Keynote Transcript

Los Angeles Times 3rd Annual Investment Strategies Conference

Andrew S. Grove
Los Angeles, Calif., USA
May 22, 1999

JAMES FLANAGAN: How do you do, and welcome. I'm Jim Flanagan. […] my purpose here is to introduce a man that I met 20 years ago. I went up to Silicon Valley, and I was told, "Oh, here's an exciting company and here's the head of the company."

And I went into what looked like a makeshift building and I said, "Where is this fellow?" And they said, "Oh, the president is over there." And here were cubicle dividers, and behind the cubicle dividers was a desk, a computer, and a man, Andrew Grove. And you looked at that and you thought, well, wait a minute. What kind of business is this? And what it was was a business for the present and the future. It was stripped for action. It didn't have mahogany paneling on the wall, he didn't have pile carpeting on the floor. He had energy. Lots of energy.

I mean, if there's one man that embodies energy and technology and vision and hard work, it's Andy Grove.

He came from Hungary in 1956 and within four years he earned a chemical engineering degree from the City College of New York, followed it up in a doctorate in chemical engineering from the University of California at Berkeley.

He participated in the founding of Intel in 1968, and he became president in 1979. That's when I met him. 1987, he became chief executive officer of Intel. Intel's stock at that time, adjusted for subsequent splits, was about $2 a share.

It was last year, he relinquished the chief executive's reins, but he is chairman of Intel, to Craig Barrett, and Intel stock was selling for 30 times the $2 price that he had when he took over. And, of course, the beat goes on.

Intel's stock today sells for 35 times that old $2 price. Andy Grove is a fascinating man. He's a man capable of writing books such as "Physics and the Technology of Semiconductor Devices" and a heck of a business book called "Only the Paranoid Survive." (Laughter.)

JAMES FLANAGAN: He's a man capable of being named Time Magazine's Man of the Year. That's an honor that's gone to people, as you know, such as Winston Churchill, Franklin Roosevelt, Martin Luther King.

Why Andy Grove, who was named in 1997? Possibly because he embodies the best of a new way of business, of technology. He embodies energy, vision, hard work.

It's a pleasure and an honor to introduce Dr. Andrew Grove. (Applause.)

ANDREW S. GROVE: Thank you, Jim. The biggest change in my life, work life, since Jim met me in that cubicle is that my cubicle shrank. (Laughter.)

It's all part of the magic of the semiconductor industry. We shrink everything. And we are hiring smaller and smaller people to fit them into smaller and smaller cubicles. (Laughter.)

What I would like to talk to you today about is the Internet economy. But first, I want to tell you a couple stories about what the Intel economy represents to me. And the stories all come from the last few weeks.

I met a gentleman a few weeks ago in New York. He runs a small business. The business is about building wooden garden toys and swing sets. It's a business located in Maine, and about as conventional and old-fashioned a business as they come.

Very casually in the course of the conversation, 25 to 30 percent of his sales come from the Internet, and the biggest unanswered question is how do people find his company's name on the Internet. But they do.

During that same trip to the East Coast I had the opportunity of visiting one of the old-line brokerage firms, one of the biggest names in the business. Probably a lot of you are customers of this firm. And I met with the chairman of the company who described that their strategy is to turn the company into a dumbbell.

The dumbbell that he has in mind on one hand is going to be the traditional brokerage business, advisor-assisted, broker-assisted brokerage as it has existed, and on the other extreme there's going to be an online brokerage arm, completely unbundled services. You pay for the services that you need, all the way down, if you don't need any services. It is going to be competitive with the up-start online brokerage firms.

All of that is not because they are particularly eager to go in that direction but because the customers are demanding that.

And lastly, the finishing touch in all of that, I read an article about Ford under the guidance of one of those Generation X managers starting a business, and the business is the Internet-assisted centralized junk-yard business. We all know what junk yards are. They are about the best example of hands and nails and elbows distributed businesses.

Ford's idea is to concentrate the process of stripping cars, storing the warehouses in a modern warehouse, cataloging them on a database, and making the database accessible to garages, repair mechanics, all over the country, and then overnight shipping the part. And you can just picture, the efficiencies inherent in that are incredible.

The point of all of this is the Internet is penetrating everything, every business, and is allowing each of these businesses to redeem themselves.

And now I want to talk about how various of us in the information technology industries are working to enable this Internet economy.

I must share with you a sense of bewilderment from an investment standpoint as all of this phenomenon unrolls. And almost every day you pick up the newspaper and you see mega-billion dollar mergers of companies that somehow fit into this economy.

It is discovering a new era without a history book. It is like discovering a new continent without a map. You must feel something very, very similar to that.

