Inside Andy Grove.

The development of the microprocessor virtually defines our current business age. Andy Grove has defined what the future potential of that power is.

Nearly everyone is familiar with Moore's Law, named after Intel co-founder Gordon Moore, which holds that the power of microchip technology relative to its price doubles roughly every 18 months. It explains, among other things, why last year's Cadillac STS has more onboard computing power than NASA's Apollo 13. It also explains why that spiffy new PC on your desktop that you reckoned to be your Ferrari Pinin Farina for vrooming along the information superhighway will in two years seem like a Ford Pinto. Moore's law has held up pretty well in that microprocessing power is about 500 times faster today than when IBM first launched the PC in 1981. Intel's own performance is also a marvel. A $1,000 investment then would be worth about $33,000 now. Since Andy Grove, 60, became CEO in 1987, its revenues have increased more than tenfold to $20.8 billion. Its average annual return to shareholders during the same period is 44 percent. For the last five years its average annual return to investors has been an astounding 61 percent. Stern Stewart ranks Intel 12th among the top 1,000 U.S. public companies for total market value added - the difference between a company's total market value, debt, and equity, and its economic book value, or the amount that investors have contributed to produce that value. (Coca-Cola, GE, and Merck lead the list.)

One might think that going from 10th place to the No. 1 microprocessor maker, whose chips power over 90 percent of all PCs sold in the world, would permit the wiry, Hungarian-born emigre to pop the champagne corks. But Grove's edginess doesn't permit much more than 10 seconds of gloat. Unlike Moore's elegant law, Grove's theorem manifest in the title of his recent book, Only the Paranoid Survive, crackles like an exposed wire. To him the wolf is always at the door. Born Andras Grof, he left his native Hungary following the 1956 uprising and found passage on a recommissioned troop ship bound for New York. He moved in with an uncle living in the Bronx, entered City College, and earned a degree in chemical engineering. He also earned a graduate degree at the University of California at Berkeley in 1963 and spurned a career in academia in favor of a Silicon Valley startup called Fairchild Semiconductor. It was there that he came across Moore and Robert Noyce, one of the two people who invented the integrated circuit. When Moore and Noyce left to start Intel, Grove tagged along. Having little taste for administration themselves, the two recognized Grove's managerial talents and made him director of operations.

The company's recent track record is due in no small measure to Grove's relentlessness, but its performance wasn't assured. Intel experienced a series of upheavals and reverses that have capsized other semiconductor firms. In 1985, a year in which the company earned a modest penny a share, Grove took Intel out of memory chips - its original business - and concentrated on microprocessors. The next two years were marked by layoffs, plant closings, and unpaid leave. By undercutting Intel and other American producers' prices by 10 percent, Japanese chipmakers were eating U.S. memory chip makers' lunch. This was one of Grove's "strategic inflection points," a critical challenge that to Grove's way of thinking represents a 10x change. The decision to commit to microprocessors wasn't the obvious choice it appears today. But it reinforced Grove's managerial "paranoia" about questioning one's assumptions, as he was forced to do when Intel confronted unfavorable publicity in the face of Grove's initial dismissal of public concerns about technical flaws in the Pentium chip.

By branding the guts of the PC with its "Intel Inside" insignia, Intel did for PCs what DuPont did with Teflon. The core technology becomes the brand of choice; the assembly or manufacture of the product becomes almost incidental. But Grove understood that the company had to be more than a supplier of microprocessors if it was to grow. In the future, Intel would seek to build demand by making computers and other devices more friendly and useful. Its dominant share would ensure continued growth. This is what he means in the following discussion about making the PC a central appliance in everyone's life. It is also why the company is investing in technologies such as Internet telephony and digital photography. Networking will become increasingly important for growth. Thus, the company is working with MCI to provide greater bandwidth for networks. To drive its technology forward, Intel is spending $4 billion on new plants and $2.5 billion - more than what Microsoft earned in net income last year - on R&D. Few companies come close to this market power, a fact that has not gone unnoticed by competitors whose paranoia rivals Grove's at times. (If Microsoft is dubbed "the evil empire" in Internet chat lines, Intel is tagged as a "dark force" by those who fear its ubiquitousness.)

If further proof is needed to underscore Intel's market dominance, it came on the May 30 announcement of a 5 percent to 10 percent fall-off in second quarter earnings. It triggered fears of a PC slump on Wall Street, panicking investors into a 50 point drop in the technology-laden Nasdaq composite index. Intel says the sales slowdown is a short-term blip due to slow demand in Europe, which will be offset by rising demand for its newly launched Pentium II and Pentium MMX products. The announcement came at a routine analysts' meeting and was not dramatic in itself, but the reaction indicates just how deep a blue chip Intel has become. (It now accounts for 2 percent of the S&P 500.) Every move it makes is seen as a signal for the wider economy, which is now so dependent on the technology it provides.

