"The important things of tomorrow are probably going to be things that are overlooked today." — Andy Grove, CEO of Intel

On Tuesday morning at 8 A.M., roughly 120 glazed eyes were focused with laser-like concentration on a Stanford business school professor as he entered our classroom. It was the beginning of my second year in business school, and, in this particular class, our professor was Andy Grove, the Chairman and CEO of Intel, a multi-billion-dollar organization and one of the largest companies in the world. The B-school rumor mill diligently had spread gossip about Grove—about his tenacity, distaste for tardiness, his spartan discipline-and many of us were both excited and intimidated by the prospect of being taught by him.

While I must admit that I am still slightly daunted by him—perhaps because of his remarkable achievements-Andy Grove is really a down-to-earth person and very approachable. He and Bill Hewlett were indeed this book's "bell cows"—once other CEOs knew that Grove and Hewlett were participating in Giants, it suddenly became surprisingly easy to land interviews with them.

The immigrant son of a Hungarian dairyman, Andras Grof was born in Budapest, Hungary, in 1936. In school he dabbled in opera and journalism. After the 1956 Soviet invasion of Hungary and subsequent "nationalization" of his father's business, Grove escaped on a boat filled with refugees, and landed in New York with only $20 in his pocket. Three years after landing on U.S. soil, Grove graduated first in his class from the City College of New York with a degree in chemical engineering, paying his way through college by being a waiter.

Three years later, he obtained a Ph.D. from UC Berkeley and used his writing skills to craft Physics and Technology of Semiconductor Devices-considered even today to be the seminal introduction to semiconductor engineering for students.

Grove left Berkeley to work for Gordon Moore and Robert Noyce, founders of one of the world's first semiconductor companies—Fairchild Camera and Instrument, later named Fairchild Semiconductor. After being dissatisfied with Fairchild's progress, Noyce and Moore went on to found Intel (INTegrated ELectronics) and brought Grove with them to head the R&D department. After a stint in R&D at Intel, Grove became COO in 1976. In 1979, he launched Operation Crush, a campaign to capture 2,000 new customers away from Motorola within one year. Intel beat that goal by 500 customers, one of which was IBM.

In 1982, IBM approached Intel to provide its 8088 microprocessor chip as the brain of IBM's entry into the microcomputer industry—the IBM PC. Ironically, Intel had already developed and patented the microprocessor in 1971, but did not expect significant demand for it.

However, in 1985 personal computers weren't shipping in enough volume. Intel had defined itself as "the memory company," yet its memory business evaporated. The Japanese chipmakers had driven prices so low that they crushed their U.S. rivals in the huge market Intel had invented. Intel, the industry leader, lost money for six straight quarters from the withering attack of Japanese firms, and many in the industry doubted Intel's ability to survive. Over the objection of several executives, Grove axed the DRAM business and thousands of employees. Unwilling to let the company come to such a pass again, Grove focused Intel on microprocessors with a paranoia and manic competitiveness that set the culture for the company today. Indeed, Intel's overarching theme of "Only the Paranoid Survive" (also the title of one of Grove's books) epitomizes the current Intel culture.

Intel's revenues grew as the personal computer industry mushroomed. Its microprocessor design quickly became the de facto industry standard for the bevy of new computer and software developers just starting up. Nowadays, the words "Intel Inside" are the words by which many Americans judge the quality of the computers they buy, resulting in about a whopping 80 percent market share for Intel.

Yet, despite the fact that the company's stranglehold on the microprocessor market shows no sign of loosening, Grove has not rested on his laurels. Intel spends a significant amount of its revenue on R&D, and consistent with his paranoia, Grove still considers the industry and his competitors to be brutal.

Intel and its competitors—most notably Advanced Micro Devices, or AMD-often take their battles to court using hardball legal tactics. Grove has referred to AMD as "the Milli Vanilli of semiconductors. Their last original idea was to copy Intel. Since they can't win in the marketplace, they try to defeat us in the courts and press." Yet, polemics aside, Intel has managed to keep its smaller competitors from undercutting the company's business. AMD has always hovered around 10 percent of the microprocessor market.