And I can't really help you with a precise map, but I thought what I would like to do is give you a bit of a historical perspective of where this current phenomenon has come from and make some observations; maybe it's a little presumptuous to say the "rules" of the Internet economy because the Internet economy is being formed day by day so the rules are being written as we go along, but some observations about what might become the rules of it; and then give you a little bit of perspective from Intel's standpoint. So that's kind of my agenda for what I'll talk about.

So going back a little bit, one of the most important phenomenon that we have recognized, all of us have recognized in some fashion, is that the high-tech industry in various ways has become the engine of the U.S. economic growth.

How does it do that? It does that by increasing the productivity of business processes and business enterprises both internally to each company and the productivity of companies working with each other, in a fairly substantial fashion. And it is exactly that productivity that has allowed the economy to grow in such a way that inflation has been kept in check.

And various people have tried to estimate this, but the best estimate gives about one percent per year growth rate, non-inflationary growth rate, over what we would have otherwise. And being investors, you're all familiar with compounding at one percent per year growth. That one percent per year growth compounds to a trillion dollars in gross domestic product that otherwise we would not have in a ten-year period of time.

So this is really the impact of the high-tech industry on our economies.

I can't say we haven't benefited ourselves in the high-tech industry because the next chart shows the participation in the gross domestic product. And it was in the one to two percent range, and it is skyrocketing.

What are some of the secrets of this economy? The simplest way of describing it is it operates by what I call "PC-style" economics. The PC is the great leveler of enterprises. What do I mean by that?

Before the PC started, the industry back in the early '80s consisted of largely vertically integrated companies. Companies like IBM, Digital, Bull, Nixdorf, a number of international computer companies, were all completely integrated. They build their own chips, they build their own software, they build their own computers, they sold their own computers, and then they competed one block against another.

The introduction of the personal computer brought a completely different economic structure, business structure, and competitive structure to this industry. It horizontalized the industry. By and large, a handful of processors got adopted and got used by all the companies, a handful of pieces of software got used industrywide through all this industry, and on and on and on.

The industry became horizontal also in terms of competitiveness. People who resided in each of these horizontal bars competed with each other very ferociously, and all of that gave an enormously productive, enormously competitive offering, much more so than the vertical industry did before.

Intel has operated as a supplier to the personal computer industry for many years, and after a period, we made some observations of what it takes for somebody to win in this type of an economy, this type of a PC-style economy.

And the rules that we came up with, kind of guidelines, rules of thumb, is first of all, in a business light, that you have to rely on using industry standard building blocks because the business is so competitive that unless you differentiate with some very, very compelling difference, nobody will pay for that difference.

Second one is very important, very significant, and very applicable to today's time, is market segment share in this industry is won at the time of technological transitions. Between technological transition it is very, very difficult, very expensive, very bloody to move market segment share by even a point or two. At times of transitions, the early movers tend to drive the technology transition and reposition themselves.

And lastly, in contrast to conventional wisdom that says you price on the basis of cost, in this industry, you set the price first by what the market -- the sweet spot of the market demands, and then you work to adjust the cost to match in a profitable fashion the prices that you set in order to be competitive in the market.

The result of all of this is an enormous performance drive delivered at a lower and lower cost.

So this is kind of the story of the last 15, 20 years. This is the period of time that Jim Flanagan and I have met from time to time to compare notes. But in the meanwhile, probably on the radar screen, his and mine, and actually most people's radar screen, something else happened, and that is that the Internet came on. And the Internet is an overnight success. As you can see it on the chart, it came from nothing in the '94-95 period of time and just mushroomed and blew sky high.

And it reminds me that at one point Jack Nicholson was asked to explain how he became an overnight success as an actor, and as I remember his answer, it went along the lines that he worked like hell for 25 years in obscurity before becoming an overnight success.

And so it was with the Internet. The Internet started 25 years earlier. Nobody paid a whole lot of attention. The early parts of it were funded by the government for defense purposes. It was used by university researchers to communicate with each other and collaborate with each other.

Then the various standards, just like the PC style economics, industrywide standards, network-wide standards, were accepted that were common and allowed commoditization to happen.

And then when the overnight success came, what has happened is two simultaneous trends kind of collided with each other. One is what I described before, the mass production of very low-cost personal computers which were ubiquitous because they were deployed by the industry all over the place; likewise, access by the customers of the standard network which has been deployed over 25 years; and phenomenal availability of capital that at this point recognized the potential of the Internet.

And I should say there is another element to this which I think I can best represent by a quote from the Internet, one of the Internet jokes. An infinite number of Internet jokes are circulating around, which is the users latching onto this once ubiquitous capital, brought to them ubiquitous access through ubiquitous computers through ubiquitous networks. And this quote kind of tells you: Give a man a fish and you feed him for a day. Teach him to use the Net and he won't bother you for weeks. (Laughter.)