"Intel won't hold the market near monopoly for much longer," says DLJ vp of Equities Research Krishna Shankar. "It will continue to grow earnings at 17 percent or better for the next five years, but the price competitiveness of AMD and Cyrix will take some share away. When PC prices start to fall below $1,000, makers can't afford to have a $400 chip in it." Others on Wall Street believe that beating back the competition is necessary but insufficient. "Intel is seeking to enable the new standards that will allow for the transition of the PC into a communications/entertainment device," says Lehman Brothers' Mike Gumport. "The 185 million PCs sold in 1994 to 1996 will all have to be replaced by 1999, and that alone should provide a healthy base for demand."

As with GM of the 1950s and IBM of the 1960s, Wintel, the complementary forces of Microsoft and Intel, have inherited the mantle of business icon of their age. It is difficult to point to any business today that isn't directly affected by operating systems and microprocessor technology.

- J.P. Donlon

INTEL NOW AND THEN

How much of Intel's future is tied to Microsoft's future?

We look at Microsoft as our most important complementer. We don't buy a whole lot from each other and we don't have any contractual relationships with each other, but their software needs our microprocessors. Our microprocessors need their software. They are the largest software provider. We are the largest microprocessor provider. And the built-in complementarity of our relative market positions kind of "destines" us to work with each other. Over time, each of us could live without the other, but it would not be easy.

Long-term, they are probably just as tied to the growth of the PC business as we are. So, that's one thing that keeps us together. But suppose we were different. What realistic chance does either of us have to encroach on the other's core business? Around the periphery we can screw around and do things to each other, but the notion that we or anybody else is going to go in and take on the Microsoft operating system or that Microsoft is going to go in and start designing microprocessors - I mean, there's nothing theoretically impossible about it. But is that a proper use of your shareholders' resources? I don't think so. To us, Microsoft is part of the environment, and I hope, to them, we are part of their environment. The environment isn't always what you want it to be. When it's pouring rain out there, I need an umbrella, not an attitude that takes on Mother Nature.

So far, Microsoft gains when Intel gains, but that isn't necessarily the way it will work in the future, is it?

Might not. But then again, it's more likely that it would. Both of us are strong beneficiaries of a healthy PC business. We both need new people in the PC business, new customers. We need to sell those new customers richer and richer experiences which require higher performance hardware and software. I don't see a break in that. We have survived the disappearance of the umbrella between the two of us, which is IBM. We have survived a lot of technological changes. We are adjusting to the Internet and moving both of our product strategies, business strategies, to be based on the Internet. And all along, the complementarity of our core businesses remains. So, what would change that? Can people want operating systems without needing microprocessors? Can people need microprocessors without needing operating systems?

Ten years ago, you were a chips company, but you're not that anymore. And it's not just a question of your being X-times bigger. You're different. How?

Probably a simple explanation is we do a lot of technology work to develop markets for our chips. At an event we had today, we brought companies together who use our technology for their purposes, and we brought them together with a technology company that builds on our technology to service a wide variety of customer contact businesses. Now, in the old days, we would've worked with a Lotus on a spreadsheet or we would have worked with Word Perfect or somebody on a word processor, and you would have understood it very simply: that's a PC application, this is a PC, while we worked with them to make sure that that application runs well on a processor. Structurally, that's exactly what we're still doing. But now, the applications are far more varied and less obviously PC applications. Buying a ticket through Ticket Master is less obviously a PC application than a word processor. But what we are doing is very similar to what we did with word processors. We're making sure that those applications get optimized on our architecture, that they get designed knowing what our technology is going to do, knowing what our plans are. We had a handful of PC software applications, now we have 100 Internet companies that we have to do the same kind of stuff. So, the nature of the work didn't change. What changed is the scope. But that scope parallels the scope of personal computing. Personal computing used to be an office tool. It's now way beyond that. We supply the personal computing industry as the personal computing industry grows into places that it hasn't been, and we grow with it. So yes, we are very different than we were 10 years ago but so is the industry that we supply.

Are Intel's current chips too powerful for their own good? Given the power of the Pentium, and most people's rather modest computing needs, why should anyone upgrade?

Provided you keep the price at a reasonable level, there's no such thing as too much pro-computing power. No software developer ever complained about too much memory or too much computing power. They always have I to make compromises in what they present to the user in terms of ease of use, richness, performance, features, because they're bound by the confines of hardware. We are moving the confined hardware boundary up.

