Why Andy Grove can’t stop

Even though it’s a glorious Saturday morning and he’s coasting downhill astride his jet-black bicycle, Intel Corp. CEO Andy Grove is hard at work. As usual, he’s lagging far behind his more athletic wife on one of their weekend treks through the tawny hills above Silicon Valley, and now, for the first time in days, he can really think. His mind, a Brownian motion of ideas, memories, and, of course, worries, floats freely.

Periodically he dwells on what he calls a “concern du jour”—mulling the promotion of a subordinate, or perhaps playing out an imaginary conversation (more likely an argument) with a prominent customer or business partner. Says he: “On my bike rides I can be very eloquent, and since there’s nobody to talk back to me, I always come back victorious.”

More often, however, he’s contemplating the future—not so much what might happen to Intel’s sales and stock price, or whether to build yet another billion-dollar chip-fabrication plant, but wider-ranging issues: Why do people watch TV so much? Does it really make sense for phone companies and cable TV companies to invade each other’s turf? If personal computers are so damned frustrating, why are people enthralled by them? How come he and Microsoft’s Bill Gates and other industry luminaries were blindsided by the sudden popularity of the Internet? Will Hollywood ever come up with something better to offer the digital world than movies, soaps, and sitcoms? Is electronic commerce right around the corner?

You’d think the man in the scarlet crash helmet could use the weekends to just chill. Since Grove became CEO in 1987, Intel’s revenues have grown nearly sixfold, to $11.5 billion last year, and the chipmaker has leaped from tenth place in its industry to a Herculean No. 1. As builder of the microprocessors that power 80% of all personal computers, Intel is on track to ship more than 35 million Pentium chips this year, as well as a half million of its next-generation P6s. Wall Street analysts predict that in 1995 Grove’s company will continue to grow like a startup, logging about $16 billion in sales and reaping $3.6 billion in profits. So promising is Intel’s future that Dan Klesken, an analyst at Robertson Stephens & Co., predicts that the chipmaker will hit annual sales of $50 billion by decade’s end. No wonder the stock price has nearly doubled since January, to $111.

Still, if you think Andy Grove sleeps easy at night, think again. The excitable CEO is just too wired. Not content to exercise a virtual monopoly on the key component of the Digital Age, Grove wants to set the hardware agenda for the entire PC industry and, while he’s at it, perhaps the consumer electronics and telecommunications industries as well.

“The PC is it,” he says. “We can make it so superb as an entertainment machine, and so vital as a communications medium for both the home and the workplace, that it will battle with TV for people’s disposable time.” It’s a power play Bill Gates might admire—if he were only a bystander. And it could lead Intel to an explosion of growth that could make it one of the most profitable and influential businesses on the planet.

What makes Grove believe in this grandiose scenario is both the steep upward trajectory of semiconductor technology and the fact that the new power users of computerdom aren’t techies at all. They’re U.S. Consumers, who, according to the Electronic Industries Association and Dataquest, last year spent more on PCs than on TVs. These eager newcomers are shelling out big bucks for multimedia bells and whistles to make their machines more entertaining—and multimedia means high-powered chips. Unlike most businesses, which have already invested vast sums in software for older, feebler machines, consumers are free to buy the hottest new boxes. That spells irresistible market opportunity for Grove, who four years ago helped launch the splashy “Intel inside” advertising campaign that made Intel a household word.

Grove now aims for Intel to define a global standard for consumer computers. He envisions machines that will incorporate, as standard equipment and at much lower cost, all the features of today’s best multimedia PCs: crystalline stereo sound, crisp digital video, gymnastic 3-D graphics, rich fax, voice, and data communications. How? The key is to use supercharged Pentium or P6 processors to handle chores that now require additional hardware. As many consumers have learned the hard way, add-ins can cost hundreds of dollars and introduce show-stopping software incompatibilities.

If the plan succeeds, Intel would lord over much more than the multibillion-dollar PC hardware business; it would also have the upper hand in devising derivatives of the PC that might subsume today’s game players, set-top boxes, and even TVs and VCRs.

Already Intel has technologies and products jumping off the drawing board that will accomplish some of these goals: ProShare desktop videoconferencing gear for your PC; a cable modem that will let PC users tap the Internet or online services via their cable TV lines and download data 1,000 times as fast as via the telephone; another type of modem that lets people simultaneously talk and swap data over a single phone line; PCs disguised as set-top boxes for playing interactive games and controlling cable TV.