ANDREW S. GROVE: And all of this has been recognized by an enormous valuation rise associated with the so-called Internet companies. And the interesting story is comparing the universe of Internet companies and their valuation with the universe of semiconductor companies who have been around 30 or 40 years, and the Internet company valuations are almost matching in aggregation the semiconductor valuation.

So this is kind of a historical perspective of how we got to where we are, how we got to this overnight success.

Now I want to shift gears and talk a little bit about can we derive some rules like we derived some rules about how the PC industry operates, can we derive some rules about how the Internet industry operates. And the answer in advance is not yet. I am certainly not in a position to give you those rules.

But let me propose a few observations about the nature of this Internet economy which might be helpful in formulating your own rules that might guide you in your investment choices.

First of all, let me start with a statement that much as we talk about Internet companies today, in five years' time there won't be any Internet companies. All companies will be Internet companies or they will be dead.

And let me give you some supporting data for this on this chart. The vertical axis is logarithmic. Each division represents a tenfold increase in revenues, in dollar volume, and the horizontal chart is just a timescale.

The two lines on this chart show the processing of the US gross domestic products. The yellow line is the gross domestic product growing at a few percent a year rate. You can hardly see the growth. The blue line is the portion of the gross domestic product that's conducted by Internet-based e-commerce.

And what you notice, first of all, in 1999, reasonably close for us to be able to estimate it accurately, about one percent of the gross domestic product is estimated to take place on Internet commerce.

Now look at how much faster the blue line is growing and closing on the total gross domestic product to the point that in a few years' time, the estimates are that as much as 10 percent of the total economy will be conducted electronically. Business will be connected electronically on Internet connections.

Now, this is very esoteric and you say, okay, anybody can project all these esoteric numbers in the future. So let me give you a little bit of support from an internal Intel experience.

This chart shows our business, the percentage of our business, that we conduct on the Internet. In 1997, it was around zero. We did not conduct any of our own business with our customers on the Internet.

In 1998, past tense, over 20 percent of our business was conducted on the Internet. In the current year, our estimate is 40 percent, 42 percent of our business is going to be conducted on the Internet.

Intel is not unique in this. Companies everywhere are working double time to move their customer contacts, supply chain, management, both ends, both downstream and upstream, to an Internet-connected operation. And this, when you add these up, leads to that growing chart of the gross domestic product moving to the Internet that I showed in the previous chart.

Now, there are basically two imperatives in doing this. One is to make it worthwhile for the customer to do this. And particularly when it comes to individual customers. Most individuals' business is company-to-company commerce. But when it comes to individual customers, you have to give the customer an advantage, an experience on the Net that is so desirable that they'll come back. Because let me get ahead of myself here and tell you there is nothing unique in this business. We're using standard building blocks, we're using standard protocols, standard network, standard business ideas. It is how we implement it that is the difference between one company and another company.

And the possibility, kind of the utopian end point that you are going toward, is to give each customer, whether it is a business customer or an individual customer, the kind of experience as if that store was designed for that particular customer.

Jeff Bezos is the founder and chairman of that is probably a leader in formulating this one customer, one store concept or moving in that direction. Bezos said once, "We have millions of customers. There's no reason we shouldn't have millions of stores."

Now, we will talk in a few minutes of what it takes to build a world like that. It is easy to say, it is easy to paint the picture, and it requires a lot of technology and a lot of hard work to actually create.

But let's assume for a moment we are moving along this line and we are creating an environment like this. So what are some of those observations?

The first observation, it's a very important observation, that this world is a buyer's medium. The buyer dominates.

My formative experience as a manager was formed in the mid-'80s when the semiconductor industry was under assault by Japanese manufacturers. And Japanese manufacturers were better manufacturers than American manufacturers of semiconductors, cars, various other things like that because they could deliver to the customer much more expeditiously what they wanted than we could. It was a buyer-oriented manufacturing world.

And since that time, the American industry has restructured itself, semiconductors, cars, and everything else, by adopting some of the same techniques. But on the Internet, what is going to be possible is an unprecedented ability to match this person who is a customer with a supplier that can orient a store of one.

But in this process, the buyer is going to be infinitely well informed. And the buyer will have the ability on this network to get together with infinitely more well-informed buyers.

If we went to a hotel, a bunch of sellers, and closed ourselves into a hotel suite, closed the doors and started discussing how we can work together as sellers in a consolidated fashion, if there would be a knock on the door we would all be very frightened because we would think it is the Feds trying to bust up a sellers' cartel. But the Internet allows buyers to get together and exert an enormous amount of power. And the implication of that on the seller, of course, are tremendous.

