Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS


Main NerdTV Page NerdTV Archives
send feedback rss/podcasts I, Cringely Sponsored by Datapipe Creative Commons License

NerdTV #13: Judy Estrin

Bob: Welcome to NerdTV. Gosh, it's - you know you are only the second woman that we have so far broadcast on our show. And after the Anina experience, you know, you're it.

Judy: So, I'm either going to be the last woman that you film or you're going to continue -

Bob: We've got another one in the can.

Judy: Okay.

Bob: But we're going to pull out certainly. But I'm really thrilled to be here and I realized driving in here, we're at the offices of Packet Design, which is your company in Palo Alto and this used to be Xerox PARC.

Judy: Right. This is a part, Xerox PARC is actually up the hill and it once occupied these buildings also and Xerox owns these buildings.

Bob: There's some real history here. What happened in these rooms? Do you know what they used these for?

Judy: No.

Bob: Orgies?

Judy: I don't think orgies but I think clearly that a lot of the great technology that came out of Xerox PARC, some of it may have happened here.

Bob: And it could still be resonating. Wow.

Judy: And when we were looking for space, one of the things we liked about this location was because Packet Design when it started was very much an R & D organization, a la Xerox PARC and so we purposely wanted to be located in this area.

Bob: Do you think of yourself as an engineer?

Judy: Yes. But I think of myself as a technologist. I mean I am an engineer. I'm an engineer by training. I have engineering degree. I think I would call myself a computer scientist more than an engineer because I don't have a classical electrical engineering training as much as computer science. But I think I so early in my career moved into the business aspects of it, I would say I never left being a technologist behind and have always been and still am really passionate about technology but I'm not a practicing engineer.

Bob: Okay.

Judy: And I'm not sure I'd call myself a nerd.

Bob: Well, what's your definition of nerd?

Judy: I knew you were going to ask that.

Bob: Do you know any?

Judy: I know lots of nerds. So, I think it is - I probably am not enough of a practicing engineer to put myself in the nerd category.

Bob: I think it's like a frame of reference. The nerd really sees things strictly from a technical frame of reference.

Judy: You know I think that when I think of a nerd, I think that a nerd would rather interact with computers or machines than people.

Bob: That's right.

Judy: And I think I'd rather interact with people than with machines, which is one of the reasons why I ended up moving into management and the business aspects but as I said, I'm still very much and have always been a technologist. And so if all technologists are nerds, then I'm a nerd.

Bob: Okay. How did you get your start?

Judy: I was kind of born into it.

Bob: Are you from a family of -

Judy: Both my parents are PhD's in electrical engineering. My father worked with Von Neumann at the Institute for Advanced Studies; was one of the founding professors of the computer science department at UCLA. My mother was a PhD in electrical engineering that was in biomedical engineering and so I grew up in a family that was consumed with engineering.

Bob: You have a sister who is in too.

Judy: I have a younger sister who is a professor of computer science at UCLA, running something called the Center for Embedded Network Sensing. And I have an older sister who is an MD. So, the joke in my family is which one of us is the black sheep. My older sister is the only one not in computers. I'm the only one who doesn't have a PhD. And so we go back and forth as to which one of us disappointed our parents.

Bob: Who's the one?

Judy: I would say I disappointed them because I don't have a PhD. She would say she disappointed them because she's not in the computer science but I also growing up was one of those kids that was so clearly math biased. When you took all those tests, the difference between how I ranked on the English and verbal skills versus the mathematical and mechanical was a huge gap and I was just always very mechanically and mathematically oriented. If computers had not been invented, I might have been a statistician or something. I don't really think so but I immediately loved computers.

I went to UCLA and you know from the minute I took my first programming class I was interested in software engineering and computer science. I did undergrad at UCLA and at that time was when some of the early ARPA work was getting done and then I did my masters -

Bob: Len Kleinrock?

Judy: Len Kleinrock, yeah. My father - Paul Baran, was a student of my father's. The other person who was a student of my fathers was Vint Cerf, who a lot of people talk about as one of the fathers of the Internet. Vint went on to Stanford and I went on and did my masters with Vint at Stanford. So, I at Stanford was involved with the very early development of TCP/IP and I was kind of the junior member of that group.