Cool stuff, no doubt. Still, if Grove weren’t innately paranoid, he’d have reason to be wary. While playing to consumers, his grand strategy risks alienating some of Intel’s best customers and strategic partners. Its standard designs would turn the PC into even more of a commodity product than it already is, leaving computer makers with fewer ways to eke out a profit.

Though Grove denies it, his strategy also challenges Microsoft’s primacy. Grove charges that the Seattle software power doesn’t “share the same sense of urgency” to come up with an improved consumer PC. He believes that despite Intel’s lucrative 15-year relationship with Microsoft and their fundamental “commonality of purpose,” there’s room in the business—indeed, a crying need—for a forceful hardware advocate. “The typical PC doesn’t push the limits of our microprocessors,” he complains. “It’s simply not as good as it should be, and that’s not good for our customers.” Grove maintains that Intel is simply trying to light a fire under Gates to make sure future versions of Windows keep pace with Intel’s designs.

A notoriously hands-on manager who for more than 20 years ruled Intel’s product-development and manufacturing operations, Grove has dropped just about everything to pursue his dream. These days chief operating officer Craig Barrett, 55, pretty much runs Intel (see box).

Meanwhile, Grove picks the brains of people like DreamWorks SKG’s Steven Spielberg and Tele-Communications Inc.’s John Malone, trying to divine how to make PCs more entertaining and better at communicating. He consorts with the young propellerheads who run Intel Architecture Labs, an Oregon skunkworks that he hopes will become the de facto R&D lab for the entire PC industry. Occasionally, he hits the lecture circuit as a demo-toting technology visionary.

One reason Grove is so aggressive is that he believes he has the laws of nature—or at least of the semiconductor industry—on his side. Anyone familiar with the computer business has probably heard of Moore’s Law. Coined way back in 1965 by Gordon Moore, an Intel co-founder and still its chairman, this axiom posits that the performance of chip technology, as measured against its price, doubles every 18 months or so. Moore’s Law is the main reason computer hardware often seems outdated within months of hitting your desk.

Equally important is a corollary of Moore’s Law that explains why the business of building computers is so fierce. Call it the Cannibal Principle. Explains Moore: “The whole point of integrated circuits is to absorb the functions of what previously were discrete electronic components, to incorporate them in a single new chip, and then to give them back for free, or at least for a lot less money than what they cost as individual parts. Thus, semiconductor technology eats everything, and people who oppose it get trampled. I can’t think of another technology or industry quite like it.”

To Moore, his law and the Cannibal Principle guarantee a wide-open future for the industry he helped create. The 68-year-old chairman, whose 5.6% stake in Intel is worth about $2.6 billion, finds that prospect reassuring. But Grove, whose Intel stake is worth $102 million, remains a worrier. Hence, a powerful third dynamic at Intel—the rule of thumb company wags call Grove’s Law. It explains a lot about Intel’s notoriously aggressive culture and why the feisty, 58-year-old Hungarian amigra is forever restless. Unlike Moore’s elegant theorem, Grove’s Law, which the CEO has espoused for decades, is a terse warning: “Only the paranoid survive.”

Life has led Andy Grove to one crossroads after another, and at each instance he has invariably veered onto a riskier, less familiar path. Grove won’t divulge much about what must have been the hellish years of his youth in Hungary, other than that he was named Andras Grof at his birth in 1936. He lived through the anti-Semitic atrocities of the Nazis and the most oppressive years of Josef Stalin’s Soviet domination of Eastern Europe.

He escaped to the West in 1957, arriving later that year at the Brooklyn Navy Yard aboard a musty, recommissioned World War II troop ship filled with refugees. “We didn’t see the Statue of Liberty or anything,” Grove recalls. “They immediately took us by bus to Camp Kilmer, New Jersey, which had been a POW camp during the war. We thought that all the communist propaganda was true, that America was just another drab, totalitarian state.”

Grove moved in with an uncle in the Bronx and enrolled in City College of New York, studying for a degree in chemical engineering. During the summers he worked as a busboy at a resort in the Catskills. A star student whose professors urged him to pursue a Ph.D., Grove applied to the University of California at Berkeley for graduate school, less because of its renown than because he hated the harsh winters in the Northeast.

