INTERVIEW:
Question 1
What is Oracle's position in the rapidly changing world of information?
Answer
There is a major paradigm shift occurring. Technology is going to dominate the world.
Oracle is only number two in the software world, but we are the leader in information
management and data-based management and that's a very great network-centric pursuit.
Oracle does one, and only one, thing well, we manage vast amounts of data and deliver
enormous amounts of data to large numbers of users across the network. We do it
economically, we do it reliably, securely and quickly. It's a very simple client server
model. The server stores data, all kinds of data, and then we operate under a variety of
other kinds of networks and then we talk to client computers on the other side of that
network, delivering data across the network. Historically Oracle began as a relational
data base company, where the kinds of information you'd expect to find in an Oracle data
base were information about inventory, or about your payroll or listed customers,
traditional alphanumeric information about a business. That changed some years ago when we
announced this thing called the Oracle media server. What this means very simply is that
our data base isn't just numbers and letters anymore, we can also store rich text
documents, that is, long strings of text and their associated images, and furthermore we
can store real time information like audio and full motion color video all in the server.
And it used to be we would deliver this information off the server to things like dumb
terminals and then later on to personal computers and now we can deliver that information
to a new set of devices, which collectively might be called information appliances. Those
are things like set top boxes that makes your television smart so you can get interactive
television on demand, personal communicators like the Newton, so that we can monitor data
base events and automatically send you a message letting you know that Oracle stock has
gone up or if Oracle stock has gone down or the traffic conditions in Paris, and then this
blank area right in here called an Internet appliance. And while Internet appliances don't
exist as yet, we think they'll be around and we think they'll be around as early as next
year.
Question 2
Your prediction of the end of the personal computer and the future of the low-cost
Internet terminal has created quite a stir.
Answer
We believe the world is moving from a desktop-centric view to a network-centric view. In
other words, the network becomes the center of gravity. It used to be the mainframe, today
it's the personal computer and it's soon to be the network. And when you suddenly have a
network-centric point of view, you don't need a device anywhere near as complicated or as
expensive as a personal computer to gain access to data on a server. You can build a
multimedia Internet terminal and deliver it for about US$400 or US$500. And the great
thing about an Internet terminal is when you get it, you just plug it into the wall to get
power, to get electrons; you plug it into the network to get information bits and you're
done. It's incredibly simple to install. If there's a new version of the operating system,
you don't drive down to the store to get it in a cardboard box. Simply the next day when
you turn your Internet terminal on, the new operating system is already delivered across
the network. To me it is very strange that we should put bits onto plastic disks, put the
plastic disks into trucks, drive the trucks to stores, empty the boxes into the stores,
have people get into cars and drive down to pick up bits. It seems far more immediate, far
more efficient to just deliver it across the network.
Question 3
Can you tell us about your new multimedia server?
Answer
A strange thing happened to Oracle on the way to the information super highway. We
announced the Oracle media server back in 1993, where we said, We're the only ones in the
world who can deliver full motion, high quality video from a server across a network to
personal computers or set top boxes. We have the technology to build the information super
highway today. That's what we said in 1993. And a couple of people believed us. Both Bell
Atlantic and British Telecom said, Well, OK Oracle, we came in and had a look at your
technology and it looks kind of interesting. We'll sign a major contract with you. And
both of those did. And now in 1995, we have over a dozen of these interactive TV
contracts, but the first two were Bell Atlantic and BT. Now, 1995. How are we doing? We
claimed back in 1993 not only that we could do it, that Oracle could build video servers
economically and reliably and scaleably, and deliver television, video on demand. And
today in 1995 or as I should say by the end of this year because we're just finishing up
deployment of two manager systems, there are two, count them, in the world there are two
working interactive television system. Two. One in Europe, one in the United States. Every
other system contract that has been signed with other vendors has either not delivered,
been canceled, been delayed, failed in some way. There are only two working systems in the
world - by working systems I mean systems with thousands of active users. There is one in
Europe and one in the United States, they're both Oracle. All the others not running. We
have thousands of users both in Europe and the United States paying customers. At the end
of 1995 we will be able to put 100,000 users on a single video library. So you record the
information once, record a movie once, record the news once, and from that one video
server we can serve 100,000 users at about the cost of a VCR for the server. About US$500
per turned on active video screen and a sever, we can do that. And we'll sign contracts..
