Intel exec on the future of Net, computing
Sean Maloney, a senior vice president of Intel, has had a front-row seat to the Internet revolution. He shares his views on the future with Michael Kanellos.
He's had a front-row seat
Q: Tell us a bit about your background:
Sean Maloney: In the early '80s, I was a kind of incurable college dropout. I was
very, very interested in software and ended up working on large mainframe
systems when I was in
my early 20s. I had written a paper on how you could do some stuff on
mainframes using
Intel memory technology. And anyway,
Intel came and approached me. I started off as one of the very few
employees in Europe who did anything on
software.
Years and years and years later I found myself working for Andy [Grove]. I was his technical assistant in September of '94. I had been working for him for a couple of years and we then went through one of the most public baptisms of fire in the Internet. [The Pentium bug.] It was an astonishing experience to go through, because you realized that there was this huge medium that was growing very, very rapidly that had the power to get ideas across to large numbers of people very, very quickly. And whether the ideas were wrong or false, it didn't matter. They could be put across very, very quickly. I ended up kind of running that activity for Intel, which ended up with our Intel.com staff.
Q: When you started looking at the Pentium bug when did you begin to
realize that it had legs of its own? At first, was Intel saying: "Well, it's a
problem that's not going to affect everybody?"
A: We knew fairly quickly that this was very, very unusual, because we had
had plenty of
errata [bugs] in processors before. There's always been errata. This is the
first time anyone in the
consumer space cared about it and this is an exceedingly minor bug. Yet
somehow this had become huge and consumerized.
But that was kind of less interesting in a lot of ways than when we realized that the Internet was a critical factor in forming opinions and spreading information. It was when we started seeing journalists and CNN starting to reference stuff seen on Usenet because none of the mass journalists really up to that point had a clue what Usenet was. And they never looked at the Net. Go back four years--there were no browsers. Most journalists quite literally didn't have a PC, let alone have Internet access. And suddenly we realized that this medium was full of opinions, one order, two orders of magnitude faster than any medium had done before.
Our immediate lesson was from that moment onwards, you cannot ignore that medium and that that medium was going to get more and more important at setting opinions. And so we started to behave accordingly.
Q: How pervasive is the Internet now?
A: Most big businesses in almost any industry whatsoever, their
information technologies now are completely driven and dominated by
Internet issues. I was with a roundtable of CIOs the other day from a
sprinkling of America's largest companies--and almost all of them were
spending more than 50 percent of their time on Internet-related issues.
Q: What are some of the main issues? Is it Y2K related?
A: No. The first issue is...I'll give them at random, actually. The first
issue is supply line
integration, integration backwards and forward, outwards to your customers
and backwards to
your suppliers.
The second issue is that the top executives of top American companies now start the day by hitting the Net. They get their e-mails and then they want to know what's going on with their competition on the Internet. And so everybody now does general market research, market information, whether you're a CIO, CEO, CFO, on the Internet. Everyone goes off and hits the medium...from 7 till 9, wham, there's all this Net traffic. So it's an opinion setter internally and externally.
Q: Where will most of the development in e-commerce occur?
A: The thing with consumer-related e-commerce is that you hit a natural
limit. You
are not going to get 100 percent of the people buying books on the Web. We're
going to hit a certain
behavioral limit, at which point people like going to bookstores or they
like browsing in CD stores
or they like buying their food in Safeways or whatever. That doesn't apply
to business-to-business
e-commerce. Business-to-business e-commerce will go to 100 percent, or
approximately to 100 percent, because
it is just a better, faster, cheaper way of doing business.
The analogy with business-to-business e-commerce is the way business toggled from being telex- and first-class-letter-based to being fax-based in the 1980s. It went 100 percent--people threw away the Telex machines. And what's going to happen in business-to-business is that people will throw away the fax machines. In small businesses, large businesses, any business-to-business commerce it is just better to do it on e-commerce. Will it take three years, four years, five years? I don't know, but it's everywhere--it's in every country in the world. In every geography the same thing is happening.
Q: What will Intel do to facilitate that or to capitalize on that movement?A: The capitalizing thing is easy, because by and large, you don't do e-commerce without a PC or?some other kind of Internet appliance. That's wonderful, because that gives us an opportunity to sell some other silicon.
From the point-of-view of driving it, I think that you have to be realistic... So then, what's wrong? The first thing is we need better knowledge management. The second is that we need a faster and more robust infrastructure. And I'll come back to the third one in a second.
Let me go back to the first issue, the better knowledge management. There's no question that the most important piece of software development developed since the beginning of the computer is the browser. The browser has, in one blow, opened up computing to the mass market and removed the need for training. And it has given ubiquity of access. But there are many things that need to be improved with how Internet information is handled.