But as a practical matter, MMX to some degree cannibalizes Pentium. And doesn't Pentium II add to the confusion of the average person who is not a technology expert?

This year we're going to go through two very major technological transitions: We introduced MMX technology and Pentium II. They are two different microprocessors, two very major steps in performance. If we could introduce two big waves of technology in one year and not confuse the world with that, we would deserve a Nobel Prize. The fact is, it is a very rapid transition. In a perfect world, what we should have done is introduced MMX technology a year ago and Pentium II this year. But it is not a perfect world and they both ended up in the same year.

In some parts of the world, particularly in Europe, nut everyone seems to be concerned with having the latest technology. Considering that you've got 58 percent of your revenue coming from outside the U.S., how does this affect an important stream of future profits?

Approximately a quarter of our revenue is from Europe. We are aware of the sluggishness of the European market in adopting new technology anti new microprocessor technology. And we see the difference between how rapidly Asia picks up new technology and, by comparison, how slowly Europe does so. These people have fallen further and further behind in an age in which business competitiveness is very heavily linked to technological savvy. Wal-Mart would not be Wal-Mart without technology. FedEx would not be FedEx without technology. These are companies that are not technology companies, but they live and die by technology. If their counterparts are technologically less savvy, they're going to lose. So, either Europe will become more competitive or it won't. If it won't, their long-term position in the world of commerce and manufacturing is going to erode. And now that they have competition, not just from the United States, but from very hungry and very commercially aggressive countries, who are also aggressive in technology adoption, this situation is going to be worse. So, you watch the German economy in the doldrums, you watch the French going from strike to strike as a response to any attempt at improving their global competitiveness; what's going to happen? They either change their ways or retreat behind the protectionist curtain. A curtain never worked too well but it works less well today than ever.

In a way, that problem is now your problem, isn't it? How do you get that to reverse itself?

There's a limitation to what we can do about it. In Europe, you notice how few people carry notebook computers. It's a night and day phenomena. You get on a flight from New York to Chicago and everybody plops out a PC. You go from Frankfurt to Munich and nobody does. And this in the face of having the best cellular digital phone system anywhere in the world. So a couple of years ago, we said we have an opportunity to do something to mobile computing in Europe by marrying GSM technology, notebook technology, and we undertook something called a mobile data initiative with people like Ericsson, Nokia, Siemens, and various other people. And we have managed to stimulate the creation of GSM cellular modems. Now, that's a constructive initiative that we undertook to marry a Europe-based infrastructure technology with an Intel-based personal computing technology for improving both our business in Europe and Europe business in doing that. Will it amount to a significant change? I don't know. But wherever we see an opportunity like this, we will take it.

THE NEW PC AGE

What's your plan for transforming the PC into a communication and entertainment appliance?

What I see unfolding is a world of a billion interconnected computers. What we have today is probably 100 million interconnected computers. And that gives us this Internet phenomenon and electronic computers and all of the buzz and the excitement and the high evaluations of things that didn't exist two years ago and all of that. I don't know whether it's going to take us five years, or seven years, or 10 years, but in some reasonable time, you're going to have a world of a billion interconnected computers that will have the characteristics that every computer can communicate with any number of other computers. Any computer with proper authorization can get information from any one of those computers, and some of those computers are servers, some are clients, some have Web pages, some have Ticket Master-like information, some of those computers have virus protection agents that you buy or banking accounts, and they are libraries, they are your medical report that is accessible to any medical provider from anywhere in the world. It is going to change how life is lived. It's a much better way.

Will the PC replace the television set or the telephone or add to them?

It's not going to replace the telephone but I think it's going to cut into telephoning. The best way to predict something is when you observe something that's happening. The amount of time I spent on the phone at work, as compared to the amount of time I spend on the computer at work, crossed over about three or four years ago. I barely spend much time on the phone. I get outside calls, that's a minority. Internal to the company, almost all of our communications are through electronic mail. So it is not going to replace the telephone. I still have a telephone. But in terms of where I do most of my work time, or how much of my work time, there's been a shift taking place that can continue.

It is not going to be a replacement of the television as a device as much as a replacement of viewing pattern, viewing time spent. And I'll give you again a present-day example of this. Broadcast television viewership is in a major decline. Those viewers are not going to the computers. They're going to more numerous, more fragmented cable television. I see this as a continuum - big time broadcast to narrow-cast to niche-cast, and the niche-cast is the Internet. TV usage is down by something like one-third in households with computers.

It's not a replacement, it's a displacement. Displacement in terms of time. And that's a very significant phenomenon. So, it's a symptom of how dependent we're going to be on those interconnected computers. And that trend has accelerated since computer usage has become a connected usage.

