SUNERGY TOKYO
"Java Today: From Smart Cards to Supercomputers"
July 16, 1997

TRANSCRIPT

JOHN GAGE: Welcome. Welcome to Sunergy. We are coming to you today live from Tokyo, by satellite to all of Asia. We are coming to Japan, of course, to the Philippines, to Indonesia, to Australia, Singapore, Malaysia, and for the first time to the People's Republic of China. Hello to all of you in Beijing.

Today's topic is Java. But more specifically, Java in devices, small devices. We have three panels, we will first discuss Java in hand-held devices, small devices that begin to become consumer electronics devices.

Secondly we will discuss Java in chips, Java chips that execute Java programs very rapidly.

And lastly, we will explore Java in smart cards, these tiny computers, the thinnest of thin clients.

So, let's begin. Let's begin by discussing Java in hand-held devices. With us today, we have Miko Matsumura, from JavaSoft.
And Yoshito Yamaguchi, senior managing director from Mitsubishi, and one of the first Java adopters, familiar with Java from a time far before Java was named Java. From the days Java was known as Oak and then Green.

YOSHITO YAMAGUCHI: Green, then Oak.

JOHN GAGE: Green, then Oak.

Now, they call you Super.

YOSHITO YAMAGUCHI: Yes.

JOHN GAGE: Should I call you Super.

YOSHITO YAMAGUCHI: Sure.

JOHN GAGE: Super, let me ask first, Miko, if you were addressing today seven years after the Green then Oak development project, if you were addressing as senior managing director of a company like Mitsubishi and saying what the benefits of Java would be for hand-held devices, what would you say?

MIKO MATSUMURA: Well, I would ask that that person think of Java as a universal fuel for computing and to think of that in those terms. If you draw attention to, there are a number of devices on a table right behind me. If you just turn to this table you will see that there are a large number of devices represented there. What we have is we have a personal digital assistant device, and we have a cellular telephone and there are a number of other devices as well that you will see, including smart cards or a camera, and in fact in the middle you have a network computer type device from Acer, this device actually runs Java OS in a localized version which is displaying Chinese characters, so you can envision, I would ask that this person envision the universal fuel for computation that will enable Java to pour and interoperate across all these diverse types of devices.

JOHN GAGE: Then the person would say does it really work?

MIKO MATSUMURA: Oh, absolutely. Java comes in several different flavors that are appropriate for multiple applications, including personal flavor, embedded flavor, and enterprise Java, there is even a Java card flavor of Java. And the goal is to have many interoperable pieces of Java that all are able to converse.

JOHN GAGE: Now, in the beginning -- I have something, a special treat for you, Super. This device, the Star Seven, is one of the only two remaining first Sun PDAs, almost a consumer electronic, because you would not think this is a consumer electronics device, it's a little heavy and expensive. This device was the very first demonstration of a language then called I suppose we called it Green at the very beginning, then Oak.

YOSHITO YAMAGUCHI: Yes.

JOHN GAGE: Each entity here on this screen is an object. So I touch the home, and the home becomes --

YOSHITO YAMAGUCHI: Bigger.

JOHN GAGE: -- bigger. It's an animation interface. And if I go back to the town where I was, I can move around in the town and you notice there is physics in the interface, there is friction, momentum, it moves and slows down.

YOSHITO YAMAGUCHI: Right.

JOHN GAGE: So if I come back to the home and I go to the place Mitsubishi is most interested in, into the consumer electronics, the living room, and I scan around and I see devices.

YOSHITO YAMAGUCHI: Sure.

JOHN GAGE: I can find out about them by touching them. This is what you first saw.

YOSHITO YAMAGUCHI: Right.

JOHN GAGE: What attracted you to this? What did you think when you heard about Oak or Java?

YOSHITO YAMAGUCHI: Right. It was about 1991 or '2, I think maybe, when the Star Seven, this device I saw at Scott McNealy's office, and this little Duke which has become so famous now, that is the little agent. That was the very beginning. We were asked as a consumer manufacturer, to cooperate to develop such a device. And we were thinking at that time, a remote control with some display and it can be the telephone with some infrared capability and download like TV guide and we can see it and/or you can call it. So we called it a phantom remote control or like a big remote control.

JOHN GAGE: I remember for you we did the TV guide so we could touch the TV guide, the TV guide would then jump up on the screen and show all the days, you could move forward in time or backward in time and pick any program.

YOSHITO YAMAGUCHI: Right.

JOHN GAGE: And simply by touching the program and dropping it on the VCR, it would program the VCR.

YOSHITO YAMAGUCHI: That's right.

JOHN GAGE: That idea had to be implemented in a programming language that was small.

YOSHITO YAMAGUCHI: Right. So that was before the Internet exploded. We thought this is very good to connect the TV to the telephone line and download it. That is an agent, as you know. So we started -- we were the first consumer electronics people to work with Sun and I signed the early access agreement and Wayne Rosing at the time, FirstPerson, Inc., that give me the free, but of course one dollar he said, so I paid one thousand dollar signature on -- my signature on it. So that was how we quoted to the M16 and we started working with that product. But it didn't work because we needed infrastructure, everything had to be -- so in that time then Internet came in. Exploded. So the Sun shifted, you shifted to Java. That is like a language or a browser of the Internet. So that's how Java came in.