One of our partners, Apple Computer makes a wonderful set top box and you buy it in
quantities of about US$500 to make your TV an intelligent television.
Question 4
Video-on-demand depends on the information superhighway which is still being developed.
Answer
What Oracle does not and cannot do is run the wire from the server to the subscriber
household, to go ahead and run that. And it looks like the world has agreed that ATM
network is what we're going to build and so we're gradually going to build this ATM video
network. What does Oracle corporation do about this situation ? Who's building the network
? Well, we've decided that perhaps there's something we can do with the existing network.
That a lot has happened in the last two years with existing networks. We've seen the cable
TV network get enhanced with these things called cable modems. Cable modems are now being
trialed and you can just hook up a computer modem directly to your cable TV and get
something between 384 Kbits to 1,5 megabits directly off the cable on an on demand basis.
Telephone networks all over Europe, all over even the United States, parts of Asia have
digital services, integrated services, digital networks. Those things are around. Digital
satellite is now out. Hughes has a very interesting system that is becoming more and more
popular delivering digital satellite, Star TV in Asia, and there's a new version of
Ethernet that is becoming ever more popular around the world called switched Ethernet. Now
all of these existing networks can be kind of glued together to deliver video on demand
without waiting for the ATM information super highway to be built, so that's what we're
working on right now. We're building an Internet video-on-demand network.
Question 5
What are Oracles VOD products?
Answer
There's an Internet video server and an Internet video browser coming from Oracle, and
with these two pieces of software you can participate in the video version of the
Internet. Let's say you are CNN and you have a satellite uplink and whoever wants to
subscribe to your news-on-demand video service just takes your downlink, subscribes, so
every time you put out a new news story, you can send it up to the satellite and I collect
that at Oracle Corporation. Let's say Oracle Corporation decides to subscribe to the news
service, we can distribute that video around our campus on switched Ethernet, a local area
network. Or we could take, for example, classes on Microsoft Windows and make those
classes available to everybody on our campus. We could deliver our own classroom material
to our customers so we're obviously delivering educational material to our customers on an
immediate basis. Stanford University could export courses on economics to other
universities by digital satellite uplink, they downlink it and talk to the different
students on a local area network.
Once you're making copies of the video on these different video servers and different
cities and organizations around the world, then you could distribute that video to an
individual viewer either across a switch Ethernet LAN or across a plain old telephone
line, an ISDN telephone line or across a cable TV network. So if in San Francisco we had
an Internet gateway provider that had one of these video servers and that Internet gateway
provider was taking CNN and took Stanford's economic courses and took Oracle's classes and
they started accumulating a big library of video in the San Francisco gateway provider so
that anyone with an ISDN phone line could just dial up and take a look at the news or take
a course from Stanford or a course from Oracle or Microsoft or what have you. They could
dial up on ISDN or they could use their cable TV ports. We think we can do this now using
the existing digital networks. We feel so strongly about that that we've built this thing
called the Oracle Web station, which we're going to announce in about 30 days. It is a
plain old Internet browser with a couple of differences. It browses not only text and
images, it also browses video; it's video enabled. Another difference is it's
programmable. It's very easy for you to develop applications in this Internet video
browser, so for those of you who are familiar with Sun's Java languages, this has a
language built into it. It has the visual basic language built into our browser but it's a
programmable browser. And it's absolutely Netscape compatible.
We also have a server because every subscriber and every publisher of video has to be
able to store and store video on their server, so we have a Web server and it uses the
Oracle 7 data base for tables, texts and images and uses the Oracle audio video server for
audio and video, and of course it has enhanced security for Web commerce.
Question 6
If there are only two working services for video on demand or interactive TV
trials, what are you doing or what has Oracle done that all the others haven't been able
to do and what was the key technical challenge you've been able to overcome?
Answer
We said a long time ago that the only video server that would work in scale was a
massively parallel computer. And that everyone who tried anything else was going to fail.