The second one is, I apologize for this word, is the Internet is a fundamentally asynchronous medium. What that means is the buyer and seller do not have to be on the phone at the same time. You do not have to play tag on the Internet. Just like electronic message, electronic mail, you can send a message and the other person can deal with it at his or her time, at his convenience. At the same time, you don't have to have the slack capacity on both ends that is necessary for both buyer and seller to be there like we had at the registration desk this morning. We had to have a face-to-face process. In an asynchronous process you can handle this a great deal more efficiently both in terms of saving time for the buyer and saving time for the seller.

All of this is going to change commerce, is the third observation. It is going to make it a whole lot more efficient in terms of time utilization.

But the way it's going to change commerce, there's going to be obvious things. I kind of have this picture of this duck serenely floating around the water, and there's not much motion you see above the water level but boy, that duck is furiously paddling under the water because it has to do a whole bunch of things in order to propel itself forward.

In a similar fashion, in adopting commerce to this world, the obvious things are all the stuff that you read about in the newspapers and analyst's reports, re-intermediation, dis-intermediation, all this stuff. And under the water line you have wholesale reengineering of the supply line to allow this store of one to operate.

So the less obvious, much more expensive and much more momentous moves are building a pedestrian huge warehouse the size of this convention center in Nevada; Barnes & Noble buying the largest book distributor to allow its back offices and supply lines to be competitive with; companies like Web Van putting in place a startup company circulating vans that would deliver groceries and sundries on a scheduled basis to neighborhoods.

All of this is to build a single medium for end-to-end commerce.

Now, this is significant enough when it comes to physical goods. But just imagine what this network can do to bits, electronic bits, that do not need vans and do not need warehouses.

Networks have been used to deliver bits for a long, long time but the bits that they delivered were generally pedestrian things like network software, like antivirus protection, stuff like that. But as digital information -- as information altogether is becoming bits -- as newspapers become bits, as magazines become bits, as music becomes bits, as movies become bits -- all of that can be delivered with this growing and efficient network without the warehousing.

Now, all of this requires upgrading the infrastructure. And again, we have to use the duck here because the operating of the infrastructure, you read a lot of stuff about one aspect of it, is how you get on the Internet. And that's the stuff above the water level.

Below the water level a whole electronic network has to be rebuilt from reusing the old telephone network, which was not designed for this purpose at all, and replacing it on the fly to what is known as a packet switching, basically Internet-oriented network, increasing the capacity of this network, moving to a broadband status, and put in store houses like the equivalent of's warehouse but for electronic bits that are called "bit factories," and we'll talk about that a little bit later.

And all of that is driven by the nature of this customer-driven, buyer-driven cartel that wants easy and fast access to information, where the manufacturer has to have deep customer knowledge, has to have the whole buying history automatically so they can customize the site for that user-- obviously, since we are dealing with a lot of transactions, there's a lot of opportunity for fraud and misuse of funds. It has to be secure, it has to be reliable, and it has to be 24 hours, 7 days a week, 52 weeks a year reliability in terms of up time and customer support.

All of this has profound requirements on the electronic systems that our industry produces. They have to provide low cost pervasive access. And these bit factories, as I said, have to be 24 by 365 days reliable operations.

So if I sum up some of the observations of the incipient rules for this Internet economy, it is going to be a buyers' medium, it is going to be asynchronous in nature, it is going to cause a reengineering of commerce in a big way. It is going to be particularly effective in distributing bits, and it is going to bring wholesale changes to the infrastructure.

And all of that is happening on Internet time, which is to say kind of like dog years are equivalent to seven man years; in this case, Internet years are equivalent to seven ordinary commerce years. That's about the rate with which it has to happen.

Now, I said earlier that there are difficulties with this. Conceptually, you can paint this on a picture or electronic slides or whatever, but to make this a reality, you have to navigate your way through real pitfalls.

Wal-Mart is one of the less-known examples, outstanding examples, of an electronically driven ordinary physical goods company. But their infrastructure was built over more than two decades. And it was built and designed and redesigned and debugged over a long period of time. We are trying to do this a whole lot faster.

This reengineering is happening as people are jockeying for a competitive advantage. So it's happening, driven very, very rapidly. And there's an interesting aside here. It is happening awash with money. These companies have money thrown at them. And that's good, but it's also dangerous.

It's good because it allows them to do big things, and it is dangerous because companies who have too much money don't have the market discipline of learning to operate with the money that they bring in part of their business and streamlining and curtailing their plans and operations and developing a pattern or focusing that that discipline brings.

And I have to rely on my own background here. In the decade of the '70s when the semiconductor industry was as turbulent, as frothy as the Internet economy is today, practically every developed and undeveloped company decided we have to have our national semiconductor company like we have to have our national airlines, and any country that doesn't do it is not going to be a significant force in the future. And what happened is we competed against a bunch of companies that were funded by the infinitely deep pockets of sovereign nations, and every single one of those vanished over a period of time.