Despite the furious competition, Grove manages to teach high-tech strategic management at Stanford's business school, and also is a prolific writer. He is the author of three management books: High Output Management, One-on-One with Andy Grove, and Only the Paranoid Survive.

We had breakfast with Andy Grove at Hobee's, a popular restaurant in Mountain View, California.

Interview Transcript

What do you think your talents are as an entrepreneur and as a manager of a large company?

One thing is important for me to correct: I never looked at myself as an entrepreneur, and even today I don't look at myself as an entrepre-neur. In Intel's case, the entrepreneurs were Bob Noyce and Gordon Moore. Gordon himself is a little uncomfortable with the word "entrepreneur." In speeches he'd describe himself as the accidental entrepreneur.

Now, that's not the case for Bob. Bob's really more of an entrepreneur. But, I just tagged along for the ride. That changed later.

Much later. But in the early stages my expectations were that I'd be head of R&D for this new company. R&D is what I did at Fairchild. I never pictured myself doing anything substantially different. But, I rapidly veered away from that ambition when I became the operations guy inside Intel. That's what I did for almost twenty years. I helped make Intel's transition from an entrepreneurial outfit, none of which was my doing, to the manager of a medium-sized company, which was actually a very difficult transition to make.

Why was it difficult?

Well, everybody can start a startup and come out with one product— you see that all over. But, making the company into something of a self-sustaining institution, with its own methods and organization—that's tough. Transitioning people from a completely task-oriented mentality to more of a process-oriented mentality, particularly since those people were attracted to a startup, was very hard.

So, I wasn't comfortable with the role of the entrepreneur as you define it—somebody who starts a business and sees a business opportunity. I was not even qualified for that. I was a researcher. But once I began to see the architecture for the company as an organization, I became more of a participant.

So if Robert Noyce hadn't left Fairchild then you wouldn't have gone?

The person I followed was Gordon. I worked with Gordon at Fairchild and actually, I wasn't that enamored by Bob [Noyce] at Fairchild. I knew him, he was two levels above me. I got to know him better later, ton and Bob was not comfortable, Fairchild was a troubled organization and Bob was not comfortable running a controversial, troubled company, so he became passive as the company was spinning out of control. So I didn't have high regard for him at that time. As Gordon was leaving, I immediately volunteered to go with him. Then, I found out that Bob was leaving, and I had momentary second thoughts. That's not the conventional wisdom because Bob was very highly regarded.

You've mentioned that Intel was a collection of strangers working together, and that made things difficult in the beginning. In retrospect, do you think there were pitfalls there that you could have avoided in that respect?

No. There are pitfalls there, but I don't know how you can avoid them. First of all, the fallout didn't depend on whether we were or weren't strangers. That attitude went away after a while. What made it difficult was not that we didn't know each other—what made it difficult was that we didn't have any rules. Nobody quite knew what to do. Then, a lot of the people—including quite possibly myself— started developing a pecking order. Influencing and elbowing gets so illogical, and so inappropriate for a group of people who are trying to get something going. In fact, I think that there was more politicking and game playing at Intel in its first year—at least in senior management—than since its history afterwards. It's not what you would think.

How, then, does one turn political shenanigans into cohesiveness? Is it something that just happens as a matter of course or do you try to mold the culture?

(Long pause] It's hard to answer that without getting awfully person-al. I don't have a big theoretical model or some five-forces diagram here. What I can say is that in the fortunate companies, certain operating methods emerge. And, at Intel, some methods of influence emerged.

For example, I didn't have an easy time with our first marketing manager and I couldn't spend a lot of energy marketing. I was the operations guy at Intel. And so, what happened was that we were second guessing each other and fighting, and then we got a new marketing manager three years later. That was a major breakthrough because we got along right from the get-go. I'd have to say that there was a personality conflict that did not exist with the second guy. But, who knows? Maybe by the time the second guy came in, some shape and order to the place had emerged. Had he been there from the beginning, we still might have been ridden with conflicts. I don't know.