And we've been working with massively parallel computers because it really is...
Delivering video on demand is a problem of getting a lot of information off of disks and
onto a network; it doesn't require much processing power, but it requires incredible
bandwidth. Then these vast massively parallel computers have 100 to 1000 times the
bandwidth of a conventional computer. And in fact it's the only way you can build one of
these things. Now, actually Nathan of Microsoft said initially that this would be a huge
problem because big iron's too expensive, but in fact massively parallel computers are
cheaper than personal computers in terms of bits moved off of disks and onto networks.
Question 7
Now, there various types, different types of massively parallel machines. Are
there particular models that work better than others ?
Answer
You want the cheapest massively parallel machine you can buy. And sort of the two
that are working both are both nCUBE machines but we're moving our software onto the Intel
massively parallel machine. We've made a major announcement with Intel. In fact the ISDN
technology you saw up here we developed jointly with Intel. We're moving our technology to
the Intel massively parallel computer, which looks very, very attractive. Tandem's coming
out with a machine next year that's massively parallel, so you'll see lots of choices in
the MPP space but, again, it's the only technology that will work. Now, Microsoft is
trying to build their own massively parallel computer by taking lots and lots of PCs and
connecting them with an ATM switch, but that's much too expensive and much too slow.
Question 8
You mentioned that one of the services is in Europe. Are there paying customers on
that or are they still on trial?
Answer
They're actually on phase three of the service, and I'm not sure what I can or
can't say. We're talking about paying customers both in Europe and in the UK measured in
the thousands or tens of thousands. I think there will be thousands in the UK and tens of
thousands on the East coast of the United States. And by the way, we have lots of other
contracts as well that are going well. And as I say, we can install this thing very, very
fast. We can get these systems up and running. If you've got a network, we can get these
systems up and running and, you know, in three or four months. There's nothing hard about
this at all.
Question 9
I assume that you would preferred to be the sole vendor in this market, some would
actually argue that's not necessarily good for the market.
Answer
We are the only vendor that can do it economically for lots of users. There are
little companies like Star Life that can do 30 users on a LAN. We can do hundreds of
thousands of users on a wide area network, a telephone network and we can also do 30 users
on a LAN on a PC. Our software runs on a PC for local area networks. We can do television
quality video across a LAN, when we just can keep on scaling since we've always built our
software such that it's portable. It runs on PCs, it runs on big Sun servers and HP
servers. It runs on massively parallel machines, so we can scale from PCs all the way up.
As far as being the only one, I go back and say, Someone has to be first. So we were the
first to do this, I'm sure others will do it in time. But we've been doing it for a couple
of years and still no one has anything that works and we're taking about huge failures.
The system in Orlando had a lot of trouble. My dear friends at Time Warner won't be happy
to hear this. They spent US$150 million with silicon graphics and they about 30 or 40
users before they had to turn it off, but they're still determined. I mean the telephone
company in the Midwest that picked a huge computer company in the United States fired
everybody and gave up when their project failed.
Question 10
So what have you learned from the trials. I'm sure there must have been some
setbacks for the company anyway.
Answer
From a technology standpoint there were no setbacks. I mean, sure it was tough to
develop the software. This is our core confidence. What we do well is to take information
off servers and put it onto networks and deliver it to lots of users. That's what we've
been doing for almost 25 years. It is natural for us to be good at this. This is the
center of our competency.
Question 11
The Internet server you mentioned confirms the idea of the Internet being the
network of the computer and the need for desktop power and shrink-wrap software to go
away. This harkens back to almost a time sharing paradigm. Does that make you nervous the
idea that you are pioneering a new time sharing paradigm that ten years later will go away
like the old one did?