JOHN GAGE: First consumer. I still remember people said you know, we went to the consumer electronics and we said we might need four megabytes of memory and everyone in consumer electronics laughed at us.

YOSHITO YAMAGUCHI: 200, 300 big, but you said four megabytes.

JOHN GAGE: Not possible.

YOSHITO YAMAGUCHI: Too heavy.

JOHN GAGE: But today things have changed, now Java became a more general programming language. In some sense you were saying it makes a cycle.

Miko, you were just saying the device can be as small as those devices you showed, a cellular phone. Are there advances we have made in Java to help applications run now in consumer electronics devices?

MIKO MATSUMURA: Absolutely. In fact what we were looking at as far as personal Java and consumer Java is a way to make these types of applications run in smaller and smaller spaces even down to the smart card space. And this all stems from the philosophy intrinsic in Java which is related to the idea of conversation and communication. So instead of having a single API that's forced to exist in all environments, we simply have environments that can communicate with each other across a number of very diverse platforms. So philosophically it becomes much easier to have device diversity.

JOHN GAGE: So a device like this, which is now seven years old, I think the first hardware version was made in 1990, I think before the Newton, we had a serious hand-held device that communicated by radio and infrared. This device forced James Gosling to make the language small. And to make it simple. And that was a lesson the computer industry had to learn from the consumer electronics industry.

YOSHITO YAMAGUCHI: Right.

JOHN GAGE: Now, have we learned our lesson well enough? Are you using Java in new products?

YOSHITO YAMAGUCHI: Yes. I can show you some of the devices like Miko had that one. So this is --

JOHN GAGE: I will show them.

YOSHITO YAMAGUCHI: Maybe he will explain later.

JOHN GAGE: Explain later, all right.

YOSHITO YAMAGUCHI: This is one. This is an NC we call MonAMI/NC. Just started marketing in the vertical market in the United States. And this is a workstation, it's engineering workstations. With this TFT display and this is MonAMI/ES, so those are the products.

JOHN GAGE: So Mitsubishi is marketing these.

YOSHITO YAMAGUCHI: Yes, we start to market for the vertical market. Because you have to have systems like the home or office, road. This device is like a cellular telephone. This we just started marketing in the United States but it's not Java on it yet. But this will be marketed by AT&T. And this can access the Internet.

JOHN GAGE: So you can download Java applications when the Java runs here.

YOSHITO YAMAGUCHI: Yes, right. That is enough -- Japanese version like a PHS, or it is a little smaller but in this size we can put enough devices and this is four by 12 line.

JOHN GAGE: So the display is good.

YOSHITO YAMAGUCHI: Display and you can, through the proxy server, you can easily access to all the stock, the market information, or like a weather forecast or whatever. You can scroll down.

JOHN GAGE: So if I put this next to the Star Seven, at some point we can move this interface right over here.

YOSHITO YAMAGUCHI: That's right. That will fit into that. Of course you can do that with a more larger devices.

JOHN GAGE: Show us this one, Miko.

MIKO MATSUMURA: This is a tremendous accomplishment. It's a network computer in a hand-held form factor, and as you can probably see, it's running HotJava, and in fact the entire Java OS is running directly on their metal. So this is straight through Java technology in an extremely compact and hand-held format. And in fact you can drive the entire system with a pen. I think it's a tremendous accomplishment for Mitsubishi.

JOHN GAGE: With the radio, then that's all. Now can you talk into this and replace the cell phone? That's the next.

YOSHITO YAMAGUCHI: Of course you can, I think. That is pretty easy. So in that volume, then you can put anything that's Java OS. But I am telling you that with this type of device you can do it. That is the beauty of Java, right?

JOHN GAGE: What is the future for the television set?

YOSHITO YAMAGUCHI: The television set will be digital. United States FCC now ruled last Christmas, and we will give you, to the older network people, give your bandwidth the next door to Europe, like NBC, ABC, we will give you a 6 Megahertz next door to your current analog channel. So you have to come transit, or change to digital starting 1998 to 1996. So that's end of 1996 FCC will confiscate.

JOHN GAGE: Take the bandwidth.

YOSHITO YAMAGUCHI: Take the bandwidth back and they will put in the auction. So that they have to migrate --

JOHN GAGE: To digital --

YOSHITO YAMAGUCHI: -- to digital --

JOHN GAGE: -- television.

YOSHITO YAMAGUCHI: -- television. If it's digital, of course you can use that 6 Megahertz. It used to be a just an one nanohertz channel you can transmit, but you can transmit four if it's standard TV, even in high definition TV you can transmit the beautiful picture and some like optimistic data.

JOHN GAGE: Now some of that data could be going into the television set --

YOSHITO YAMAGUCHI: Sure.

JOHN GAGE: -- and the television set could be smart enough to ran Java.

YOSHITO YAMAGUCHI: Sure. And you can download some of the information and put into the, of course two or three screen, or if you want to see the normal record you can immediately see on the screen. A lot of information at the same time. So digital, TV on digital meaning new possibilities erupt.