The problem with a lot of these stories is that they are very personal and very self-centered. I'm telling you the story from my perspective and, on top of it all, filtered through 25 years of memories. Those stories make what I did seem wonderful and what other people did, less wonderful. But all our histories are like that, I suppose.

Anyway, that particular interpersonal conflict was the most important to me, but by no means the only one. We had productive people flip out or go through a kind of burnout process. When you're 30 years old and encounter that... It's not an ideal startup situation.

What was the first critical juncture at Intel-the fork in the road that could have made you or broken you? Was it when IBM decided to use your products?

The first critical juncture was the introduction of our first successful product, the 1103—the first DRAM. It was enormously difficult to get it into production. DRAMs were so difficult to make and test. We were learning totally new things—a new process, a new product, new testing, new customers. And, customers were seriously relying on these things for mainframe memory. We weren't playing with calculators here. We were playing with big time memory, a very difficult period.

So, that's probably what I'd put my finger on—this process and these events when we became a company. Up until then we were a new startup just coming out with our first products, none of which were that successful. Then we made the 1103 and it became very successful. That's when Intel's systems as we know them today emerged. Had we blown that [product] we probably would have followed a completely different trajectory. We might have survived, but with the introduction of the 1103 into market, the real Intel was born.

From the get-go, did you think that the 1103 was the way to go? Was it part of a master plan at Intel?

No. We found that the 1103 as going to be a sexy product. And, in the last year of production, we had to systematize it. It was the end result that dictated the style. There was no master plan whatsoever. We were running out of money and licensed the products and technology to another Canadian company. Things were much simpler and much more haphazard in the early '70s than they are today. In the midst of this the microprocessor was born. That was not a big event, though. It was an interesting thought and nobody paid a whole lot of attention to it.

It [the microprocessor], was done in response to a customer request, right?

Yes. We obviously didn't realize how important it was.

Do you think that's true of many important technologies that you don't realize their importance?

Probably. The important things of tomorrow are probably going to be things that are overlooked today.

The important thing to realize is that a lot of major business decisions, such as Intel's shift to microprocessors, are not that obvious. In retrospect, they might be, but not when you're looking for-ward. If you read the newspapers about Apple, what they are going through is very similar. Nothing is obvious about Apple-other than the fact that they are in trouble. That's very obvious, but what to do about it is not obvious. If you lined up three or four of the people in your book and gave them a blank sheet of paper that said, "If I were CEO of Apple I would...," each would have a different answer.

It was like that at Intel in 1985. You could have asked five people where the company should go and you probably would have gotten five different answers. In retrospect, we know what happened with the 386, but up until the 386, the microprocessor was really not that big. Our new product was very complex. When we threw ourselves into the microprocessor business, we didn't know if we could make it.

This was 1985-before we introduced the 386. And the microprocessor business was much much less than Intel's memory business at the time. So, everything was very murky, and it wasn't an obvious You asked about when IBM adopted our product. I didn't think it vas colossally important at that time. And, I maintain that nobody at Intel did.

But, here's the rewriting of history: We just had the ten-year anniversary of the PC a few years ago, and the Intel newspaper want ed to interview a bunch of people about it. All of a sudden everybody remembered what a big celebration our microprocessor was at the time it was introduced. Well, I must have been on vacation. I don't remember any celebrations. I don't remember anything like that. It was good, but keep in mind that IBM thought the lifetime production of PCs was 200,000! IBM thought that the projections were too optimistic in the first place. Why would we have been more knowledgeable about how successful IBM was going be with the PC than IBM?

So, it was not a big deal at the time. What was a big deal was then they came around a year later, unrelated to the PC, and wanted to buy into Intel.

And they actually did take an ownership stake.

