Intel Is Betting All of Its Chips on the Future

Breathtaking technological change in semiconductors, which has already transformed the world economy, will continue at its torrid pace for at least another 10 years and probably longer. “We’ll be putting hundreds of millions, maybe even a billion transistors on each silicon chip,” Gordon Moore, co-founder and chairman emeritus of Intel, said last week.

By the time computer microprocessors--which now contain 6.5 million transistors--get to the hundred million level in 2004 or so, speech recognition will have advanced to the point at which you will talk to machines that do your bidding. The droid R2D2 from “Star Wars” will have become reality.

That’s a serious forecast, not science fiction, and Intel is acting on that forecast in hopes of dominating the future market as it does semiconductors today.

Moore has defined technology’s pace since 1965, when he wrote that the complexity and capability of electronic computing would double every year, later amended to two years. He also predicted that prices for computing power would fall constantly and thus spread computer usage.

His prediction, known as Moore’s Law, proved visionary. Microchips became pervasive, controlling car engines as well as automobile assembly lines, and adjusting the fit of bluejeans even while making heart surgery routine.

Moore understood that the development of computing would not be in large machines but in the densely packed circuits that form the processing units, or brains, of the computer. As chip manufacturing technology--known as photolithography--allowed ever more circuits to be etched on wafers of silicon, chips were able to mimic in a small way the neural networks of the human brain.

With all of those heady thoughts and numerous inventions, Moore and his colleagues built Intel from a start-up in 1968 to a world leader in technology--and lately a bellwether of the stock market.

Last week, as Intel reported profits of more than $5 billion on revenues of $20 billion for 1996, it was hailed as potentially the world’s most profitable company. The Santa Clara, Calif.-based firm now has 48,500 employees and a total market value of $120 billion, larger than almost any other global company. Moore’s ownership of roughly 5% of the stock is worth $6 billion at current prices.

Yet Intel has reached a critical point in terms of competition, technology and management. Moore, now 68, was named chairman emeritus last week, and even though he says it will make no difference to his three-day-a-week schedule of recent years, management change looms. Andrew Grove, 60, remains chief executive; Craig Barrett, 57, a production specialist, was elevated to president last week.

The challenge, as well as the opportunity, is that technology and markets are changing.

Intel’s 25% growth in revenues surprised many analysts, who weren’t paying attention to the areas in which Intel’s sales are growing. “The acceptance of personal computers in the developing countries is astounding,” says Moore. “In China, parents plunk down a year’s earnings to buy a computer for their children’s education.” The same passion for computers and education is true for South Korea and many other countries of Asia--a boon for Intel, which reaps 20% or more of the value of every computer sold. What that tells you is that Asia’s growth is founded on a bedrock of hope for future generations.

The business is fast and furious. Intel had to alter its distribution to make sure to get its chips into the hands of small manufacturers in Asia that turn out 10 computers at a time and then order chips for 10 more. That tells you that even Intel has to watch out for competitors who are eager to supply such new growth markets.

To be on top is to invite competition, of course. Advanced Micro Devices, a longtime rival of Intel, has a new microprocessor--developed by a former Intel executive--that it hopes can win a share of the 85 million personal computers to be sold worldwide in 1997.

More serious for Intel is the challenge of changing technology. The shift of emphasis to computer communications, for Internet access and corporate Intranets, could spur demand for simple network computers rather than the full-power PCs driven by Intel’s microprocessors.

Moore counters that the Internet represents an opportunity, not a problem. Intel’s microprocessors already facilitate communication of data, he explains, and the company is backing multimedia ventures. The stakes are rising. Soon Intel will move to larger silicon wafers, so that more chips can be produced at one time with significant cost savings. But such wafers will demand fabrication plants costing no less than $2.5 billion apiece, a 70% increase over today’s costs.

What can Intel do to stay on top in such a challenging field? Adapt to change and avoid arrogance, the flaw that cut down other giants from General Motors to IBM to CBS.

Intel by and large has not been arrogant and has adapted. The company started out making memory chips but needed an investment from IBM to stay afloat in the early 1980s, when Japanese competition made memory chips unprofitable.

Soon, Intel dropped memory chips to concentrate on microprocessors, which combine several functions. IBM’s choice of Intel’s microprocessor for its PC was the greatest boost the company ever got.

Intel has continued to invest heavily and work diligently to maintain its leadership, with fabrication plants in the Western U.S., Israel and Ireland, and simpler assembly plants in Malaysia, the Philippines and, soon, Costa Rica.

The future, Moore believes--whatever the fashion in Internets and stock markets--will see microchips increase in complexity as they move toward closer imitation of the brain.

He knows whereof he speaks. Even now major research laboratories, including one that Moore endowed at his alma mater, Caltech, are working on microcircuits that recreate the musical thought patterns of Johann Sebastian Bach and the simple learning processes of a child.

Intel and Moore, along with the whole computing industry, have been lucky: It’s not every business that can count for growth on the human thirst for knowledge.