Answer
Well it's not exactly time sharing in the sense, Microsoft Network is time sharing
and America Online is time sharing. The Internet really isn't. The Internet is a
publisher-subscriber model. Anyone can be a publisher of information. Anyone can be a
subscriber and user of information. It really is an open model of computing. I know it's
very exciting because what we have right now... I remember a long time ago, it probably
was ten years ago, I made a speech saying that IBM was in deep trouble and I was almost
laughed off the stage because, again, it was the peak of IBM's influence, it was the mid
1980s. The PC is a ridiculous device. The idea is so complicated and so expensive. The PC
world made fun of the mainframe; they said that the mainframe is much too complicated and
much too expensive and, therefore, it has to die. Well, you can make the same statement
about the PC; it's now a mainframe on your desktop. You have to back it up. You have to
have experts keeping the LAN running, you have to train your users. It costs between
US$4000 and US$5000 per year for an organization to maintain each and every personal
computer. That's unbelievable the cost. And if one breaks and you didn't back it up, you
lose all your information. It's really a nightmare to run client server computing; these
devices are enormously complicated. What the world really wants I believe are information
appliances that I plug into the wall to get electrons, to get power, and I plug into the
wall to get data. Now people say, Well, now I don't need to get data across a network.
Well, that's nonsense. You get electric power across the network. Why shouldn't you get
data across the network? It's much more economical, it's much more reliable, it's much
more secure. Again, the bottom line is it's much cheaper and these appliances will cost
US$400 to US$500 and give you video and audio and if they break, unplug it and throw it
away.
Question 12
So I guess, do you have any time frame for I guess the death of the PC or local
on-line services of today, which will go first?
Answer
This reminds me of ten years ago when I predicted the death of the mainframe. It's
not really the death of the mainframe or the death of the PC because there are more
mainframes today than there were ten years ago. There are more mainframes today then ten
years ago. The mainframe did not die. The mainframe is just no longer the center of the
universe. The PC isn't going to die. It's just no longer going to be the center of the
universe, the network will be the center of the universe. And the bulk of spending and IT
interest will be around the network, not around the desktop. And that makes complete
sense. So you'll see in ten years from today they'll probably be more PCs than there are
today. But there will be a lot more of these appliances than there are PCs because these
appliances will replace every telephone, every television, they'll be all over the place.
When you check into a hotel room, there will be one in your hotel room. There will be one
of these information appliances in your pocket, all of whom can connect with the network
across radio networks, across wire networks and give you immediate access to information.
Question 13
Do you see the event of these appliance being the same vendors, the PCs vendors or
are we talking a whole new genre of companies?
Answer
I think you'll see the PC vendors get into this once it becomes obvious that they
are important, but I think you'll see a whole new genre of vendors, where I don't know if
people watching closely, but what's going on in Korea there are these enormous companies
like Samsung and Lucky Gold Star that are completely vertically integrated. We've seen
D-RAM manufacturing, memory manufacturing move from the United States to Japan and now
from Japan to Korea. Most of the value in computers is in the parts, not in the
manufacturing, manufacturing. Marketing may be a big deal and distribution may be a big
deal, manufacturing is easy. It's certain key components. And the key components are the
microprocessor, the memory, the D-RAM and the screen, the thin film. And Korea is a leader
in D-RAMs and they're soon going to be the leader in flat panel displays, thin film
transistor displays. So Samsung and Lucky Gold Star will have the ability to build these
devices where you can imagine factories with petroleum and sand on one side and appliance
computers coming out the other side at very low cost.
Question 14
Which software vendors do you worry about most as competitors, Microsoft or
companies that haven't even started yet?
Answer
Everyone tries to find an answer to that question and say someone other than
Microsoft, but you know when Bank of America and Citibank worry about Microsoft, I guess
we ought to, too. Clearly everyone is concerned with Microsoft not so much as the supplier
of technology but because they have this incredible brand power. You know, it really
harkens back to the mid 1980s when IBM said, Do this, and customers used to do what IBM
said. Customers were looking to IBM to tell them what to do. This is an enormously
complicated industry and you have to rely on some vendor to plot a road map, a technology
road map. You have to rely on somebody. Well, in the mid 1990s, who do you rely on to
provide the technology road map? It's perfectly reasonable for customers to want that.
Well, today Microsoft has more brand power than anybody else and we are concerned about
that brand power. As far as a real technology provider, we are less concerned but let's
say brand and cash goes a long way.
Question 15
Speaking of cash, what about Netscape? Are they not a business partner of yours?