We struggled with that, whether we should let them or not. For some reason, we turned it into an existential debate: Should we or should we not? Finally we did, and the whole thing turned out to be a non-event. They never did anything related to that investment with us and they sold our stock. It was a joke—"a long-term investment" that a few years later they sold because the stock went up and they wanted to bail-out their earnings. So it turned out to be nothing. But at the time it was, "Oh, my God!" All of a sudden I was flying to New York and was meeting with the Chairman of IBM for a big announcement. It was a big deal at the time, and yet turned out to be nothing. That was in '82.

But, the big milestone was the DRAM case [shifting from DRAMs to microprocessors]. That was the crisis that shaped the Intel of today. Just about everybody in management thought it was the end. That's where the paranoia and all that stuff came from. You can't go very far at Intel without running into people who went through it.

My father worked at Intel during that time, at the Santa Cruz plant, which was shut down.

What did he do?

I don't know. When you're young, you never know what your dad does.

I had no idea.

So, when you decided to lay off a third of the workforce, obviously that must have been difficult. How does such a decision get made?

Well, we never laid off a third of the workforce in one fell swoop. We wanted to do. I have this saying: you never cut enough soon enough.

At the time, Gordon said, "Let's make sure we cut enough so we can take care of all of this in one layoff.” And, I said sure.

We tried to do that. But, we had just barely finished the layoffs, and it became clear that our business had sunk some more. Nobody ever forecasts that revenues will go straight down. Everybody thinks the low point has been reached.

So, we went through excruciating detail figuring out who we needed to let go. It wasn't enough so we kept doing it until finally it was enough. The transitions that we talked about earlier made us a company: employees, organizations, doing things correctly, and doing things well. But when we had to cut people back—we did that very seriously too- from the standpoint of doing it right from the employees' standpoint. As right as you can do it. Did your father get laid off?

Yes.

Then this will be a weird conversation. The prevailing industry practice at that time was to cut back on seniority-where the most recently hired person goes first. The justification was that, unless you cut back that way, you gave unions an opportunity to come in and say, "We're going to make you guys do it right." I took a very strong exception to that. We had some layoffs in the '70s-also pretty big for a small firm—and the personnel people wanted it done in a very organized way. I really leaned towards us laying off on the basis of merit, which we didn't. We gave it lip service but it was too complicated. By the time this '85 situation came up, we went to the point where we took people's performance reviews and systematically tabulated the ratings of the performance reviews and all that stuff. We did it on the basis of measurable, documentable performance. That was the compromise. We were going to do it on the basis of performance, but it had to be objective enough, so that if somebody questioned it later, you knew you had a well-defined system. We used seniority to break the ties. We spent incredible time working on it. I remember the first layoff—it was in Oregon. I went through the list name by name by name to make sure. There were hundreds of names. Most of them I didn't know, but I wanted to make sure that we were really doing what we said we would do. We took layoffs pretty seriously.

In 1982 we had a bump in our business and everybody else was laying people off. Luckily we thought it was just going to be a blip. We turned out to be right. We decided to stop hiring and did all kinds of weird things like making people work an extra ten hours a week for the "125-percent solution." We really held expenses down and stretched our employees' days longer. Our people did that literally for six months. Then we gave them a big party. Then we had another blip and we cut pay by 10 percent. So, we tried a bunch of different means.

We were very anti-layoff, because in those days people laid off left and right. Management didn't give it a second thought. We were trying to build a solution—you can't treat your people like an expense item. We really tried [not to have layoffs), but in 1985, it got beyond that. Actually, I didn't agonize too much over the Santa Cruz plant.

What I really agonized over was Barbados, an assembly plant that largely built memories. There was nothing in Barbados but us. It was a resort. You would either go fold sheets at the resort or go work at Intel. It was a very stupid thing for us to put a plant there. It was very very difficult to run, but by 1985 it was a good plant. There was no history of any industrial undertaking on the entire island of Barbados, so we had to teach everybody. We went through plant manager after plant manager until we finally got it right and the plant was good. They were so proud. It was a high profile thing.

The layoffs were a major part of our maturation as a company.