JOHN GAGE: Now is Java prepared to take advantage? Do you think that programmers will target the television set as a destination for Java applications?

MIKO MATSUMURA: Oh, absolutely. In fact at the Java One conference we actually demonstrated a television that was running Java OS, and the name of the hardware company that was associated with, I can't disclose, but the software provider was Applix of Japan.

JOHN GAGE: So Applix wrote Java to run in the television.

MIKO MATSUMURA: And it literally ran on the television set, it was quite impressive.

JOHN GAGE: I do remember the first time the Star Seven showed a way to index hundreds of television programs and all you needed to do to program the VCR or program the television set was touch the program, as I was doing before. Move it over and drop it on to the television set. And suddenly the television set would simply tune to that channel. These are all the controls for the television set. It makes it easier.

YOSHITO YAMAGUCHI: Sure.

JOHN GAGE: And that's our goal, I think, to broaden the market. Now with bandwidth coming from digital television, bandwidth coming from cellular telephones, bandwidth coming here and there linking everyone, we are making a world that seems ready for thousands of applications. How do people develop today? Do they have a warm helpful development environment? Are Java Beans giving them the power to start building new things?

MIKO MATSUMURA: I believe that Java development environment is going to become an extremely ubiquitous and scalable method for deploying applications. So what I mean by that is if you are developing a multimedia application there will be Java support. For example, Random Noise Coda is a product that allows you to develop very easily web based interactive applications in Java. So you don't really have to necessarily code in the Sun's Java Developer Kit, which is a fairly complex way of programming. Certainly more complex than programming your VCR, for example.

JOHN GAGE: Are there hints for the programmer? If I'm destined, if my application is destined for a screen like this, I would need to know it's smaller so that I would not try to display a hundred characters. Are there hints for a programmer about the destination of an application or can the application modify itself depending upon the device it finally runs on?

MIKO MATSUMURA: Yes. Rather than the notion of hints, what we wanted to accomplish with Java is the ability to run in a variety of environments, even ones that the programmer did not envision. And what that means is, we have recently announced the Java Foundation Classes, which are based on the abstract window tool kit. What that means is that when an user presses a button they could be tapping a pen on this screen, or they could be tapping the button on the surface of the phone, or they could be using a joystick or a mouse that the button tapping has been created on a layer of abstraction. And it's actually a very powerful thing that allows these devices to operate that way.

JOHN GAGE: I think what we should do, Super, is find this device, find you one to take home. So the memories of how long ago it was when this first idea made Java compress into a small environment and then be ready to run in every kind of device. I want to go talk about the smallest device, the chip level, the Java chips.

YOSHITO YAMAGUCHI: Good.

JOHN GAGE: So let me thank you, Super Yamaguchi, it has been a great pleasure to have you here. You must write down the first days of Green then Oak then Java.

YOSHITO YAMAGUCHI: Okay, I will.

JOHN GAGE: I will be back with you in a moment, Miko.

YOSHITO YAMAGUCHI: Thank you.

JOHN GAGE: Let me join Dennis Tsu, and Kozy Kubota. Dennis Tsu, the global marketing manager for Sun. And Kozy, NEC Semiconductor, you are building Java chips.

KAZUHIRO KUBOTA: Right.

JOHN GAGE: Small, fast Java chips.

KAZUHIRO KUBOTA: Right.

JOHN GAGE: Tell me about the chips. What are they, how big are they, how fast do they go? What do you think the future is for Java chips?

KAZUHIRO KUBOTA: Now, we are working with Sun microelectronics and developing the Java chip and actually I believe a variety of JVMs will appear soon as business products. And for instance, high performance JVM, low cost JVM, low power JVM, and so on.

JOHN GAGE: You are making chips at these different ranges for different applications.

KAZUHIRO KUBOTA: Yes, that's right. The important thing is the fact that there will be a lot of JVMs which the system designer can select. So we will provide our own chip to our customers, of course including Java chips solution. And our target is do it five times faster than Pentium processor.

JOHN GAGE: Five times faster than Pentium? At every clock rate for Pentium?

KAZUHIRO KUBOTA: At the same frequency.

JOHN GAGE: At the same frequency. So a Pentium running at 300 Megahertz, the NEC Java chip would run five times faster, run Java five times faster.

KAZUHIRO KUBOTA: Running byte code.

JOHN GAGE: Running byte code. A five time factor makes -- Dennis, what does a five time speed improvement over Pentium mean to you?

DENNIS TSU: Wow. It means multimedia, it means user interfaces that can be customized and localized in any language anywhere in the world. It means search capabilities that you don't have to wait at Internet snail crawl time to try and get the information that you are looking for. So it's a really neat sort of thing that we are looking forward to.

JOHN GAGE: What volumes are we talking about? NEC makes 30 million chips a month? How many --

KAZUHIRO KUBOTA: Actually 50 million per month.

JOHN GAGE: 50 million.

KAZUHIRO KUBOTA: Yes.

JOHN GAGE: 50. Intel doesn't make 50 million chips a year.

KAZUHIRO KUBOTA: Yes. Actually 50 million microprocessor and microcontroller we are shipping every month. Of course we are now focusing on the embedded area, Miko introduced those kinds of devices. Now we are focusing on this area with Java chip.

