CNET también está disponible en español.

Ir a español

Don't show this again

Desktops

Intel's Barrett knows PCs inside and out

The PC market saved Intel from becoming a technology industry footnote in the mid-1980s, and it hasn't looked back.

Intel Chairman Craig Barrett has seen a lot of PCs pass by his desk in the last 25 years.

Barrett has used a PC from just about every major PC supplier at one point or another during his career at Intel. The company's role in advancing the PC market came almost out of desperation, as it struggled in the face of competition from Japanese memory makers.

Intel decided to invest in PC processors, and that has worked out fairly well for the Santa Clara, Calif.-based company, which went on to become the dominant PC and server chip supplier. Even in the face of renewed competition from Advanced Micro Devices, Intel still commands 73 percent of the market.

Barrett, who steered Intel throughout the late 1990s and early part of this decade, spoke last week with CNET News.com about the development of the PC market and the future of the PC. Here are some excerpts from that conversation. To hear the full audio interview, click the play button in the box below.

Q: What was the first PC you used, in your home and in your office?
Barrett: My first PC was an Osborne, which I won in a contest run by (former Compaq executive) Ben Rosen at one of his conferences once a year for guessing what the compound growth rate of the high-tech industry would be during the last couple of years, next year or something. It was the old sewing box version.

Listen up

Intel's chairman
on the PC's 25th

Getting older and still getting better.

Download mp3 (4.4mb)

When did you first realize that the PC market would turn out to be as large as it has?
Barrett: I'm not sure (that) when we started the industry--and I'm going back to the mid-'80s now--that anybody had an idea that it would grow to be its current size. You heard forecasts from various folks that the objective was to put a PC in every household. You heard equally adamant comments that that would never happen.

But 10 years ago, we started making projections, not only about Moore's Law and the semiconductor industry, but also started making projections about the number of PCs and the number of Internet users and things like that. And most of those projections over the last 10 years have been reasonably accurate--basically saying that you'd get to sell a few hundred million PCs a year; you'd get a billion or more people online attached to the Internet; you'd do trillions of dollars of e-commerce over the Internet. All of those projections have pretty much been accurate.

What are the most important factors that allowed the market to take off the way it did?
Barrett: I think it's been the flexibility of the machine, the fact that it's not a dedicated, hardwired, hard-programmed machine. It started off as the spreadsheet machine and a word processor and went from that into different forms of media, and then it went from that into current substantiations with music and video and the Internet. So the fact that the machine has this ability to transform itself and evolve with time, as opposed to just being a single-function machine, I think that's been the primary value of the PC.

"Maybe the dang thing will be flexible and you'll roll it up; it won't be a solid metal-encased entity."

It's simple to think of it from the standpoint that when PCs started, it was basically a word processor and spreadsheet and everybody was looking for the killer app to supersede those two applications. The killer app happened to be 101 or 1,001 different applications, not a specific application. And the fact that it was 1,001 was made possible by the reprogrammable nature of the machine, its adaptability.

Around this time in the mid-'80s, Intel obviously made a huge bet on this market, switching production away from memory chips to processors. Tell us about some of the things that convinced you that was going to be the right decision for the company.
Barrett: There is a lot of folklore about that decision. In reality, if you look at the time in the mid-1980s when we really made that final decision, it was one of those decisions that had already been made for you. Our market share in the DRAM market--the memory market--had declined substantially; we were a single-digit percent player. I think the company was realistic in where its future was going to be: We could either continue to play in a business where we were not doing well, or we could cast our lot with a new and potentially very exciting business which was just starting.

If you look at the history of Intel, which has been (an) innovator in bringing new technology to market, I think it made sense for us to pursue the new technology and to see what opportunities it would bring rather than aim to slog it out with a whole series of very competent Japanese DRAM manufacturers.

One thing that we've been looking back on is what the PC industry might have looked like had IBM kept tighter control of the platform, had it not allowed Microsoft to license the operating system.
Barrett: Over the years, the beauty of the PC market has been its relatively horizontal nature, which means you had different folks doing the various components--different folks doing the platforms, different people doing the software, operating system, applications, etc. That horizontal nature just interjected competition into the system at all levels, and there is nothing like competition to move the technology forward.

If the PC had stayed a vertical, it would have moved much, much slower in terms of bringing functionality and capability to the end user. That goes for whether you looked at Apple, which kind of stayed a vertical, controlling--to a degree--the hardware and the OS and the apps. But Apple had to move rapidly in response to what was going (on) in the PC market to maintain a competitive spirit. The horizontal nature of the PC market is what made it move as fast as it did.

Twenty-five years from now, what do you think the PC will look like?
Barrett: The last 25 years, its trend has been to basically follow Moore's Law and bring more and more functionality with lower cost and better form factor and better user-interface, and all of those things will certainly continue. Ten years ago, when we were thinking about what would the ideal laptop be today, nobody dared to think about something effectively not much bigger than a bunch of 8-1/2 x 11 sheets of paper stacked together, about a quarter of an inch thick, and having a full-color screen, weighing about a pound, all-day battery life, great graphics, etc.

You know that was a dream 10 years ago (and) that's pretty much happened today. I can dream (that) 10 years from now we'll continue on those same lines. Maybe the dang thing will be flexible and you'll roll it up; it won't be a solid metal-encased entity. I can imagine that the form factors would continue to increase; we'll see real PCs that kind of look like and feel like BlackBerrys--the little handheld intelligent phones that we have today but with real PC Internet-interface quality. I can see all of that happening in 10 years. I am not even going to speculate what happens in 25; that's way too far.