Grove Calculates Intel’s Course With Confidence--and Caution

Andrew S. Grove’s impish, expressive face could be a metaphor for the promise and problems at Intel Corp., the company he runs.

One minute he is beaming, his engaging grin exuding nothing but confidence, his deep-set eyes betraying just a glint of cockiness. Intel has a virtual monopoly on the complex chips that form the brains inside personal computers, and profits are pouring in.

But the next minute, Grove’s expression turns fierce, the grin vanishes into a stiff pucker, his eyes turn dark beneath a furrowed brow. Intel’s monopoly is under attack, from newfangled competitors who make faster microprocessors and from old-line chip companies that are cloning Intel’s products.

A moment later Grove is pensive, instinctively stroking his chin, the eyes grown absent-mindedly wide open. The computer chip business is fascinating but treacherous, important to the nation’s future yet under relentless attack from bigger and better-financed Japanese competitors.

No matter how good things look for Intel now--and most analysts think they look very good indeed--Andy Grove doesn’t think the battle is ever won. Coming off its best year, Intel plans to spend about $1.5 billion on research and development and plant and equipment in 1991 to head off new competitive threats, even though the economy is in a recession and the company’s position in its core market appears unassailable in the near term.

“Success has a lot of dangers,” Grove, 54, said recently as he munched a sandwich in his cubicle at Intel’s Santa Clara headquarters. “You can breed an over-inflated opinion of yourself. We have to be very careful.”

For the intense but personable Grove, being careful means continually probing every aspect of Intel’s business, challenging employees to think through problems with the same dedication and intellectual rigor that he brings to the job of chief executive and president. He combines a professorial love of intellectual exchange with an insistence on discipline and a take-no-prisoners attitude toward the competition.

That managerial recipe, together with the technical vision and more subdued leadership of Gordon E. Moore, company co-founder and chairman, has helped Intel gain a truly extraordinary position in the computer industry. In essence, any company that wants to build an International Business Machines-compatible personal computer--from IBM itself to Compaq Computer to the raft of Asian clone-makers--is obliged to buy the most important component, the microprocessor, from Intel.

Intel’s enviable situation is in part a result of good fortune: IBM’s decision in 1980 to use an Intel microprocessor, or computer-on-a-chip, in its original PC. That meant that the software for the PC was written for Intel’s chip and that Intel was the only company in a position to supply more powerful microprocessors compatible with the same software as the early models.

The result has been a family of microprocessors--first the 8088, then 80286, 80386 and 80486--that have, in their turn, established a standard for a generation of personal computers. Intel gave other companies the right to produce the earlier versions under license but chose to remain the only supplier for the now-standard 386 and the more advanced 486 (the more common names for the 80386 and 80486, respectively).

As 386 sales boomed during the past several years, Intel’s revenues and profits soared. Last year the company earned $650 million on sales of $3.9 billion at a time most other semiconductor companies were struggling to break even and leaped from eighth place to fifth in the annual rankings of the world’s largest semiconductor companies compiled by Dataquest, a San Jose-based technology market research firm.

The coming year, however, promises to be tougher than the last one.

“Intel will not have the degree of dominance in the future that they’ve had in the past,” said Michael Slater, publisher of the Microprocessor Report. “Their percentage of the desktop (computer) market will shrink, though the overall market will grow faster than Intel shrinks.”

The challenge is coming on two fronts. In the bread-and-butter 386 business, which analysts estimate accounted for about $1 billion in revenue for Intel in 1990, there will soon be competition from clones.

Leading that charge is Advanced Micro Devices, which has for years been fighting a multi-front legal battle with Intel over the rights to produce the 386. Although all the legal issues--including whether AMD will be allowed to call the product a 386--have not been resolved, AMD has nonetheless begun selling the product.

Nexgen, a Silicon Valley start-up, is developing multi-chip sets that are compatible with the 386 and the 486. And several other companies, including Chips & Technologies, are rumored to be working on 386 clones.

Grove, however, says he isn’t worried: “From the moment the 386 saw light (five years ago) we knew this would happen.”

Intel’s response is straightforward enough: new, innovative versions of the 386 designed for specific applications and a host of new 486 products--some cheaper and some fancier--to help push PC vendors toward that processor.

“We’ll compete with ourselves rather than let someone else in,” he says. And Wall Street agrees that the threat isn’t too serious.

“AMD will pick up some business that Intel can’t service anyway,” says Thomas Kurlak, semiconductor analyst at Merrill Lynch Research, noting that demand for the 386 has been outstripping production. “It’s a one-shot product, and it’s becoming long in the tooth.”

A greater concern is the long-term challenge from microprocessors that use reduced instruction set computing (RISC), a design technique that increases raw computing power. Commercially available RISC products from Sun Microsystems and Mips Computer Systems, as well as proprietary RISC chips made by IBM and Hewlett-Packard, dominate the market for powerful engineering workstations, and some observers believe that they will inevitably make major inroads in the PC world.