JOHN GAGE: So the volume for each of these different Java chips could be very, very large.

KAZUHIRO KUBOTA: Yes. I think so. I believe so.

JOHN GAGE: Now the low power chip, what are the destination applications for a low power chip?

KAZUHIRO KUBOTA: The second item, low power is very important, I believe, in embedded area. Of course the digital phone and the digital camera, including this platform.

JOHN GAGE: So with those chips you could reach into the cameras, not only the television cameras but the small hand-held cameras, cameras everywhere, cellular phones, and in the automobile, is that a target market for you to put embedded chips?

KAZUHIRO KUBOTA: Of course. Automobiles of course. There is not only automobile but also actually the microwave or some kitchen equipment or something.

JOHN GAGE: Refrigerators, air conditioners, all of those appliances.

For power control, Dennis, do you think this is an interesting market?

DENNIS TSU: Oh, I very much think so. This, the sort of embedded controllers that Kozy has been describing here really get to the example that Scott and others at Sun like to talk about of integrating all of your household appliances so when you get up out of bed in the morning you can step on the bathroom scale and if you weigh too much the door to the refrigerator locks, and all of the Java processors are talking to one another. So those are cute examples, but there is actually practical application to that.

JOHN GAGE: So a smart house could be a very stern house that disciplines you if you don't do things properly.

DENNIS TSU: It could indeed.

JOHN GAGE: I will never have to tell the children to turn the lights off again, the house will know.

DENNIS TSU: Right.

JOHN GAGE: At NEC, Kozy, you have been showing a device you built that might be in the home. This is -- I will hold this up so people can see, this device, with connectors and infrared and a large screen, tell me about this device.

KAZUHIRO KUBOTA: This is called Web Panel.

JOHN GAGE: Web Panel.

KAZUHIRO KUBOTA: Yes. It's a rebalanced platform Java chip, power chip. And also we, the nickname is Kotastu, not desktop, not table top, it's Kotastutop.

JOHN GAGE: What does that mean?

KAZUHIRO KUBOTA: The kotastu is a very common furniture in Japan, that is a kind of table. You keep it with a stove.

JOHN GAGE: A stove in the table?

KAZUHIRO KUBOTA: Yes. Please show this cartoon.

JOHN GAGE: All right.

KAZUHIRO KUBOTA: This is kotastu.

JOHN GAGE: I see. So the family sits around and the table has a stove beneath it and it keeps them warm. And while they are warm they can either eat or they can use this to play Go.

KAZUHIRO KUBOTA: Right, exactly. It means very warm and closeness, the word kotastu, that's what it means.

JOHN GAGE: Well, this device could also be used in vertical applications, in a hospital or in any application that needs mobility.

KAZUHIRO KUBOTA: Yes. I think the first time frame we will penetrate the vertical market, you know, hospital or some perhaps in supermarket. And then the final role will be home network client. Actually we are developing for my wife and daughters.

JOHN GAGE: The best marketing, absolutely the best judges of what is needed.

KAZUHIRO KUBOTA: Yes, I agree.

JOHN GAGE: Now do you ask your wife, your children about what should be done with Java, with Java chips?

DENNIS TSU: Well, I don't necessarily ask them what should be done with the chips, but the application is one my wife is already looking forward to. One of her favorite sites on the Internet is www.epicurious.com.

JOHN GAGE: Epicurious.

DENNIS TSU: Epicurious meaning --

JOHN GAGE: Eating.

DENNIS TSU: -- eating. And it has a wealth of recipes and you could actually have this device hung on the kitchen wall right next to the stove, and you could be surfing the web and pulling down recipes, and as you are cooking it could be giving you line by line, even video instructions based on the new bandwidths and the new chips that are coming out.

JOHN GAGE: Now we have new TFT or other plastic display technologies that give us very thin displays. I don't know what the time for Mitsubishi or for NEC is, but it's within the next year or two, we will have very, very thin and we hope low power displays. In the development by NEC Semiconductor of these devices, what market projections do you have? If you segmented, you said there were low power devices, then there are the very fast devices, these are to run different styles of virtual machine. Can you make projections, have you any ideas of the various markets that would absorb them? 50 million automobiles are manufactured every year, each automobile has many chips. What guess would you make about the possible market?

KAZUHIRO KUBOTA: Now actually I'm talking with not only system designers but also network operators and content providers for the Web Panel. Web panel is just egg. We need chicken.

DENNIS TSU: This is the seed that needs to be incubated.

JOHN GAGE: I see.

DENNIS TSU: So this is really a reference platform from NEC, as an example to the industry of what can be done with the chip and the technology, and NEC and Sun are working together to try and proliferate a variety of experiments and pilots in the industry to take advantage of the technology.

JOHN GAGE: Now would your chips go in smart cards? Does NEC sell into the smart card world?

KAZUHIRO KUBOTA: Yes. Actually the chip for smart card, yes, we are shipping from one dollar microcontroller to one thousand microprocessor, very long range. So of course now we are planning the Java chip for smart card. One dollar, two dollar, I don't know, but very strict business.

JOHN GAGE: Yes, huge volume of business.

KAZUHIRO KUBOTA: Correct.