Answer
Netscape actually uses an Oracle data bases inside of every Netscape server. That
doesn't cause us to be critical of their market value. It's absolutely astonishing. I
think the evaluation of Netscape was responsible for the collapse of technology prices
because it was so insane. Here's a company that got 75 per cent market share, by giving
away their most popular product for nothing, in the face of no commercial competition
whatsoever. And that most popular product is a 9000-line program, the Netscape browser is
a 9000-line program in an era when million-line programs are commonplace. Now, my cat can
write a 9000-line program. It's very easy to build a better browser than Netscape. All it
does basically is display stuff on a screen. I mean it's incredible. And then George
Gilder, in an article, predicted that Marc Andreissen would become the next Bill Gates,
you should ask Bill about that.
Question 16
While we're on the subject of software companies, what's your personal opinion of
the IBM purchase of Lotus? Good news, bad news?
Answer
I think the IBM purchase of Lotus was good news for IBM and it's good news for
Lotus. I think IBM will makes Lotus Notes more successful than Lotus could ever have made
it because they can put more money into marketing it. Lotus Notes is a data base product,
an application development tool and it has two interesting technological characteristics,
one is that it gains its scalability through replication technology, it makes copies of
the data onto several servers, and because of using replication, it operates
intermittently connected disconnected operations and it does a good job of handling text.
Now, the questions is, Do you really want a Lotus Notes data base or let's say an Oracle
data base where some of the data base is in Oracle and some of the data base in Lotus
Notes? In an ideal with mature technology you'd like one data base that does everything
Lotus Notes does and everything Oracle does, so you only have one data base and you can
carry it around in your laptop and you can replicate and do all of these things. That's
what Oracle 8 is all about. The next release really will do everything that Lotus Notes
can do. And that's really focused on the very high end, for mainframe replacement and a
very low end you know, Lotus laptops and Lotus Notes. But how do you decide what
application to put in Notes and what application to put in Oracle? You'd like to have a
unified data base. Therefore we think Notes is a very interesting point solution, where
the world really wants an integrated solution, but it's doing very well because there
isn't an integrated solution available. When there is one, people will choose that.
Question 17
What do you think is your single most important personal challenge in the next
three years. What mountain do you feel you need to scale?
Answer
Well, I think the Internet opportunity is enormously important to us. We think it
is a paradigm shift. We have watched for more than a decade, where IBM helped create
Microsoft, then IBM helped create their own downfall. I jokingly said that the
introduction of what was mistakenly called the IBM PC (it is really the Microsoft Intel
PC), was the biggest single business mistake in the history of enterprise on earth. It
created Intel and Microsoft, which are US$120 billion of market value, which belonged to
IBM at the time. You don't get the chance to make US$100 billion mistakes very often. So
I'm impressed by that one. Those with the most, have the most to lose. What's going on
right now with this next paradigm shift away from the desktop into network-centric is
playing into our strength. IBM helped Microsoft become Microsoft; now with the network as
the center of the world there's an opportunity for other companies to really take
advantage of that and it's right at the heart of our core competence. We have servers,
manage lots of data distributed across networks to lots of users. This is our real
strength; this is a once in a lifetime opportunity for Oracle Corporation and for me
personally, so my real focus is on that.
Question 18
Is there anything unique about the European markets that you've discovered in the
trials and the tests and in your daily business?
Answer
I think it varies from country to country; as yet Europe is not unified. Someone
said jokingly the last person to successfully unify Europe was Napoleon, and we're making
progress towards the European Union but in some countries in Europe the telecom companies
have been completely deregulated and are aggressively competing and assuming new
technologies. In other countries that deregulation has yet to occur so it really varies.
It's hard to speak of Europe in a single sentence. You have to go from country to country,
from Germany to France, to the UK to figure out exactly where you are. But one thing is
absolutely true in Europe: that ISDN is much more popular in Europe than any place else in
the world. And we have a huge chance to take advantage of the ISDN deployment in Europe,
which is way ahead of anywhere else, and the demonstration I just did was all ISDN based
and it's easier to install than it is in North America.
|
 |