So, given the choice between hiring people and laying them off in bad times versus not hiring enough people and making them work at greater expense to the company, which would you do?

The second. I have a very strong personal dictate that we as a company not lay people off again like we did before. The memory of the incident remains.

Perhaps you'll never have to make that type of decision again because of the industry you're in. It seems like you're in a pretty nice industry. Do you ever see a downturn in the semiconductor industry?

Well, it did go down in '85. There were all kinds of arguments about the semiconductor trade agreement. We were in bad shape as an industry then. There was no PC industry to keep us up. I don't think semiconductors is a nice industry. I don't think it's a cushy industry. It's a very tough industry. Look at the people that we compete against. You don't know the meaning of competition until you've dealt with a Samsung. We have had incredibly dedicated competitors.

So, which one of your competitors in particular concerns you the most?

Nobody in particular because competition is not the issue in our case. It's in navigating change from the fairly well-defined PC industry into the network computing industry and it's very, very complex. I think probably my biggest fear is the process of navigating into the next stage of computing. Whenever microprocessor power becomes less important than before, that means a loss of business for us, much more so than a loss of business to a competitor. Competitors are much easier, this threat is just kind of ambiguous. There exists a rising tide that controls the industry. Competitors either matter or don't matter depending on that tide.

Speaking of this tide, then, do you envision the idea of just an empty set-top box with a network connection?

No, I don't see that, but what I do see is the need to develop the type of application that uses the hidden, latent microprocessor power. At home, which is where all the talk about appliances are, the benchmark set by television, the visualization of television is what people expect in terms of processing power. So we're going to need processing power, but that's only half of our business. The other half is with businesses. So we have to take the initiative on stimulating applications, visually compelling applications, business applications. That's our job, but it's not about competitors.

Or, increase demand.

Right. We have to generate our own demand.

Let's switch gears a little here. Why do you teach at Stanford?

Because I enjoy it. It's very simple. I've always liked teaching. I taught both during my engineering days and then during the period of time when Intel made the transition into more systematic management. I started teaching management practices at Intel, out of which came High Output Management.

I like exchanges. Harvard did a case on Intel. I became aware of it, and I talked to the class. I liked that. I wasn't going to go back and forth to Harvard each year to teach the case so I approached Stanford and really liked it. It's a laboratory to try new ideas. Cases allow you to think through a particular industry much more. The process ended up being very helpful to me in understanding the interaction of the internet, the World Wide Web and the like. I didn't really learn any new facts about it, but preparing the case and going to class really helped. That's a secondary reason.

Given all of your extra-Intel activities, such as writing a management column, and teaching a course at Stanford, what kind of hours does the CEO of Intel put in? Do you work less or more than before?

It hasn't really changed over time. I spend a lot of time thinking about the industry. Our industry has an impact on every other industry. Telecommunications—I used to pay no attention whatsoever to the telecommunications industry. It has now become vitally significant. International events have become vitally significant. Now the media industry is where the telecommunications industry was a few years ago—on the fringes—it gets increasingly significant. We capture people who are in the media and going into digital technology.

We capture their efforts on PCs. That's our business.

So, I have to read a third of the newspaper, even though it doesn't mention Intel or the article doesn't have anything directly to do with Intel. In that context I think a lot about the world. Not so much about Intel, per se. So what are my days like? I start at 8 o'clock, end at 7. I've got another hour of work that I do at home.

Seven days a week?

No, five days a week. And over the weekend, I have maybe three to five hours of work. Not administrative things, but really the longer things that I didn't do during the week. I don't do a lot of administrative work anymore. Craig Barrett [President of Intel] does that. I figured I'd written all the performance reviews that I needed to write and I washed my hands of all of that some years ago. I really do what I want to do, but what I want to do adds up to be more than I would like it to be. This is a problem. I will be sixty years old this year and soon l'll probably have to stop. Ten years ago, I said I would retire at age 55, and I don't want to hang on too long.