JOHN GAGE: For developers, you say that you need, if this is the egg you need the chicken, is that what you say?

DENNIS TSU: Yes.

JOHN GAGE: So all of the Japanese chickens that would develop applications for this, how do you find them? Are there web sites or places that they should read or should they send you mail personally, if they think of something good?

KAZUHIRO KUBOTA: You mean e-mail?

JOHN GAGE: Well, or a web site. You were talking about --

DENNIS TSU: You mentioned you were president of an association here in Japan.

KAZUHIRO KUBOTA: With this Web Panel.

JOHN GAGE: With Web Panel there was a FJ.

KAZUHIRO KUBOTA: Oh, I see. Actually I'm a board member of a Java conference, Java conference in the kind of committee to penetrate Java technology in Japan market.

JOHN GAGE: So many people can participate in this conference?

KAZUHIRO KUBOTA: Yes, that's right. Over a hundred companies.

JOHN GAGE: 100 companies.

KAZUHIRO KUBOTA: Yes. Joined the Java conference.

JOHN GAGE: What web site is this?

KAZUHIRO KUBOTA: www.java-fj.or.jp.

JOHN GAGE: All right. I will hold this up. Is the site in Japanese?

KAZUHIRO KUBOTA: It is in Japanese, but English home page, also.

JOHN GAGE: Also English. So anyone in the world that is writing Java applications can go to these pages and be part of a conference of a hundred companies looking for the chickens developing the eggs.

DENNIS TSU: Hatching the eggs.

JOHN GAGE: Hatching the eggs.

DENNIS TSU: Yes.

JOHN GAGE: Well, NEC has made a large commitment to Java. How long has the NEC semiconductor division been developing Java chips?

KAZUHIRO KUBOTA: One year and four months.

JOHN GAGE: That is fast work to be able to make microprocessors across a range.
Dennis, what do you think the market size might be, not for the chickens, but for these chips? These are very large numbers --

DENNIS TSU: Right.

JOHN GAGE: -- if the applications work. Is there any way to guess who the first adopters might be?

DENNIS TSU: Well, John, if you take it back to the network computer market, or the Java station market, which I think most of the viewers have probably heard about today, the projections from leading market research firms in the United States is that that will be about a $1 billion market in calendar '97 into calendar '98. And that's a $1 billion market with about an average sales price of roughly a thousand dollars. So that's a hundred thousand units that we are looking at, or is that a million? That's a million units that we are looking at in this time frame just for network computers.

JOHN GAGE: These are the larger devices.

DENNIS TSU: These are the larger devices.

JOHN GAGE: A million units for NEC is just a blink of the eye.

DENNIS TSU: Right. But now network computers as well are going to go through an evolution and a proliferation much as Kozy was describing the evolution of the chip family. So you will see higher powered more scalable expandable network computers selling for perhaps $1,500, $2,000, and you will see more specialized focused network computers that will approach the ideal $500 price point or perhaps lower. And so that would be the initial market, given the initial chip set and the price points and so forth. That's going to be one market that develops but then we are really looking forward to and working with NEC and other partners on is really trying to drive the chips and the volumes up by entering some of the lower end consumer device markets.

JOHN GAGE: Embedded devices.

DENNIS TSU: Embedded devices.

JOHN GAGE: Tiny devices, all of the appliances.

DENNIS TSU: Right.

JOHN GAGE: This is the beginning of the possibility to make, well, among other things, to make the efficient use of the power in a household. If the refrigerator is smart.

DENNIS TSU: Right.

JOHN GAGE: And the air conditioner is smart, and they know when to adjust their usage to keep their electric bill low, every consumer will be happy.

DENNIS TSU: Just to pick up on that example, John, again to put this into very human terms, I have today a calendar that's on my little Pilot device that's linked up and hooked up and synchronized over the network with my calendar on my workstation. There is no reason, once they all have Java embedded, that those two could not talk to my alarm clock at home. So that if I or my administrator were to schedule a 7:30 morning breakfast meeting, which is earlier than my normal 8:30 start at work, it would reprogram my alarm clock to get me out of bed one hour earlier and I wouldn't need to do a thing other than enter it on my workstation.

JOHN GAGE: So this is a new market opportunity for NEC, the alarm clock business. This is something I'm not sure NEC has really penetrated before.

Well, thank you very much, Kozy, for this news about these chips. I want them soon and I want a lot of them. But you mentioned the reason that this all works together is the network. As the network links us all, it links these devices, it makes the value of Java network more valuable because programs can move across the network. We have a video that talks about the spread of the network, first in South America and then in China. So let's look at our world news segment which doesn't deal with Java particularly, but with what makes Java so important for everyone, the network. Let's look at that video now.

(video playing.)

THE NARRATOR: The advent of network computing has opened new doors to learning around the world. Here at the University of El Salvador in Buenos Aires, Argentina, systems integrator Intersoft is helping to build a network of Java stations.

MR. RACCA: This was for us, the light of hope that we were waiting for. To start a journey of creativity and innovation and to not give in to the notion that all avenues have already been explored.

THE NARRATOR: The University of El Salvador was the ideal site for this web based information system, a system which students are helping to create.