On that theme, what is your view of the so-called network PC, or so-called simplified Internet appliance?

The interconnected computers will not be uniform computers. They will be powerful ones and older ones, thin ones and thick ones, muscular ones, and big disks and little disks - but they will all have to be compatible, at least in terms of this web operation. The limitation of the kind of devices you can hang on that thing depends on the usage pattern of those billion computers. If you think the usage pattern is going to be very simple, character-based, very lowbrow, basically like dumb terminals, then a significant portion of those billion computers could become very simple-minded devices. But I don't think so. If I thought so, I couldn't have given you that answer about television. I couldn't have given you an answer about telephones either, for that matter, because Internet telephony probably couldn't take place on those simple-minded devices. I think the usage pattern is going to be rich. It is going to be human-friendly in characteristics: visual, 3-D images with points of view, video rich sound, possibly voice recognition.

But couldn't the server handle all of that or most of it?

A server could handle those but at the expense of much more load on the server, and at the expense of bandwidth requirements that are beyond what is likely to be available, certainly in the consumer space.

Java software, long term: friend or foe?

Java software is friend, absolute friend - because it brings network richness for the applications designable for networks much more effectively than many other languages.

It has been written that the cost of supporting and maintaining a PC is many multiple times those of its replacements and that this is going to be an issue when companies choose to replace PCs or think about network computing. What role will Intel play in this?

We are playing a major role in building in management, network management, which is the most important asset. These computers, when they're wired, they get more expensive. But the very act of wiring them also gives you the escape in allowing a remote help desk, a remote administrator, or a remote machine to fix those problems that require human intervention today, which is where most of the cost is. So the problem contains the seeds of the solution and it's up to us to engineer solutions that come built-in to every computer. It's a major thrust for us.

Gartner and others say that the total cost of maintaining and servicing computer PCs are around $40,000 per desk top. What do you think of that estimate?

I think it's going to come down over time to about half of that in five years. Most of the problems are software-related, most are user-induced, and all those problems can be remotely solved. One capability we have today is you can have a computer turned off and a machine can "wake" that computer up, turn it on remotely and update the software while the worker is not even there. We can build those into a computer, and we are building those things into computers today.

BIGGER THAN ANY OTHER THING

Intel's overwhelming market dominance inspires awe - and fear. Just as Microsoft has sometimes been branded in the industry as "the Evil Empire," Intel is sometimes depicted as the dark force, or as a threat to the competitive environment for innovation. Do you believe this is an issue for Intel and the PC industry?

When all else fails, it always helps to look at some facts. And the fact is that we brought down the cost of computing year after year at the rate of 30 percent per year. The fact is, we've increased the power of computing, which is doubling every 18 months. The fact is, in a very competitive industry where we were almost at a death bed due to other competitive factors, we've become the leading architecture on which computing is built, not by some kind of a state-endowed monopoly but because we earned it, one design after another and one shipment after another by a bunch of discriminating and sophisticated customers. Name-calling is cheap. Innovation, cost reduction, performance improvement, year after year after year is hard, and I think that's what we have done.

What did you learn from the Pentium-flaw flap of a couple of years ago? And how did that experience Inform your reaction to the current one over Pentium II?

It has had a major impact. The most important thing is, we have to respect our customers, our users, and part of that respect is not making decisions for them. What was wrong - and it's most fundamental with the Pentium problem - we decided on our own that we knew best, that the problem was not important, and everything we did was based on that attitude. We have had lots of problems, and pushing technology involves getting into these problems. I wish it didn't, but it does. We have followed a very simple principle from that: we inform and let the customer decide. That's what we did here. It clearly worked a lot better. It absolutely makes no difference whether we were right or wrong in the first instance, technically.

I imagine a number of CEOs, not technology CEOs, have probably asked you "Is this Internet overhyped or not?" What do you tell them?

It's both overhyped and underhyped. It is overhyped in the near-term and it's under-appreciated in the long-term. It's not this Internet thing, but a world in which a billion computers and different kinds are all connected with each other in this web. You're going to change education; you're going to change health care; you're going to change social policy in ways you can't even begin to understand; you're going to change the relationship between companies and their customers in very profound ways. It is going to displace large portions of the population who are in that chain of the relationship between customer and supplier. The magnitude of that could be really breathtaking. I don't know how broadly people appreciate it. It may be a measure of unease about it, but not a profound appreciation of it. In the meanwhile, near-term people are saying, "Where is the such-and-such that people were talking about a year ago and is still not here?" That's what I'm saying. So, in near-term it's overhyped, long-term I think it is going to be bigger than any other thing. I kind of fear it's going to be bigger than what I'm thinking also. It's very hard to keep up with it.