After earning a doctorate in 1963, Grove probably could have gone on to make a name for himself in academia. But again at a professor’s urging, he interviewed for jobs at Bell Laboratories and an intriguing California startup called Fairchild Semiconductor. Recalls Grove: “The choice was very easy: Bell Labs was the place to work back then. So I picked Fairchild.”

Among Grove’s new bosses were Moore and Robert Noyce, one of two men who independently invented the integrated circuit. Four years later, when Moore and Noyce announced plans to start what would become Intel, Grove was the first to volunteer to join them. “I was supposed to be director of engineering, but there were so few of us that they made me director of operations,” he recalls. “My first assignment was to get a post office box so we could get literature describing the equipment we couldn’t afford to buy.”

The founders quickly learned that Grove was a natural whip cracker. Recalls Moore: “Andy always made it hard for me. I would be all excited that we were under budget or ahead of schedule on a product—and he’d ask why we couldn’t do it faster and cheaper. He got very interested in the art of management, and that served us very well.” In 1987, Grove was named president and CEO.

For years Andy Grove had no use for personal computers, even though Intel owed its livelihood to them. Visitors to his office in the late Eighties would be treated to a salty harangue about how counterintuitive PCs were, how they required users to become experts just to operate them, and how fooling with them was a monumental waste of his time.

That all changed in 1989 when Intel began using electronic mail companywide. Grove had no choice but to deal with his PC every day if he wanted to keep up with what was going on in the company. After Microsoft introduced Windows 3.0 in 1990, using a PC became less of an ordeal. It dawned on Grove that the PC was most useful as a communications medium.

Grove’s attitude toward PCs also was mightily influenced by events in the marketplace, particularly by the phenomenal success of the “Intel inside” campaign. By establishing the Intel name in the minds of consumers, the company hoped to dampen sales of “clones” of Intel microprocessors made by Advanced Micro Devices and others. As part of the campaign, Intel asked PC makers to label their machines with a distinctive “Intel inside” logo that echoed the advertising slogan. Not only was Intel able to hold the high ground against its rivals, but customer feedback revealed as well that regular consumers, not just the techies, really did care what made their PCs tick.

Grove, who was becoming a typical PC user himself, also had a disquieting insight. He realized it didn’t matter how zippy Intel’s microprocessors were because PCs didn’t make full use of their capabilities. To his frustration, this seemed beyond Intel’s power to fix: The chips and other components of the PCs relied completely on Microsoft’s operating-system software to tell them what to do. And even though multimedia was beginning to attract consumers’ interest, Microsoft, loath to maroon business customers who owned older, dumber machines, had been slow to add the code necessary to make the newer PCs better at handling sound, video, and graphics.

Savvy computer users could add these features themselves, of course, by installing multimedia circuits, software, and peripheral devices. But the market for those extras quickly became fragmented. That, in turn, has caused software developers to get bogged down designing their games and other flashy programs to anticipate all the possible competing hardware variations, many of which are incompatible. Fumes Grove: “With no common platform to target, applications developers haven’t been able to be as creative as they could be. That deprives PC users of what their machines are capable of.”

It wasn’t until 1993, when multimedia was becoming the hottest buzzword in personal computing, that Grove decided Intel had to throw its weight around. He turned to the Intel Architecture Labs, which he had established two years before. Its original mission had been to look for basic improvements in design that Intel could quietly share with PC manufacturers, whose R&D budgets had been increasingly hammered by price wars.

Thus was born a technology strategy called native-signal processing, or NSP. Behind the nerdy name was an explosive goal: to develop a standard design for PCs that would provide multimedia capabilities using the Pentium itself, rather than additional hardware. Explains Michael Slater, publisher of Microprocessor Report: “Intel basically decided to do an end run around Microsoft and the add-in hardware makers, and go straight to the PC manufacturers and software developers with a whole new platform standard.”

With its assumption that the Pentium and its successors will improve in performance and subsume many functions that now require extra chips and add-in circuitboards, Grove’s strategy depends heavily on both Moore’s Law and the Cannibal Principle. That makes NSP an obvious threat to dozens of small add-in companies. But NSP also involves software, the domain of Grove’s PC-duopoly partner, Bill Gates. And therein lie the seeds of a potentially ugly rift in their lucrative relationship.