So too much money is a dangerous thing to deal with. And the point I'm saying is exponential -- huge opportunities bring exponentially huge opportunities for failure as well.

So that's what I want to say about the rules of the Internet economy such as they are. And I'm sure if I came back here a year from now, we'd probably change these, because things change and we learn about what works and what doesn't work as time goes on.

In the next section, I would like to comment about Intel's role. Intel's role is, of course, very significant to me because we are running an enterprise that has to live in this world, but also interesting because it is fairly pervasive in this Internet economy. So our perspective can kind of give you a different view, a triangulated view, on the Internet economy.

Our charter is to be a building block supplier to this Internet economy and to be as large a building block supplier as we can manage to do.

Now, what does the Internet economy mean? I'm going to first give you a glimpse, and then I will explain this in a minute. But the Internet infrastructure kind of looks like this: the computers through which you get into this big cloud, Internet service providers, company Web sites and the like. You don't know where they are, I don't know where they are, I don't know where the Intel version of those, where they are. They are someplace in this big cloud, but they do the work of the Internet.

So before we go back to that cloud, I want to show you what the access devices look like. And the access devices are the stuff that you use yourself to get on the Internet, to get to your online trading, to get to your online sports.

And luckily for us, because it is based on personal computers, that is the most prevalent access device on the Internet, and consequently, all the people who develop software on the Internet know that an overwhelming percent of the access devices are based on Intel chips or Intel's competitor chips, similar to that, and so all of that software, first and foremost, is written for Intel Architecture based machines.

And we realize there are people who are more ambitious in terms of the visual user experience they want, and we realize people have needs. Some people have needs for the economies of mass production and are more value oriented buyers and we have segmented lines to serve both of those, and it is out of that segmentation that the latest phenomenon, which is the PCs provided by other suppliers in the cloud, the so-called subsidized-PC movement, has emerged.

But let me demonstrate what Intel-based client access to the Internet looks like.

Let me welcome first Frank Martinez from Intel who is operating on what?

FRANK MARTINEZ: Good morning, Andy. What I have here is a Celeron based PC -- I'm sorry. An Intel-based Celeron® processor platform.

ANDREW S. GROVE: Thank you. (Laughter.)

FRANK MARTINEZ: The great thing about it is it's a great second PC for the home that has the price and the performance to allow me to access timely information from the Internet. I can manage my finances on the site. As you can see here. And what I'm going to do is go back to the CNBC Web site. This is my financial center. And if I take a look at the investor toolkit, you can see that I can have access to both domestic and international news. And in addition, there's some interesting utilities such as this retirement planner here.

ANDREW S. GROVE: Well, I'll let you work on your retirement. You have a few years to work on that. But what I really want to ask, this is the Intel Celeron based computer here. Where is your Internet connection.

FRANK MARTINEZ: The great thing is I'm using Intel's AnyPoint Home Network. This allows me to share a single connection to the Internet with all the computers in the house.

ANDREW S. GROVE: Let's call attention to this little gray device, which comes in any color as long as it's gray until Apple puts its hands on it, allows you to connect this computer to another computer in the house. And if that computer is connected to the Internet, I can share that connection for the same price and at the same time, even?

FRANK MARTINEZ: Correct. And the great thing is it uses existing telephone wires in the house, so you don't need to install any new wires. And in addition to that, you can share files, you can play games between the computers, and you don't have to be on the same site as other people in the household.

So I'm interested in this, and Lisa over there, in fact, is surfing and going to the Web on a different location.

And while you go take a look at what she's doing, I'll work on my retirement plan here.

ANDREW S. GROVE: Well, I hope Lisa is doing something more interesting than your retirement plan. Goodbye.

Hi Lisa.


ANDREW S. GROVE: And here is the rest of the connection, the AnyPoint™ Home Network device.

LISA WARREN: Exactly, I'm sharing the same connection as Frank, but I'm accessing it with the more powerful Intel Pentium III-based processor system.

ANDREW S. GROVE: Very good.

LISA WARREN: And I'm about to visit the Van Gogh exhibit that was recently here in L.A. on a new site called

ANDREW S. GROVE: Did you get to the Van Gogh exhibit?

LISA WARREN: No, I wasn't able to get there. It was sold out even though it was open 24 hours a day. But I can access it anytime on the Internet.

The thing I'd like to show you is something under development in our labs back in Santa Clara. It is a reconstruction of the exhibit as it appeared at the National Gallery of Art. We can move through the exhibit using the keyboard keys.

ANDREW S. GROVE: This is the layout of the original Van Gogh exhibit, and you can find your way through it through this navigational scheme.

LISA WARREN: Exactly. Or if we know where a specific painting is, we can use the 3D map to go directly to that room.

I'd like you to see one of my favorite paintings, Yellow House.