Today the semiconductor industry is not one that a young Ph.D. can get into and start a company on a hope and a dime.

Not on a dime.

Many dimes. If you were starting over today, what industry would you recommend to young people trying to start a company? Would it be in semiconductors?

First a point about startups. I can't look at a startup as an end result. A startup to me is a means to achieve an end. A lot of your business school friends come up to me and say, "Hey, Andy, I want to do a startup. I want to do an Intel. So, what should it be?"

It just doesn't work that way. Instead, you should first figure out that you want to do, say, semiconductor memory. You want to do semiconductor memory so bad that it hurts, and you can't do it where you are. So you do it within a startup. Otherwise, it's such an ass-backwards way. People who did it exactly like that saw a mountain of opportunities that weren't there.

If the question is what field would I go into as a young person today I would probably get into genetics or biotechnology. I know very little about it, but it fascinates me from the following standpoint: it is the new frontier that can use computing power and increasingly even semiconductor technology in terms of preparing and building genes. It is in converging technologies which I always liked.

I see the possibility of cross-fertilization of genetics or biotech with information technology, which is very advanced and very powerful. Computational tools are very powerful. The concepts are very powerful. Software designs are very advanced. Semiconductor technology is very advanced. I'd apply it to a field like genetics which is still very, very raw. It's kind of like the semiconductor industry was in the 1950s. So I would prefer to go there. If I had to go back to school again, I would get a double degree in computer science and genetics. I don't exactly know what I would do with that, but it would be a combination of those two subjects.

You wrote an article for the Wall Street Journal about the importance of immigrants to this country's future. How do you think immigrants have helped Intel?

Fifty percent of our Santa Clara technical staff is Asian. There's no question about it. The degree is not as high in the other locations, in Oregon it is probably 20 percent. Our company is interwoven. The man who discovered EPROMs is Israeli.

It's very interesting that you mentioned the article. I wrote a regular management column. I never got responses to any of my columns. But the response to this one article was incredible e-mail after e-mail. I had a lot of feedback. In one message I got a one liner, and in another, three pages of philosophizing.

Why?

I don't even know. Are you asking a larger question?

The political thinking in the air today. The Gingrich approach to things. Proposition 187. It has become an atmosphere of intolerance and, although it may be just a minority standpoint, this attitude seems to be more prevalent than before.

It's incredible to me that with the monumental problems that we have financially as a country, there are people who target welfare mothers. It rubs me the wrong way. It really does. I had a very positive experience as an immigrant here. Nobody has ever given me a hard time. Nobody has begrudged me anything. It's never been an issue. I've become a rabid American as a result of it. I haven't gone back to Hungary at all. I have no love lost for my origins. But I hate seeing this kind of thing developing. You know every trend in this country starts in California. I can't say that it's a major trend yet, but it's enough of a trend that California has Proposition 187. Think about it.

Is there anything you can do to change that?

Probably the most important thing I can do is to run Intel in a good way. Intel's a big community, you know. There's very little I can do outside that's going to affect a community of forty-five thousand people. You can take that number and multiply it by four for the number of people that are affected in terms of employees, families, and customers. What I do at work, in some ways, affects hundreds of thousands of people. That's pretty big.

For more, order a copy of In the Company of Giants: Candid Conversations with the Visionaries of the Digital World.

Browse more of history's greatest speeches →

See Also

Find more from Steve Jobs and others related to this lecture:

• Who was Steve Jobs? Wisdom From The Man Who Built Apple and Pixar.
• Steve Job's "The Lost Interview from 1995"
• Jony Ive's "CCA Commencement Address and Advice for the Graduating Class"
• Steve Jobs “Stanford Commencement Address“ in 2005
• I, Steve: Steve Jobs In His Own Words
• Make Something Wonderful: Steve Jobs In His Own Words
• Insanely Simple: The Obsession That Drives Apple's Success
• Designed by Apple in California

Learn more about Steve Jobs: Who was Steve Jobs? Wisdom From The Man Who Built Apple and Pixar →

‍