Not many people know it, but 2,000 of Intel’s 32,600 employees are programmers. Most of them write code that is embedded in the company’s various chip products. Others, however, are working with a small Santa Barbara company called Spectron to develop software that is central to the NSP design, code that invisibly enhances Microsoft’s Windows by controlling the way the Pentium allocates its time. In essence, the software, called IA/Spox, gives top priority to tasks that affect what the user sees and hears. Result: a livelier multimedia experience. Meanwhile, other software tasks, such as routine data handling and other Windows chores, are forced to wait. NSP-equipped PCs will begin hitting the market this year.

Microsoft isn’t too crazy about the prospect of Intel’s providing, of all things, software to enhance Windows. Indeed, Microsoft is already working on code for the next version of Windows that improves sound and graphics somewhat. Compaq, meanwhile, has proposed a standard design for consumer computers that accommodates digital video with the help of about $100 worth of specialized chips.

Steven McGeady, a vice president at the Oregon skunkworks, maintains that Intel’s goals with NSP are entirely benign. Says he: “What we are doing is putting rabbits out there to run ahead of hounds like Microsoft to make them run faster. Sometimes the hounds might catch the rabbit—Microsoft could well provide its own alternative to our NSP software that ultimately wins. But to do that they have to run faster, and that’s what we really care about.”

To buttress NSP, Intel is expanding a business that drives Compaq and other top PC makers crazy: its motherboard manufacturing. A motherboard holds the main circuitry of a PC, and Intel has been supplying them for years, especially to smaller manufacturers that prefer not to have to design and assemble PCs from scratch. Even some big-name PC makers have become customers recently, because buying motherboards from Intel is a good way to make sure their product lines feature machines with Intel’s latest processors. Analysts estimate that of the first five million Pentiums Intel produced, fully four million were shipped out affixed to motherboards.

But big fish like Compaq don’t like it when Intel helps smaller fry get the hottest technology to market fast. Who can blame them? Intel’s motherboard operation is a key reason the PC hardware business remains a free-for-all in which makers are forced to compete on price, sales channels, and other nontechnical advantages. Meanwhile, Intel reaps extra profits—and without a kingpin PC maker, has an easier time shaping the industry.

To hear Grove tell it, the NSP and board strategies are just a means to an end: to inspire the development of software applications that make PCs much more engaging and useful, and ultimately to broaden the market for PCs and Intel microprocessors. As Grove puts it, “We want the whole pie to grow.”

Grove thinks hot new applications will come from two quarters—communications and digital entertainment—and that some of the best applications will combine the two. Intel is hard at work trying to create trendsetting offerings of its own.

“My personal obsession is the popularization of computer communications,” says Grove. For nearly a decade, he has championed the idea of digital video for the PC. Intel’s first attempt at the technology involved special chips; its latest, called Indeo, is pure software that takes advantage of the Pentium’s speed. (It’s a perfect example of how Intel’s own products are subject to the Cannibal Principle.)

Indeo is at the heart of two commercial products that Grove concedes are somewhat ahead of their time: ProShare, a desktop videoconferencing system for the PC, and CNN at Work, which uses office networks to feed PCs a version of CNN. (Prices: $999 for ProShare; $4,995 for CNN at Work, plus $10 a month for the video feed.)

Grove is also a big promoter of cable modems in the home. Hooked to a cable system and supplemented with a standard telephone line that the consumer uses to send requests and messages back to the cable provider, a cable modem makes possible what techies call rich interactivity. Translation: the ability to instantly receive photographs, illustrated documents, CD-quality sound, and ultimately digital video clips. Intel currently is conducting field tests with Viacom and several smaller partners in 180 homes in Castro Valley, California.

Says Grove: “The cable modem is our answer to the bandwidth problem. Right now, using the Internet is like reading a magazine where you can turn the page only once every 30 seconds. With higher bandwidth it can be more like every three seconds. We know it could be big, because when we tried to reposition the trial from one set of customers to another, people started freaking. They didn’t want to have it taken away.” Commercial versions of the technology probably won’t appear for a couple of years.

Grove also is intrigued by the prospect of more sophisticated computer entertainment, but he’s less sure what form it will take. “It’s like a soapy elephant—slippery and big. The problem is that neither we nor the people in Hollywood seem to know what to do with digital media. Whoever figures it out will have the key to pulling people away from their television sets.”

Having spent days each month hobnobbing with Hollywood types, though, Grove concludes that Hollywood and Silicon Valley are worlds apart: “Oh, boy, we don’t even have a common language. We don’t have common shared experiences. It’s not a matter of trust or interest. We can talk for hours, but at the end of it we don’t understand each other or the possibilities any better than before. It really bothers me.”

