Hello Frank, Interview with Andy Grove ( from recent edition of The New Yorker magazine]
ONLY THE FAST SURVIVE
An E-mail exchange with Andrew S. Grove, the chairman and C.E.O. of Intel. KEN AULETTA
I NTEL, the computer-chip company that Andy Grove co-founded in 1968, manufactures the chips that are the brains of more than eighty per cent of all P.C.s. During the past decade, as Intel managed to double the speed of its chips every eighteen months, the company's revenues rose an average of thirty per cent annually -- to twenty-one billion dollars last year.
Last month, Intel stunned the high-tech community with its announcement that the memory of Intel chips would soon double at no extra cost. The company's dominance is such that the Federal Trade Commission is now looking into its hegemony. (A previous investigation was closed.) Recently, I conducted an E-mail interview with Grove about what is next for Intel and communications.
Q. Last year, you wrote a book, "Only the Paranoid Survive!," advancing the notion that a company must always be fearful of competitors. Is the future business model for communications companies predicated on more warfare or more competition?
A. What's the difference? Competition is warfare. Mostly, it is played
by prescribed rules-there is sort of a Geneva Convention for
competition-but it's tough and often brutal. So, yes, the future of
business being based on technology (which changes rapidly and gives new opportunity for renewed attacks) and being global in form (those attacks can come from any corner of the world) is going to be warlike.
Q. On a panel at the World Economic Forum in Switzerland last February, a telephone-company executive asked you to explain Intel's business plan for the Internet. With barely concealed contempt, you asked him, "Did Columbus have a business plan to reach the New World?" So you're just guessing at the future?
A. Well, "guessing at it" may be an overstatement. Or maybe it isn't. On any given day, those of us who work in the high tech industry make decisions that are basically educated guesses about the future of technology and-equally important-about future market trends. Today, at Intel, we decide on building factories (or not) that will come on stream in 2000, or 2001. They will build products that our engineers are designing to day. We are sizing those factories based on little more
than guesswork that involves such things as market demand-in 2001! for these products. It seems crazy or irresponsible. But is there really
another way? Those factories take three to four years to be built and
get up and running. They will, by necessity, build new products. There
is only one thing known about them: if we don't build those factories,
those new products will not be built. It's like that for Internet and
telecom infrastructure, too. They deal with unknown and unknowable
market demand for services. I just wish some of those industries were a bit more aggressive and optimistic about building their infrastructure,
too.
Q. How is the Internet a boon, and a potential menace, to Intel's
business?
A. The Internet means universal, global connectivity. It means that in a
few years there will be a billion interconnected computers operating in
the world. Such an incredible network will turn into a medium for
everything-entertainment, commerce, medicine. This picture means there will be a huge market for microprocessors. They will be needed in the billion computers, and in the servers, switches, and routers that make up the network. That's what we make, so this is good news. The fact that this network is universal means that all kinds of microprocessors can be used for all these purposes. That means more competition. That's the threat.
Q. You often speak of what you call the "strategic inflection points,"
the unforeseen threats companies must anticipate. What do you believe the next strategic inflection point is for Intel and the computer
industry?
A. It's the Internet. It's the staggering need for reliability and
ruggedness and security that such a network requires be fore we can
entrust our commerce, our work, our medical records to it.
Q. Larry Ellison, of Oracle, and others believe that consumers will want
cheaper computers-in the five-hundred-dollar price range-with the
ability to pluck software off various networks; including the Internet.
Do you believe that price will prompt consumers to embrace this carpool model? Or do you believe consumers want to drive their own souped-up car and will pay extra to own and control their own cool stuff?
A. Limited computers will be adopted broadly only if the software
community stops developing new and attractive products. Until that
happens, the industry will be driven by a demand for the new. Having
said that, I think it is worthwhile to experiment with what can be done
by simpler, network-connected devices. It will be years before they
become popular, if they ever do, and using those years to develop
expertise in them is worth while.
Q. If Ellison is right in his guess, what happens to Intel and its
expensive, souped-up chips or microprocessors"
A. They move into the servers. Computational needs are driven by the
application. Where the data is stored doesn't really impact the need for
powerful microprocessors-somewhere. If computation becomes more
server-centric, servers will need to be beefed up. We'll hope to be
supplying those beefed-up chips.
Q. Such apostles of the Internet as George Gilder argue that companies like Intel and Microsoft represent the past, and that the emerging battle in information technology is between those who wish to control the desktop and those who wish to use the Internet as a mass-transit system to liberate citizens from expensive chips and software. Do you feel like Marie Antoinette?
A. Well, I don't really know how she felt. I feel energized and
motivated by the enormous change that we are engaged in. I scour the papers (and my computer) for the events of the day, every day, as we progress through these changes. As for George, well, he is George. In 1984, he wrote a review of a book of mine. He predicted that Intel would be wiped up by Japanese microprocessors. We weren't. Then he wrote a book, called "Microcosm." In it he predicted that microprocessors will all be customized by systems designers. That didn't happen. Now he is merrily predicting a bunch of new things, unfettered by any accountability for his predictions. I envy you writers such a fairyland life. |