MR. HIDALGO: Our big concern is that we are entering a time of new technology and we would like to be ahead and not behind.

THE NARRATOR: At the University of Chile, the AccessNova project featured on Sunergy will soon be housed in a state-of-the-art multimedia center reconstructed from a historic but earthquake damaged library.

MR. VERA: It's information where you access the content in various ways so you have books, you have also CDs, you have software, you have direct access to libraries elsewhere, so the concept is quite different from a conventional library.

THE NARRATOR: A partnership with the University of Chile and NTT in Japan, AccessNova is developing broadband multimedia applications. Dr. Vera sees this as a way to bring people together. This virtual plaza will become physical and students are inspired to connect with one another outside the center's walls.

MR. VERA: Once you make this communication bandwidth available massively, people are going to meet and that will create the need to meet face to face.

THE NARRATOR: Universities develop the Internet in China. China's first link to the Internet was created here at the Institute for High Energy Physics in Beijing. The scientists needed an e-mail link to colleagues in other countries.

MR. RONGSHENG: In the past, we put all the data on a page and shipped it by airplane. Because of the Internet, we can directly communicate with America and other parts of the world and send data and software to each other with easy access.

THE NARRATOR: In this room Professor Xu proudly shows us the first e-mail server in China. He is optimistic about the future of the Internet in China, but he is also cautious about the Net's initial impact.

MR. RONGSHENG: First of all, our culture is different from that of the west. This development needs a greater transition. We should not be too rapid, otherwise the results may not be good.

THE NARRATOR: But as China turns its gaze to the world following the return of Hong Kong, Chinese entrepreneurs are making great strides connecting the east with the west, and there is little doubt among them that as Chinese commerce develops it will be shaped by the technologies first put into practice in the Institute of High Energy Physics.

(End of videotape.)

JOHN GAGE: That Internet development from science then linking ultimately thousands of universities in China will make an enormous change in the ability of people to communicate. When people are linked, then that means they can begin to do something else, they can begin commerce. But they need a way to authenticate themselves. I'm back with Miko Matsumura and we are joined by Yuki Mitani, from the Smart Card Division of Schlumberger, Schlumberger Japan. It's a pleasure to have you with us.

YUKIHIRO MITANI: Nice to meet you.

JOHN GAGE: Now the devices you make, these smart little computers that everyone can have, and as we have seen in the story of China, in the story of Chile linked to Japan with very high speed links, people can now send things across these net connections, they can send money across these net connections.

YUKIHIRO MITANI: Sure, sure.

JOHN GAGE: They can link themselves but they have to authenticate themselves. You have here a number of smart cards. Could you describe what they are, what kinds of smart cards exist?

YUKIHIRO MITANI: Okay. Well, there are various types of smart card actually in the market. Well, those are the microprocessor card.

JOHN GAGE: I'll hold them and you tell me what they are.

YUKIHIRO MITANI: For example, this one is used for -- as a SIM Card, Subscriber Identification Module, for the mobile phone, like GSM, so every mobile phone has the smart card inside, except for Japan, unfortunately.

JOHN GAGE: Oh, my.

YUKIHIRO MITANI: Well, but it will be coming soon.

JOHN GAGE: Talk to Super, he will fix that.

YUKIHIRO MITANI: Oh, yes. That's a good idea.

JOHN GAGE: What is the memory, what is in this card? This is a memory card alone or this is a processor and memory card?

YUKIHIRO MITANI: There is a microprocessor inside and the maximum size will be now 8 kilobyte or 16 kilobyte memory, to compare with the supercomputers, it's very tiny.

JOHN GAGE: Yes. But you can fit a Java virtual machine into that space.

YUKIHIRO MITANI: Oh, yes, yes.

JOHN GAGE: When Schlumberger began the Java Card Forum, Schlumberger developed the Java virtual machine and now with all the smart card vendors are establishing this as a standard. You were going to describe the other cards you have. Now here is a second one, it says Smart Village.

YUKIHIRO MITANI: This is a brand new card, still not on the market but prototype, this has contact card as well as the address, inside there is antenna, so there is two interfaces.

JOHN GAGE: So this is a radio smart card.

YUKIHIRO MITANI: Exactly.

JOHN GAGE: Wireless smart card.

YUKIHIRO MITANI: We call it the Combi-Card.

JOHN GAGE: Combi-Card, so I could pass by a receiver and it can interact.

YUKIHIRO MITANI: Yes.

JOHN GAGE: Is there encryption in this card?

YUKIHIRO MITANI: Oh, yes, sure. This card can have arithmetic co-processors for encryption algorithm.

JOHN GAGE: I'm sorry, which algorithm?

YUKIHIRO MITANI: Encryption.

JOHN GAGE: A variety of encryption algorithms.

YUKIHIRO MITANI: For example public key or secret key, whatever.

JOHN GAGE: That gives us the ability to use this to identify yourself. The applications for this should be very, very large. Now what is this card?

YUKIHIRO MITANI: Yes. This is the world's first Java card. Java developed.

JOHN GAGE: So Schlumberger developed this card and this runs Java.

YUKIHIRO MITANI: Yes, it is.

JOHN GAGE: Right now. I can see it running. No, I'm joking. So the size of the virtual machine is 4 kilobytes?