So what advice would you give to a non-technical CEO who asks: "I'd love to get my hands around this. I'm not in the technology field. What can I do?"

Your question contains the answer: get your hands around it and overcome your pride and your fear and get yourself a tutor and learn the elements of it yourself. Not presentations, not charts. Sit down with a computer and try to experience it an hour at a time so you gradually develop a context in which everything you read starts to mean a little bit more. At the Herb Allen Sun Valley conference, we're going to put on a training course for the attendees there. I'm not going to give a talk. We're going have a bunch of Intel training people there. We're going to drag attendees in there and tell them, "Just check your embarrassment outside." Then we'll teach them: this is a mouse, this is a browser. Hey, you've got to do that. We did that for our own board; we did that for our own management.

You and Gordon Moore are driving near Roswell, NM, and are abducted by aliens. The board is forced to appoint a new CEO. What does this new CEO do?

Reorganize the company to provide the elements from inside the company that we are supplying today. And we are part of the Intel environment, so snatching us out of there will upset the internal balance that needs to be deliberately addressed. In terms of Intel strategy, I don't think we have a major thing to do right now. A year from now, or six months from now, it might be something. But if I had a little capsule that I wrote in case I disappear altogether, there wouldn't be a business to get out of. I think we are execution-limited in what we try to do because of factors like Europe, communication, bandwidths. We know that. Ten years ago, we were frozen at the stick, frozen in strategy. It was clearly not working for us, and we've lost a lot of market share, and we needed to pump ourselves up to do something that we kind of knew we needed to do but we didn't have the courage.

You talk a lot in your book about strategic inflection points. What were the one or two most important SIPs for you? Would it have been your fight with prostate cancer, for example?

That's a daunting and awesome experience, but I don't think it changed me. It changed my diet, my behavior, and my worry patterns. But I don't think it changed me as a person in particular. That incident in '85 was a strategic inflection point for me. I was an operations guy, part of the R&D, engineering, manufacturing, and almost overnight I learned how important it is to think through what you're trying to do and articulate it and get everybody onto the same page. It was a brutal lesson, but that was one. I think getting out of Hungary and getting to the United States was a very important one.

In Hungary, then, you had to suppress your opinion, your views. When I got out and got to this country, there was almost a kind of a break-like resolve. I'm not going to pretend. And there's a lot been said about how blunt I am and that kind of. thing. It's an outgrowth of that. I had to pretend to believe in things that I didn't believe in before. I don't have to do that. I'm not going to do that. So, I probably haven't been a perfect follower of that feeling. But that was a breaking point.

I think that Pentium episode had a major impact on me. I like to think - one kids himself on matters of this - but I like to think that I have personally become less self-righteous as a result of that incident. Personally, I was so damn sure I was right. So that was a humbling experience.

Were you self-righteous before?

Well, I still am. That's a fault of mine. But I think that was a big break. It's not that it's gone from me. It's tempered significantly. I give the other point of view more room than I used to.

You'll be 61 in September. Sixty is a time when people take stock of themselves, look to the future. When you look back on what you've accomplished, what are the things that stand out in your mind?

I'll tell you, I'm happy with my career. I'm delighted with Intel. I'm delighted with Intel's role. I'm just very comfortable, very pleased with how we did it, a few minor aberrations to the contrary. But we did it basically with an eye on the ball. We have three constituencies: customers, employees, shareholders. We've treated them well over a period of time. We're not a flash in the pan. We've built an institution that's capable of doing that. And I don't think I've made major personal compromises. I've got family that I'm happy with, and that most of the time is happy with me. I've enjoyed my life. I've enjoyed my work.

Gordon Moore is a formidable presence. Did you ever feel that you were living in his shadow?

No, I don't think so. At the annual meeting, I went through a little process of acknowledging him. I said I was very fortunate 34 years ago to have run into him and to have met him because it's been a 34-year working relationship. He never made me feel I was in his shadow. I always felt like I was his partner. And it doesn't feel any differently now.

So in that respect, will Craig Barrett be your partner?

Yes. He has been. That's the art of this transition. There used to be two people, then there were three of us, and there was one who unfortunately died, and there were three of us again. In the meantime, Intel prospered and grew and retained these essential characteristics. I just want to continue with that, so I can look back and say, "Hey, I was there, grew with it, I did this, I did that, served several different roles. And it's now time for someone else to continue." And hopefully, like Gordon, I would play a constructive role beyond that but not in the same position, so you have to write about somebody else.