While we're here, we can learn a little bit more about the art work by listening to an audio clip.

(Audio playing): "The Yellow House is certainly one of Van Gogh's best loved paintings. The bright blue sky and vividly yellow building..."

LISA WARREN: Or one of the great things about exploring art with the Internet and a PC is we can zoom in to a very high level of detail and explore some of the people sitting at the cafe. But don't try this in the real museum or you might get thrown out.

ANDREW S. GROVE: Can you switch to another painting and show something even more rich? I'm not much of an art connoisseur, so paint brush techniques don't appeal to me.

LISA WARREN: Let me show you something that might be a little more interesting. Intel has been working very closely with the Van Gogh Museum to try to create more innovative ways to teach people about Van Gogh's technique and his life, so we've created a 3D reconstruction of this painting, The Bedroom, to help people learn more about Van Gogh's art.

ANDREW S. GROVE: We can go into the painting and turn around.

LISA WARREN: You're inside the bedroom right now.

And one thing I'd like to point out to you is many people believe that this painting was painted with a skewed perspective, but actually if we look at a photo of the yellow house where Van Gogh lived at the time he painted this, we can see that it's correct and the roof is slanted.

Something that's really neat, using the power of the Pentium III, we can move around inside the room and even look at the paintings that Vincent had hung over his bed. (Applause.)

ANDREW S. GROVE: Now, suppose I wanted to have a souvenir from the Van Gogh exhibit. Would you help me out?

LISA WARREN: I could. One thing I like to do when I'm done cruising around the galleries is visit the museum store where we can buy the same souvenirs as were at the actual exhibit. Here's a key ring with The Bedroom on it, and here's one with the Yellow House. So if you'll just give me your credit card, I'd be happy to -- (Laughter.)

LISA WARREN: … process this transaction and it will be waiting for you when you get home.

ANDREW S. GROVE: What kind of security software do you use?

LISA WARREN: You can trust me, really. (Laughter.)

ANDREW S. GROVE: I'll think about it. Let me see what Frank is up to in the meanwhile.

Did you figure out your retirement plan?

FRANK MARTINEZ: Actually, I think you caught me here. (Laughter.)

FRANK MARTINEZ: I think my retirement plan is going to be with the Lakers. I'm trying to figure out the jump shot for Kobe Bryant and help him win the game tonight.

ANDREW S. GROVE: Well, keep practicing.

Thank you. Thank you, both of you. (Applause.)

ANDREW S. GROVE: Now, this is the front end of what you see and experience.

Let me, at the risk of possibly telling you more than you want to know, and I keep thinking of the visit to the sausage factory. (Laughter.)

ANDREW S. GROVE: Let me take you inside the sausage factory to that famous cloud of what has happened behind these movements and what happens every time you do anything on the Internet.

First of all, what you see is either Frank or Lisa sends a bunch of bits to the Internet service provider. The Internet service provider has a bunch of computers in its location, and those bits arrive and they get into something called an access concentrator-- I'm going to use a bunch of words that I don't really expect you to remember even though there's going to be a quiz at the end of the demonstration.

Fortunately, from our standpoint, being a building-block supplier, we supply a bunch of network blocks that go into this networking deal, and the bits go into a server and from there they go into something arcane called Internet naming server, domain naming server, where the previous location, which is the location that Lisa was looking for,, is translated into an Internet location that the Internet can understand.

All of this is done on a bunch of Intel microprocessor-based servers. Or I shouldn't say "all" of it. I wish all of it was done on Intel-based servers. But much of it is done on Intel microprocessor-based servers.

And once it is done, we leave the Internet service provider, we go through the telephone backbones and all the stuff that these big companies with their billion dollar mergers and you go from there, you can tell the Web site that offers up because you can see perhaps in that little vignette the Web page, which is what you see on your computer.

On that Web site, the bits arrive. We want, but what we really want to do is shop. So we go through a bunch of servers with arcane names like traffic management servers that decide where to send the bits next. Fortunately, we have some Intel architecture products that help the Web site build these things. From there they go to the Web server that serve up the Van Gogh pages or serve up the Laker pages. Each of these are bits located in storage in somebody's bit warehouse that's located at this Web server location.

But then when we want to go into a transaction like buying the key chain the bits leave and go into something called an application server and conduct the transaction again asynchronously, because all of this can take place at a time not precisely the same time that Lisa clicked the button. And therefore, a million transactions can be processed at the same time because they don't have to shake hands with that application server at a precise time.

And once that happens, the bits start coming back out of there, move through the Internet backbone, move through the Internet service provider, and come back and serve up the Web page that we looked at before.

This is the gist of the Internet. Computers passing data to each other, managing them, calling up data, calling up customer records, conducting transactions. And this is where the sausage factory does things in a way us customers don't care about it, don't know about it. But some of those of us -- as I said, this is the Intel perspective -- that make a living helping to build this sausage factory care very much about it because this is our market.