YUKIHIRO MITANI: Yes, only 4 kilobytes inside. It's working.

JOHN GAGE: It works.

YUKIHIRO MITANI: It works.

JOHN GAGE: When will we see these as commercial products? Can people buy these now?

YUKIHIRO MITANI: Yes. We are releasing the pre-release developer's kit, so still not commercial yet, but soon. Hopefully from October this year.

JOHN GAGE: From October this year. Would they be able to find this at the Schlumberger web site?

YUKIHIRO MITANI: Oh, yes, yes, possible.

JOHN GAGE: Good. We will show this in a minute.

YUKIHIRO MITANI: Okay.

JOHN GAGE: So what market do you think these cards address? There are some cards that are memory only, there are some cards that are memory and processor, the new ones have radio.

YUKIHIRO MITANI: Yes.

JOHN GAGE: What range does the radio --

YUKIHIRO MITANI: Depends on the frequency, but most of the applications needs the so-called close distance, such as within ten centimeters.

JOHN GAGE: Ten centimeters.

YUKIHIRO MITANI: Something like that.

JOHN GAGE: Well, that's close enough if you are going past an automobile toll or in a retail store.

YUKIHIRO MITANI: For automobile that is too far but for the ticket, railway ticket gate, that is sufficient.

JOHN GAGE: That is good.

YUKIHIRO MITANI: For application, I think.

JOHN GAGE: What volume do you think these cards? What is the volume today of the smart card?

YUKIHIRO MITANI: Smart card? Total market is about 600 million per year, I mean annual demand is 600 million. And about 20 to 25 percent are for the microprocessor card. And actually it is getting to be popular and popular for various applications now.

JOHN GAGE: So 150 million microprocessor based smart cards sell today, and this is before Java applications are running in them and allow them to do new and more powerful things.

Miko, you described the development environment before, the range for smart cards through the most powerful machines. Are there any tools that are now available, Schlumberger will have a development kit, are any of the other API developments in dealing with smart cards?

MIKO MATSUMURA: We have a very extensive smart card API that's public and available on the JavaSoft.com web page. And our commerce API deals directly with these card APIs and allow people to very flexibly address card applications. What's very interesting is that it changes a little bit with the nature of the cards, in the sense that you are running a number of applications on top of the card so the Java card actually becomes a platform, so Schlumberger is actually a compute platform vendor, whereas people like Visa and Master Card are actually software application providers. And it's a completely different way of looking at that, that market.

JOHN GAGE: Now today's smart cards will be in vertical markets. There will be a card for telephone, or a card for train or subway. If you are able to move Java applications on to these cards you can reprogram them. Is this a plan of you Schlumberger's? Is this a market opportunity?

YUKIHIRO MITANI: We are doing that because thanks to the sophisticated concept of Java, multi-platform capability, which enables us to build the different applications separately but combine into the Java card. So that means multi-application card.

JOHN GAGE: So you could sell to banks, but now would the banks develop the software to run on the card? Could they?

YUKIHIRO MITANI: Oh, yes. Well, if someone who has the Java language capability, everybody can write the software for every application. Banking, telecommunication or customer ID, royalty, whatever.

JOHN GAGE: Let me see the architecture because you had a diagram that showed how this worked. Here I can hold this so the camera can see this.

YUKIHIRO MITANI: This bottom part here I mentioned as the smart card primitive is the conventional smart card structure.

JOHN GAGE: The dark green there is the processor.

YUKIHIRO MITANI: And then here the operating system, this is a chip dependent operating system. And we put or develop the Java interpreter, actually it's a Java virtual machine, we named it Solo, but this enables to interpret the Java applet which is written in each script's own area, can be downloaded or simultaneously interpreted into the operating system. So while those applications are all written in Java code.

JOHN GAGE: So the immediate application -- is there room, do you think, on your smart card in the future for processors that execute byte codes directly? Because of the limited number, one would think given the size that you would not have all -- you probably would translate byte codes so that you would not have a complete vocabulary of byte codes but a subset, and then reoptimize your code to fit into the processor.

YUKIHIRO MITANI: Might be possible, yes. If the capacity of the microprocessor and also the -- well, the capacity of the memory resources are increased, then that is realistic, I think.

JOHN GAGE: So in 1997 a high end smart card could have 8 kilobytes or 16 kilobytes. What will be the state of smart cards two or three years from now? Will they continue to increase in memory?

YUKIHIRO MITANI: All the chip manufacturers for smart card are now making effort to increase that capacity, so in coming couple of years, for example, the 32-bit RISC machine will be in smart cards. That is possible, I think. Also the memory capacity will be ten times or 20 times as much as now. So then I go back to the beginning, but multi-application card will be, how do you say, realized.

JOHN GAGE: Well, it seems close. Because the drive with all the applications you see, every place you go, the thought that you could run some of them on something like, where did we put it, here we have lost -- this is the problem with a thin client, you can never find it at the moment you need it. So to put something on this device with a 32-bit processor and with a serious amount of memory is a future target for many developers.

Do you hear a lot of developers talking about smart cards as a future for their applications?