The thread running through all these strategies is that Grove and Intel are focusing on the needs and desires of PC users rather than on Intel’s more direct customers, the PC makers. Not even the scorching experience of last winter’s Pentium fiasco has deterred Grove from courting consumers directly. In that episode, Grove at first denied that a generation of flawed Pentiums could go haywire on typical users. But after a hailstorm of criticism, he backed down and offered to replace the arithmetically challenged chips free. Intel reserved $475 million to cover potential costs, but few customers have asked for new chips. Today the Pentium brand seems stronger than ever, especially among buyers of home PCs.

As hard-boiled as he seems, Grove has an idealistic streak. Just ask him what businessman influences him the most. While Bill Gates names management genius Alfred P. Sloan, the longtime General Motors CEO, as his inspiration, Grove points to—get ready—Steve Jobs: “Look at his history. He was the first to see what the PC was about. The first to recognize the value of the laser printer, the graphical user interface, and object-oriented software, which will become very important sooner or later. And now with Pixar, his computer-animated movie and game company, he’s pioneering real digital entertainment. Not bad for 40 years old. True, he made a few mistakes along the way, but he continues to believe in his own technological vision.”

Grove seems to enjoy being in the limelight after laboring so long in the wings while Noyce and Moore held center stage. He enjoys his role as an industry advocate and spokesman too. Lately he’s visited Washington to put in his two cents about how to deregulate the telecommunications industry. (The more diverse an industry, the better, he opines.) With all this on his plate, plus his duties at Intel, he still finds time to work on his third book, about managing change.

But Gordon Moore, the business associate who knows him best, says that while Grove seems to have mellowed a bit, he doubts that the CEO has it in him to relax. Says Moore: “In the end, thank God, it doesn’t really matter how well things are going at Intel. Grove’s Law will always keep Andy awake nights.”

Now that Andy Grove has shifted into full visionary mode, do the trains at Intel still run on time? If anything, the company is more efficient than ever, thanks largely to managers Grove has hired and groomed. Says Jack Carsten, a venture capitalist and former Intel vice president: “Andy always was the professor. He loved to bring in young engineers and teach them to be good general managers and strategists.” His main man, Craig Barrett, 55, was a Stanford University professor himself before joining Intel in 1974. He became head of manufacturing in 1990 and chief operating officer in 1993. What exactly does Barrett’s job entail? He explains: “Andy does what he wants, and I get to do everything else.”

Whatever that is, Barrett’s good at it. Since 1990, he has supervised Intel’s $7.2 billion capital-spending spree to bulk up its factories while simultaneously managing to crank up return on assets from 12.1 to 16.6 and more than double revenue per employee to over $370,000. Barrett even makes Intel’s “bad” years look good. In 1994, for example, the company ranked No. 4 among the Coins2Day 500 in return on revenues, even after setting aside a $475 million reserve for fixing the notorious flaw in the Pentium chip.

Intel’s microprocessor business is its gold mine, of course, but the company exploits other lodes as well. It has pioneered a technology called “flash memory”: chips that remember data even when the power is shut off. These have proved especially useful in cellular telephones, laptop computers, and other portable devices. Intel also makes specialized microprocessors that translate digital video signals into analog images, and others that control the innards of laser printers and fax machines. A small division makes supercomputers.

The most demanding part of Barrett’s job is supervising the company’s massive capital spending. (Grove says he still sits in on capital-authorization meetings—but only as an observer.) The company has elaborate chip plants, or “fabs” in industry parlance, in California, Oregon, Arizona, New Mexico, Ireland, Israel, and Malaysia that are almost perpetually being upgraded or expanded. Intel also has seven new factories going up, and recently announced plans for an eighth that will be its first in China. Barrett tries to visit each facility at least once a year.

One type of meeting Grove can’t stay away from is marketing-strategy sessions. Says he: “I’m not personally involved in the nuts and bolts of planning and designing a new microprocessor, but I get very involved in how we roll it out, because it’s very important in the early stages to emphasize it in the right way. The P6 does a multitude of things better than the Pentium. We have to decide whether we should immediately position it as a multimedia engine or emphasize other, more business-oriented things. There can be some really heated debates.”

Paul Otellini, senior vice president for sales, can testify to the CEO’s ongoing obsession with marketing details. “Andy may tell you he stays awake nights thinking big thoughts. But he is also the one who keeps me up nights.”