Who's afraid of the big, bad NC?


6 min read
CNET News.com Newsmakers
September 9, 1996, Andy Grove
Who's afraid of the big, bad NC?
By Margie Wylie, Richard Hart, and Brooke Crothers
Staff Writers, CNET NEWS.COM

Andy Grove isn't a pop icon, but he ought to be.

Grove is one of only a handful of people responsible for transforming the computer into a cultural phenomenon. A Hungarian immigrant, who remains just plain Andy to friends and business colleagues, Grove is more professor than icon, more elder statesman than rock star. Intel may have rung up $16.2 billion in total revenues last year, but Grove still teaches a college course, writes management books, and following a bout with prostate cancer last year, wrote with candor about the intimate details of his treatment.

1607K 1623K
Grove likes to say that "only the paranoid survive." What does the head of the world's biggest personal computer chip maker have to be paranoid about? The consumer electronics market. While making up half of the "Wintel" consortium that dominates the personal computer market, Grove still manages to count among friends and admirers not only Bill Gates, but Microsoft foe Larry Ellison and Steve Jobs. Grove thinks that not only will the personal computer do just fine in the home, it will flourish; no need for the network computer. Or will it? We chatted with him about his faith in the PC, Intel's vision for the Internet and its abandoned plans to hook into the consumer market through cable modems.

Intel is a CNET investor.

NEWS.COM: You've said in the past that the PC is the TV of the future. Yet you haven't been very vocal about the network computer (NC). Why not?
Grove: I haven't been because I thought that the NC was going to become a PC. When the Web started people would do very simple static Web pages. Static Web pages are as boring as cold fish. So, we've basically got to get a high-performance PC as the network computer.

NEXT: NC a marketing ploy

Andy Grove

Age: 59

Grove's Law: "Only the paranoid survive."

Thrills: Teaching college business courses, writing books.

Latest book: "Only the Paranoid Survive," Doubleday

CNET News.com Newsmakers
September 9, 1996, Andy Grove
NC a marketing ploy

Why not just connect a CD-ROM to an NC?
You know, the definition of an NC has changed three or four times in the last month. The fact that you mention a local CD-ROM drive, that is no longer a network computer. You just described a personal computer; I'm not sure what the difference is anymore. I think the world is going to be personal computers, connected to the Internet, that have the full resources of a high-performance personal computer available to augment the whole experience.

So has the NC just been a marketing ploy?
It was a ploy to exclude Microsoft software. It's not going to work. How many Netscape browsers operate without being loaded on Windows?

About those static Web pages, don't small pages load faster?
If you're going to track a UPS package, you don't want dancing girls. There's a range of Web applications.

Have you made any projections as to what percentage content, say in the year 2000, is going to be UPS kind of content and what percentage is going to be dancing-girls kind of content?
A year ago would anyone believe that PBS would carry URLs? All you can do with this is get in the water and let it sweep you along. It's a revolutionary phenomenon. And revolutionary phenomena come from one of two rules: shit happens or Moore's law. What's happening here is data transmission has improved 100 to 1,000-fold. That's enough for two revolutions.

That leads us back to the question of the network computer. You may not need more (of a computer) to track a UPS package, but, hey, you might as well buy a little bit more processing. It's only going to cost you $100.

Why wouldn't a competitive product be a telephone handset on a packet-switched network?
It may be a decent product to do a telephone handset on a packet network. It may be useful and could be a very cheap instrument: $100.

I think what is likely to happen is the computer interface that's going to solve all these problems is going to be the browsing interface. You can't convince me that a browsing interface is too hard to use. I can sit down anybody in front of a browser today with the only qualification that he has used the TV remote control before. The browser paradigm will simplify computer use and for untold numbers of people. It's like a VCR control.

What is hard is downloading plug-ins and making sure that the software is up to the right levels. And those things are going to get simplified.

If the NC does pick up more than you think it might, isn't it antithetical to your efforts to try to sell more faster chips?
I don't think so. If all people will want to do is pull up UPS forms on them, then computing as we know it is over and that's a catastrophe to us and to the user. We go back to the 3270 terminals. I don't think that will happen. People will not go back to that.

NEXT: Bandwidth

2316K 2383K
CNET News.com Newsmakers
September 9, 1996, Andy Grove

Did Intel abandon cable modems because it's just too difficult to change the cable infrastructure?
I think there's two things you need to consider. First of all, cable modems are a new infrastructure. And new infrastructures are hellishly difficult to put together end to end. It's a difficult job. Our interest is to provide the best experience that computers and the existing Internet can provide to the 35 million (existing connected PC users), and as cable modems or other high-bandwidth technology (usage) creeps up..., for those who have them that provide a better experience. But our principal objective is to raise the baseline for everyone who's online.

How long will a "hybrid system" of local storage plus network feeds serve our needs?
Really forever. No matter the bandwidth you can provide externally, we can always provide higher bandwidth access from your local disk. Local memory is where all your data winds up anyway and we can match the data rate coming form local memory much easier locally than over the network.

XDSL technology was missing from your list of bandwidth advances. Why?
XDSL technology is an experimental technology. To my knowledge there is no real proof of it in large scale production and no real commitments to employ it. I hope that the proof will come and I hope the deployment will come. Because in principle, on paper and in laboratories, it's a dynamite technology providing high bandwidth. But, again, when it comes, it will come to a small portion of the total connected population. Our interest is to take care of the 60 million, 70 million people who buy computers this year.

If you were gambling and you had to put your money on ISDN or ADSL, which one would you bet on?
I would bet on a combination of POTS (plain ole telephone service) and your local storage, the hybrid application concept.

You're really high on this?
Yes. We concluded this after several years of working with ISDN, with cable modems and we continue to do that because they all provide better experience than the cable network, but the time when they will be available to ordinary users, is unforeseeable, decades away.

You've said that only the paranoid survive. How can you possibly be paranoid of anyone with the success of Intel? Isn't it true that you can virtually dictate standards?
No. This phenomenon is bigger than any of us. Nobody can dictate standards. We all have to work in such a manner that the other parties buy into what you are trying to do. We cannot provide quality of service experience by ourselves. That's why we entered into a partnership with MCI. We do not provide content, or means of delivering DVD, in fact, we don't provide DVDs. There are very large forces and companies at work whose specialized efforts all have to becoming building blocks in this total picture.

990K 1058K