MIKO MATSUMURA: Without a doubt. And I would be excited if I were developing to the smart card API. Simply put is that if a bank provides its customers with the smart card that's compliant to the Java card API, that means that I can actually develop to the same API and deploy an application on that bank card, which is exciting. So, for example, an extremely trivial application would be a card that identifies my friends, and then you could actually have a little algorithm, a little program that hops around and actually rests on your friend's cards but not on other people's cards, so I could write that myself using the JDK and I could just code my own little application that lives on smart cards. So as you can see, a developer should get very excited about this prospect of having yet another exciting platform to develop for using the Java programming language.

JOHN GAGE: Now, I do know that Visa, I think in cooperation with Schlumberger, this is a Visa card, there is real money on this Visa card, real dollars. And if I place the Visa card inside this reader, this viewer, as I slide it in, it makes contact, I don't need to with the radio cards, the radio cards could be read without making contact, I slide it in, and I press this button and we can see that it has, can you tell me, Yuki, how much money is on this card? $10 I think.

YUKIHIRO MITANI: About, yes.

JOHN GAGE: I will press it again. $10, yes it is $10. United States dollars. Now it would be possible to place this on the network. Does the Java card application enable encrypted authentication so each side could verify amounts? Then they would be able to transfer money across the net.

YUKIHIRO MITANI: Sure. The future of authentication is really important in the financial transactions. So, well now the Java card or the other smart cards as well, but the encryption algorithm is available to use for this. So why not?

JOHN GAGE: Real money.

YUKIHIRO MITANI: Yes.

JOHN GAGE: So in the future, do you think that the 600 million smart card number today will double or triple, or how rapidly do you think this will go up?

YUKIHIRO MITANI: Actually, it's about 20 percent annual growth in '96. So, well, within a couple of years it can be double or more.

JOHN GAGE: I found this device, a compact smart card -- there is one.

YUKIHIRO MITANI: Yes, this one.

JOHN GAGE: So a PCMCIA card to go in any machine, and you slide the card in. I will slide a card in, and suddenly this computer is interacting -- go ahead.

YUKIHIRO MITANI: There you are.

JOHN GAGE: What applications have you been experimenting with using this kind of device?

YUKIHIRO MITANI: Well, for such a card readers can be implemented into the personal computers. So, for example, the network access purpose to Internet or intranet as well, well, it can be used. And also the, how do you say, the purchasing through the Internet, so in that case you have to use the credit card, for example. Such a reader/writer is mandatory.

JOHN GAGE: To be able to do that.

YUKIHIRO MITANI: To transfer your data.

JOHN GAGE: Now the SET protocols for credit card are well established, people all agree with them. With the smart card I could implement SET protocols on this computer, on this smart card.

YUKIHIRO MITANI: Not really yet.

JOHN GAGE: Not yet.

YUKIHIRO MITANI: Because of the capacity limitation.

JOHN GAGE: They are not big enough yet.

YUKIHIRO MITANI: Not yet.

JOHN GAGE: I see. So we have to merge protocols when the computers become big enough, then we can do all these transactions at one time.

Well, Miko, what do you think Java's future is? We have ranged from chips, we've ranged from PDAs, hand-held devices down to a device like this. What do you think, what's your projection? What should a software developer think of? What should they do tomorrow if they are going to build something for this new world?

MIKO MATSUMURA: Well, I would tell them to dream big dreams. Because simply put, Java has fundamentally created a global platform that scales across devices. What that means is instead of being in the industrial age you are in an age where the rigid behavior of machines no longer stays rigid. Suddenly the machines become client in their behavior, in fact that means machine behavior is distributable, but beyond that now, this particular type of smart card enables value itself to travel across the network. So you can picture me with my web camera standing in Namibia saying I would like to build 300 houses for these homeless people here. I want everyone to stick their card in now and give me the money to do that. And I can broadcast that across the network, get the value into that town, and build those houses.

JOHN GAGE: I can see that every person watching Sunergy at this moment is thinking we could use a television and the Internet together and smart cards.

MIKO MATSUMURA: Instantly gathering value.

JOHN GAGE: This is a new world. This really does change things. Thank you both.

Now I have a book list that I want to show people. I have two books on this broadcast to recommend to people.

The first book I would like to recommend is a book entitled Factor Four. Doubling Wealth, Halving, cutting in half, Resource Use. The author is a man named Weizsacker, hard to pronounce, Weizsacker, and Amory Lovins. This book describes the new automobiles, whose weight is half, whose mileage, whose ability to use gasoline, whose efficiency is four or five times current automobiles, automobiles being built today by Toyota and Honda and General Motors. He makes the point that in today's industrial societies we can become much more efficient using intelligent devices.

My second book, I should tell you this is published by, it's the New Report to the Club of Rome, a group very concerned with global resource use. And Lovins has 50 different engineering examples of ways to be efficient.

And my second book, written by a Spanish sociologist, Manuel Castells, is called The Rise of the Network Society. A powerful book about how to think about networks.

With that, I want to thank everyone for joining us on Sunergy from Tokyo. Be sure to look at our web page at www.Sun.com for the schedule for the next Sunergy. Thank you very much for joining us.

RealTime Reporters

831.335.7792

dhoyman@RTReporters.com

http://www.RTReporters.com/