We are deluged in electronic mail. The world has moved from communication by post to communication by e-mail. And what's happened is that we're ending up moving into complete information overload, and we're drowning.
Q: Traffic has geometrically expanded, it seems like.
A: Yeah, with no apparent sign of stopping. And so people are moving to the
point where it's
absolutely common in our industry for people just to do mass deletes on
email. The outgoing side, how I go out and get information, is also
nowhere near where it needs to be. When I go do a search for example, the
information I get back from that search is 95 percent inappropriate, it's
overwhelming frequently, and unless I
understand Boolean operators, which most people don't, I'm going to get
flooded with stuff.
Q: What about infrastructure?
A: We're very excited about VPN [Virtual Private Networks]. If you look at
the voice infrastructure now, the voice infrastructure and the data
infrastructure--once you get beyond the local access point, it's the same.
Voice calls are ending up going over pretty much the same backbones.
Everybody is moving towards using the IP infrastructure for everything. And
so the idea of VPN is that rather than making long-distance connections,
you can make local hops, and you'll get a better quality of service by doing
that.
Q: How much money can people save--20, 40, 50 percent--if they switch from
ordinary phone usage to a VPN?
A: Yeah, and higher. I could probably go off and dig up some examples for
you if you want. But
I'd say very substantial?You
know, we bought
Shiva a few months ago. Shiva is a VPN specialist company.
Q: In layman's terms, the advantage of VPN is that it takes you off
the telecommunication
lines and puts you on the IP Internet structure, correct?
A: That's correct, plus it also gives to you better-than DES encryption.
So you have security of
transmission?I think it's rather drowned in the public eye at the moment
because of the Year 2000 issues, but when you see what Melissa did in the
space of 24-48 hours or that New York Times hack, one of these days fairly
soon there is going to be a successful big under-the-water body blow
against someone's Internet
site. And the way that you protect from that is you do some non-trivial
encryption. Really,
anybody who is doing 48-bit encryption has to recognize that that stuff can
be broken on a
Pentium II-class PC very, very, very quickly. Recent German case law or
consumer case law is
saying anything less than 128-bit is basically lack of diligence by the
company.
Q: A lot of analysts recently have pointed out that communications could be
the next big thing for
Intel. It could even dwarf the growth of processors. How will this whole
networking and
communication business begin to manifest itself as a substantial part of
the Intel
revenues and profits?
A: The PC long ceased really being a computational device and really became a
communications device. And obviously the Internet is what's speeding that
up a hundred fold.
PCs are primarily used for communications now. And so everything to do with
Intel is now
centered around the Internet and communications. Our client side
development is really centered
around optimizing the Internet experience. The server side stuff we do--the
Pentium II Xeon and
Pentium III Xeon, all of the development architectural work and planning is
all based around how
do you make better Internet servers. The whole server industry is going to
be consumed by
Internet issues within two years or three years.
Q: This "PC is dead" issue: Do you see the PC declining in future years as
a result of cell phones, Internet access devices, and the like?
A: No, I don't. I think that what the computer industry has with the
Internet is an
almost infinite demand driver. You can make a case five years ago that
every person should have
a PC. You could make a case now that every room and every car should have
some form or method of accessing the Internet. So the potential requirement
for connectivity devices or computers has gone up dramatically as a
consequence of the Internet.
It's not all going to get serviced by PCs...I think you're going to see more and more simple messaging devices that interface to the Internet. You're probably going to start to see some very stripped down browser appliances over the next year or so.
But I don't see them supplanting PCs, because the price of the PC has come down rapidly, the performance has carried on going through the roof, and there are now hundreds of millions of people who are used to using them and like using them.
Q: What does this imply for Intel's business model? A number of these
devices are going to cost less, so you will have to make more low-cost
processors.
A: I guess I would worry if we were anywhere near saturation. And as far
as Internet access is
concerned, we're only scratching the surface?I
consider them a kind of
upside potential. It would be very nice, for example, to have an Internet
access device in the
kitchen.
Q: The "Intellifridge"
A: Right. Let's say a screen that was in the kitchen where I could go and
hit "epicurian.com," or at
the same time I could look at my mail or I could look at my telephone
messages. Am I going to
want a keyboard and a mouse on that? No, not really, because I might end up
tipping stuff on it.
So a conventional PC in the kitchen is probably going to be a fairly
unusual device. You're more
likely to have just simple Web access devices there.
On the other hand, in other rooms in the house I'm absolutely going to want a full-function PC. So I think that the market potential is large enough that you want to keep a very open mind about what these access devices would be.