And the way it is predominantly our market, we develop servers and we use chips for servers, and we use the same economic rules that we have perfected and honed in supplying microprocessors for personal computers.

Our current product line is the high volume servers, 32-bit in terms of how many bits they can operate with at the same time, and they have the same virtues as PC-based microprocessors have, high performance at a very low cost, and offering the designers of the locations of the Internet service providers or the Web site a great deal of choice in what kind of software they can use to build that Web site.

A very major undertaking for Intel in the last several years has been extending our architecture so it can operate with 64-bit data, which is twice as long data. And in the course of the next year we will be starting to go into that.

Now, the significance of that in Internet applications comes when every customer has his/her own data, own purchase records, own purchase preferences, own credit information, all of these are located in a database. This database becomes very, very large.

If you can go to 64 bits in size, you can store very large data bits in the server memory and, therefore, find everything without having to go out to a disk and bring memory in there, and the response of the site is going to be very, very much faster.

So this 64-bit transition comes exactly at the right time as databases are growing so big.

As databases are getting so big, it is very difficult to find your way around them. And as we visualized the Van Gogh exhibit, visualizing buyer patterns, customer patterns, supply information is one way to cut through and digest a growing amount of information that's stored in these servers.

And, of course, security and operating system flexibility are all necessary things as well.

In the last decade we have managed to build up a fairly decent-sized networking device business. The Internet is about networking. Every time two computers are connected on both ends, just like illustrated in this AnyPoint device in its simplest, home-oriented application, two pieces of connected gear are necessary. And in the cloud, in the sausage factory, likewise, a lot of networking devices are used.

And our claim to fame here is addressing the needs of particularly the medium and small companies, each of whom, just like the swing-set manufacturer, have to go on the Internet, and they can't afford to have the real high-end solution to that. And using our semiconductor expertise, try to allow these devices to be flexible, so as a customer's requirements grow and change, the networking devices can be adapted to that.

More recently we announced the new initiative going into basically allowing a customer to buy the services provided by these bit factories on a service basis rather than having to buy their own. So our friend, for instance, the swing manufacturer, may decide that he doesn't want to hire IT professionals and he doesn't want to use his capital to build his own Web server and Web presence. He can rent this capability or outsource this capability from a bit factory that we will have in his neighborhood.

We will attempt to grow this globally because an enormous amount of growth on the Internet will come from countries outside the U.S. And as the gross domestic product is moving to the Internet, a lot of small and medium companies will have to get into the act. And we think they will want to use this capability as a service rather than put this on themselves.

So the sum total of what we are up to is we are expecting that the Internet economy is going to be horizontalized in the same fashion as the PC infrastructure was horizontalized a decade ago. There is going to be horizontal aspects of the infrastructure -- where your presence right now is Intel-based clients like the two you have seen here before. We have acquired a new device called StrongArm™ which has allowed us to participate in the mobile handheld Internet clients. We have Intel microprocessor-based servers, Intel networking devices. This is our infrastructure business unit.

In the high-volume data application services we described these bit factory undertakings. And finally, when we deal with content like, it's very important for us that content be developed taking advantage of the infrastructure that we developed; that it should be tuned on that infrastructure. And we use a fairly large investment activity to participate in, as a minority investor, in these companies that build up the content part of the Internet economy and provide them with technical support, guidance, and financial wherewithal to help them to do that.

The whole point of it is driven, the whole, all of these activities are driven by the desire and the vision that networked bits anywhere on the Internet should be shaped and shipped primarily by Intel silicon and services.

And that completes my story, and thank you very much for bearing with me. (Applause.)

JAMES FLANAGAN: Thank you, Andy. That was a heck of a speech.

ANDREW S. GROVE: Thank you.

JAMES FLANAGAN: Charles Piller of the Times business staff, and myself are going to ask Dr. Grove a few questions.

CHARLES PILLER: Even as Intel jumps full-fledged into the Internet economy as a supplier and as a builder of that economy, you've expressed considerable skepticism about the extremely high valuations of many Internet companies that look pretty distant from making a profit. And I'm wondering what you think the implications are for investors and for the high-tech economy if and, many would say, when the high-tech Internet stock bubble bursts.

ANDREW S. GROVE: Well, I said earlier in the talk, this is a challenging time to be a participant and a challenging time to be an investor also, because what you have in front of you is the wholesale reengineering of the U.S. and, for that matter, the global gross domestic product delivery.

Remember my PC rule that market share is gained and lost at times of technology breaks. This is the mother of all technology breaks. So it is with this wave of reengineering sweeps through the economy, people will surf it and get ahead of other people who become irrelevant and fall behind.