Charles M. Boesenberg, president of Mips, asserts that within three to five years RISC chips will be the predominant processor for personal computers in the $5,000 to $10,000 class. That will be made possible by the emergence of a new software operating system from Microsoft that will allow RISC-based PCs to use the same new software as those that use Intel chips.

Grove thinks that’s nonsense. “What’s mandatory at the desktop level is continuity, and the desktop is weighed down with X86 processors.” He says that the constantly improving 486 will be a match for the RISC chips and that Intel has RISC products of its own for special applications like graphics processing.

Intel also has the spending power. Grove points out that the company will spend more on R&D; this year than the RISC chip vendors will take in in revenue. Intel plans to spend an intimidating $800 million to $1 billion on new plant and equipment and $600 million on R&D; in 1991--more even than Japanese giants like NEC and Toshiba.

But the challenge for Intel is more than technical. Some personal computer vendors are unhappy with their heavy dependence on Intel--particularly because key chips have been in short supply--and clone vendors and RISC companies are betting that PC manufacturers will leap at the chance to gain another source of chips.

Intel has also irritated some customers by competing against them. It supplies complete personal computers to companies such as American Telephone & Telegraph that sell them under their own name.

Intel’s aggressive legal strategy on patent issues has also won it few friends, and the company has been accused in a lawsuit of intimidating customers out of buying competitors’ products.

Grove and Intel’s willingness to play hardball was illustrated clearly in an arbitration decision issued last year as part of the legal wrangle with AMD. The arbitrator was sharply critical of Intel’s conduct, stating that the firm was duplicitous in its dealings with AMD and had unilaterally abrogated a technology-sharing agreement because it was in its business interest to do so. Damages in the case have not been awarded.

All this adds up to a perception that Intel has grown arrogant. “I think there’s a lot of resentment out there,” Slater says. “It’s hard to say how much that might hurt them, but Intel may have to change its attitude.”

Grove’s approach to this problem is typical of his direct, goal-oriented and sometimes confrontational style of management. He commissioned a study of the company’s image among computer buyers to see if that was a problem. It wasn’t.

He established an objective of being one of the top two vendors in terms of customer service in the eyes of the computer companies that buy Intel’s chips, and No. 1 in quality. Currently, he says, 87% of Intel’s customers rate the company that high. The goal is the high 90% range.

Intel-watchers say it’s Grove’s ability to attack specific problems like that head-on, while still remaining attuned to broader issues, that sets him apart from other managers. “In most cases, management is made up of smart people, but they only focus on one thing at a time,” Kurlak says.

“He (Grove) can think about a lot of conflicting issues simultaneously. At Intel, they prepare for a lot of different eventualities, so they don’t get blindsided.”

Robert Burgelman, a management professor at Stanford University who has studied Intel, agrees that management’s “alertness” has been a key to the company’s success. “They have an extraordinary knowledge of the industry, and they’ve been able to recognize what’s important.”

He notes that the microprocessor itself was not planned--it came about as part of a product development for a customer. But when it happened, Intel had the foresight to buy back the rights. Burgelman points to the willingness, even eagerness, on the part of Grove and Moore to intellectually engage people at all levels of the organization as a key factor in nurturing that alertness. Good people are attracted to such an environment, he adds, and good ideas are acted upon.

And that kind of discourse, whether it’s about obscure scientific questions or management issues, is what Grove likes best. Nowhere does he look more comfortable than in front of a class at Stanford Business School, prodding students to consider all angles of that day’s case.

“I like a good intellectual pursuit of something, engaging with groups of people and coming to grips with a problem. What the issue is doesn’t matter,” he says.

The pursuit of the solutions doesn’t leave him after a typical 11-hour day at the office. “Sometimes I’ll be taking a bike ride, and I’ll still be conducting these arguments in my mind.”

These days, Grove is the “outside” man at Intel, focusing on the complex relationships with other companies in the industry while Executive Vice President Craig Barrett assumes much of the responsibility for internal operations. That recent change has helped Grove maintain his enthusiasm for a job that he once said he would leave by age 55.

He also spends a lot of time lobbying and spreading the word about the severe problems he believes the United States will face if it continues to do nothing about the Japanese challenge in high-tech. Like many immigrants from Eastern Europe--he came to the United States from Hungary after the Soviet invasion in 1956--Grove has an instinctive interest in politics, but he doesn’t much enjoy his lobbying role.

“I don’t share the view that people in government are turkeys. But the system is not wonderful, and people give in to the system instead of fighting it,” he says. Getting the government to do something about the high-tech industry’s accelerating problems, he laments, is just about impossible.

At Intel, by contrast, “almost every problem is solvable.”

So far, at least, he’s been right about that.