previous estimates because of declining demand.

Intel may have to start worrying about missing another inflection point in the computer market in the form of "convergence" devices, so named for their blend of PC and consumer electronics technologies.

General Instrument, the largest manufacturer of TV set-top boxes, said yesterday that it would use a MIPS processor from Quantum Effect Design for its next-generation digital set-top box, a device that will include many features similar to a personal computer. (See related story)

In spite of Intel's newfound interest in the market, it lost out in the decision, which is important because General Instrument last year won an agreement to supply 12 large cable companies (including TCI) with up to 15 million new digital set-top boxes.

The potential market for set-tops with computer-like features could one day exceed the market for costlier PCs and represents a critical area of chip growth. The TV set-top box isn't yet a large volume product, nor is another the market for handheld computers.

Intel's has been loath to enter the market for consumer devices because it puts the company's historically high profit margins in peril. But to sustain its rapid growth, the company can no longer afford to ignore these burgeoning new markets.