So the opportunity to try to handicap that and bet on the horse that's going to get ahead is very, very tempting.

At the same time, the ability to discriminate between one choice and another choice in this very noisy investment environment is becoming very difficult.

So in the aggregate, I think there's going to be a lot of losers and some winners. There's no question that if you took a snapshot of the economy today and a snapshot of the economy five years from now, the lineup of participants are going to be very different. But just the same way as previous technology we have seen and the PC industry swept in and the semiconductor industry swept in, if I mention the starters, a lot of people would say "Who?" A lot of the high fliers today are going to be "Who?"

And it is certain that when Frank does his retirement planning he has to be extremely careful of how much risk he can tolerate, because his visibility, even as he faces enormous opportunities and retires to go to work for the Lakers, that is there, but the risk for placing the bet on the secondary horses is also very high.

CHARLES PILLER: Words of caution.

ANDREW S. GROVE: Absolutely. Words of caution and words of opportunity combined at the same time.

CHARLES PILLER: I'm going to have to check with my stockbroker after that. Thanks.

JAMES FLANAGAN: Right now we are on the cutting edge of this mother of all waves of technology. And I think the United States economy has more usage of the Internet than any other economy. You had global distribution up there. We have a world market that is going to get into this.

How fast do you see that world market developing? You and I go back to the days of fierce international competition and people predicting that Intel and everybody else was going to be buried here. Well, now we're in a new wave. Just because the United States is ahead in Internet usage, do American companies, do companies like yours and the ones we read about, have an edge going into the international market?

ANDREW S. GROVE: This is a pretty complicated question. Both of you are asking pretty complicated questions so far. The competition in the direction that you're going with is more between global -- between national economies versus other national economies, it's almost like the economic health of a regional economy can be measured by degree of reliance on connected technology, Internet technology.

As I pointed out in the early part of the talk, one of the reasons clearly emerging that the U.S. economy has been so rugged is because it has such an enormously persistent productivity growth and allowed to grow with low inflation, and it's increasingly evident that's tied to high technology.

When it comes to individual companies, all of us are global players. All of the substantial players are global players. More than half of Intel's income comes from abroad. So when you invest in a large multi-national, you are getting a market average of this phenomenon, a basket of this phenomenon, that cuts across some Japan, some Asia, some China and some Germany and a lot of U.S.

But when you look at an economy, like the U.S. economy, it clearly has an advantage. The American company doesn't necessarily have that much of an advantage, but the American economy does.

CHARLES PILLER: I was wondering if you could talk about one of the big concerns that consumers have that despite the incredible growth of e-commerce, one of the stumbling blocks has been consistent concerns about personal privacy on the World Wide Web. Even Intel has experienced the sting of that issue with the controversy over your Pentium III chips.

People want to come on the Web, they want to buy things on the Web, but they hesitate because of that. What is it that Internet e-commerce companies must do to gain the degree of confidence of consumers that they must have in order to really bring the economy forward?

ANDREW S. GROVE: Well, this is two questions. One is the concern about privacy, and the second one, what we should do about reassuring buyers.

The answer to the first question is of the ilk of the sausage factory comment again. Long before the Internet became the medium for commercial transactions, we have lost, each of us as consumers, lost a lot of our privacy.

I get in my mailbox catalogues and mailers from companies that I never heard of, never thought of, never was interested in. People are buying and selling databases of credit card records and phone records and stuff like that. There's a whole economy going on which has nothing to do with the Internet.

And in actual fact, it kind of, as a first rule, everything that is happening on the Internet is just an adaptation of what has happened in non-Internet or pre-Internet life in some fashion, whether it's pornography, it wasn't invented on the Internet, crime wasn't invented on the Internet, fraud wasn't invented on the Internet. These are all, just like the GDP, are being moved over. All of these undesirable aspects of human life that have existed are moving over to the Internet.

So is the loss of privacy. The loss of privacy precedes the Internet, but it also is present on the Internet just like the other things.

As to what can be done about it, the best thing we can do about it is shine a light on it and expect certain commercial standards from people who collect that information. They have to collect the information to do all the stuff that they're talking about. But we don't have to get into the business of reselling that information for other users and other purposes.

So I'm a great believer in the emerging standards, industry standards, that says when you do business with me, I promise to honor the information that you give me and use it only for the purpose that you gave me.

And increasingly, these kind of standards will become a necessary element of doing business on the Internet, because you will simply avoid the sites that don't make that commitment.

JAMES FLANAGAN: Andy Grove, I really want to thank you and the Times wants to thank you and I'm sure everybody wants to thank you for a great kickoff to this conference and on a subject that is just right today, cutting edge.

Thank you.

ANDREW S. GROVE: Thank you. (Applause.)

* Other names and brands may be claimed as the property of others.