February 23, 1998

Read the transcript of the Craig Barrett interview or discuss Intel's Enterprise Push in Shop Talk

Andy Grove Interview

INFORMATIONWEEK: We've seen you explain Intel's Enterprise processor strategy. We also see Intel making very big advancements to stimulate the use of new types of applications. I'm thinking of the work in visual computing, Internet-related startups, E-commerce, videoconferencing. Can you give us examples of how this strategy will change the way enterprises use technology and run their businesses during the next two to three years?

GROVE: I've gotten used to giving all kinds of caveats before I talk. I feel compelled to start answering you with disqualifying ourselves as f orecasters of future trends. It's interesting, if you bear with me for just a minute. It's an interesting exercise because how I've performed in the past in answering questions like this may give you some idea of how accurate further prognostications are going to be.

Where we were right in answering this question in 1989, 1990, '91, we came up with this simple concept that PCs could predominantly [become] the communications device and [we] started investing our technology in the direction that made that possible. At the next level of detail what that meant to us was videoconferencing. We started investing on that in a pretty heavy way, fully anticipating that if someone would ask me what the situation would look like in 1998, I would have said that videoconferencing -- or whatever phrase I would have used to describe it -- is going to be as ubiquitous as word processing.

What happened if I look back almost 10 years, is the PC became a communication device, videoconferencing remains a niche ap plication, and Intel on an absolute scale within the confines of videoconferencing has done all right. Videoconferencing did not become anything like word processing, but the Internet, which absolutely would not have made the top 10 of anything that we would have looked at that time, has. Do I make sense with this?

INFORMATIONWEEK: Yes, in the sense that it's hard to predict.

GROVE: Well, you can predict the general trends. I think we were right in that the PC did become a communications device. Reduce it to the next level of detail of a specific application and we were wrong -- in two ways. The things that we thought would happen, didn't. And the things that we didn't even think about, happened in an incredible way. So, with that as an abject lesson, I'm answering your question as follows:

Yes, the PC will continue to be and evolve [as] the communications device. It is a window to other computers. That part we were right on and we continue to believe in. The impact of c ommunications is going to become more and more ubiquitous, and as it becomes more and more ubiquitous, it is going to be the computer-to-computer transactions of communications [that] are going to be the default of business communications. If 15 years ago the default was "give me a call," maybe seven or eight years ago it would have been, "send me a fax." I think going forward it's going to toggle to computer-to-computer communications. From the most trivial, which is E-mail, to the more sophisticated, which could impact videoconferencing 10 years later than I thought.

Because the communication elements can be made more instantly digestible by visual means, increasingly visualization is going to come in. By the way, the World Wide Web is a perfect example of this. The difference between success and the ubiquity of the Web and the previous niche use of the Internet is visual. Having visualization is going to be a continuing driving force and the big payoff in that is that it is increasing the tempo of business, [the] rate of exchange of information, rate of immediacy. It is going to do the same as electronic funds transfer did to finance: Electronic information and transaction transfer is going to make [just in time] business absolutely possible.

It happens to finance. It happens internally to the computer industry. It is going to happen in a broader [set of industries.] It's almost like a set of concentric waves [after] you drop a stone in the water. The waves that are the stone in the water that you put in is computer-to- computer communications and the waves that move outward are turning business in to a more immediate higher rate, higher- speed process.

One of the things that I feel a little frustrated about is the people [who] look for productivity improvement associated with information technology, and the way they measure productivity improvement is completely oblivious to the impact of timeliness. The productivity impact of electronic funds transfer is zero. The impact of electronic funds transfer to the world economy is being demonstrated in real time in front of our eyes. So, that argument to me exposes the enormous innate weakness of an economic measurement of theories. It's a commentary on economics more than in IT.

INFORMATIONWEEK: With regard to the point about changing or advancing productivity, how are some of these nontraditional investments that you're making going to help Intel? How are they going to change their companies culturally and organizationally moving forward?

GROVE: I'm not sure what you mean by nontraditional investments.

INFORMATIONWEEK: Such as networking, [software.]

GROVE: It would fall under the umbrella of what I've said before. Just about everything we do at Intel has to do with computer-to-computer communications which are increasingly visual. Everything from data mining, to videoconferencing, to CDs, to networking. If you look at the hundred-odd investments we made in different companies (a min ority of them that have to do with manufacturing) but other than that just about everything fits into this umbrella. So, we passionately believe in that. The payoff is going to be people doing the work of business in front of a rich, visual computer monitor communicating in an increasingly visual way.

Visual, don't think videoconferencing only. Our mistake was concentrating on the "talking head." I showed a [data- mining] demonstration at the Intel Development Forum. It is an amazing technology looking at enormous bunches of data and downloading into your machine and letting this data mining sort it through in various different ways. Data becomes a pattern, and pattern becomes information. Before there was no way you could have dug out the subtle changes.

If I know specific data or patterns faster, I can communicate it to those people who need to know it, [and] I have a competitive advantage. Other people will follow me. I have to move along that same line. The net result is I pursue my busine ss more knowledgeably and faster, which is really all there is in competition: knowledge and speed. Notice I did not say productivity, because it is a third order of fact. Having the right transaction and taking a half-hour as compared to having 20 transactions at five minutes each. They're wrong. You're not even talking on the same plane.

INFORMATIONWEEK: Can we talk a little about the segmentation of the industry, different chips for each segment?

GROVE: That is a subject I can hold forth for an hour. Can you get a little more narrow in your question?

INFORMATIONWEEK: At the high end, do you see a point at which Intel-based systems will be able to scale to run the biggest enterprise applications?

GROVE: Yes.

INFORMATIONWEEK: When?

GROVE: When will we all be happy? [Laughs.] The pursuit of happiness and the pursuit of the highest reaches of enterprises is a never-ending and dynamic race. We are moving up, and databas es are getting bigger, and the need for speed is relentless, so the application is moving up. In a way, we [in the microprocessor industry] continue to chase [the development of applications] and maybe we'll never reach them. But I think we're moving upward faster than applications are moving upwards. So we're gaining on them.

None of you were at the Intel Developers Forum, were you?

Continue...

Return to "Intel's Enterprise Push"

Send Us Your Feedback

Top of the Page

---