So, I used to get to go in at three o'clock in the morning and test the TCP protocols with the people at the University of College London. That was my contribution to the TCP protocol development. So, I was involved in the Internet from the very, very early days. That was 1976 when I was at Stanford.

Bob: I want to jump back for just a second. I didn't realize this, you literally grew up in this world, so you were a little kid presumably meeting some of these people. What do you remember about Von Neumann, no one ever talks about Von Neumann. Did he smoke Marlboro's?

Judy: Actually I didn't know Von Neumann because by the time I was born my parents had left the Institute of Advanced Study.

Bob: Sure.

Judy: But a lot of the people who are the pioneers of the computer industry, yeah, they were probably about my parents' houses for parties and I probably serving them platters of the food. I don't know that I can remember specifics.

Bob: You don't remember juggling with Claude Shannon?

Judy: No but I essentially grew up in that environment. I can remember my father bringing home tapes about FORTRAN when I was in high school and making me sit down and watch these tapes. Little did I know then that I was actually going to become interested in it as a field.

Bob: Now our last show before this one was with Bob Kahn and you mentioned being involved in the development of TCP/IP and of course he and Vint were right there doing it as well.

Judy: Right. So I was part of Vint's team at Stanford. There was a group that were - Carl Sunshine who was one of the primary people doing his PhD with Vint on the TCP protocols. Yogan Dalal, there was a whole group of people.

Bob: Yogan, what a great guy.

Judy: Yeah, yeah and so I was part of that team while I was doing my masters. And then when I left there, I went to a little company called Zilog.

Bob: Not so little.

Judy: It was very little then - 50 people.

Bob: Really.

Judy: Yeah. So that was, I got out of school. Had offers from Xerox and HP and Intel and then there was this little start up called Zilog.

Bob: Why Zilog?

Judy: Because a friend of my father's told me the smartest people he knew were at that company. And I decided to give it a shot and that changed my life because I ended up going to a start up and remember this was like in 76, when start ups were not a dime a dozen as they are in the Valley today. But because I joined such a small company, it gave me opportunities and exposures to different things in a way that I wouldn't have if I had gone to a larger company and my career probably would have gone in a different direction.

Bob: When we think of Zilog it is primarily for the Z-80 Processor, which was the big mother there for awhile.

Judy: Right. Right.

Bob: What were you doing?

Judy: So, when I first went, the Z80 had just came out and my first job was to work with the chip architects who were working on the Z8000 and the Z8 and my job was to look at the instructions from a software person, from a programmer's perspective. So what did you want that instruction set to look like so it would be easy to program with that chip and I worked with the architects of the 8000 and then the 8 from that perspective.

Then I went on to do some compiler and assembler, developing the compilers and the assemblers but a couple of years in Zilog also had a systems business. And we ended up starting a program called Aerial, which was its code name at the time. The product was Z-Net and it was actually the first commercial local area network to be offered onto the market. And what it came from was Zilog had all these microprocessors. The systems group had the early notion of let's build systems around these microprocessors and let's tie them together and by tying them together, we can build a distributed system.

It was way ahead of its time. Very early in the cycle. Today no one would think of a computer environment that doesn't have PCs connected but this, again, was in the kind of the mid-70s. And one of the early -

Bob: How tightly integrated - were you doing remote procedure calls to other processors or?

Judy: You know it was interesting because the operating system, this was before Windows, before even CP/M and Zilog had an operating system called REO, which was a real time operating system that actually allowed you to send messages back and forth between the different systems. We had a system called the MCZ, which had a big shared - one system was a shared disc and microprocessor based systems that accessed it.

So, I was responsible for the development of Aerial, which became Z-Net. And it was an interesting experience for me because we didn't sell any. And we didn't sell any because there was a complete mismatch between the marketing and sales organization and what the customers of Zilog wanted and what we were doing as a technology group. We were having lots of fun developing really advanced technology. Then it came time to sell it and the marketing and sales organization didn't have a clue what to do with it. And that for me was when I said to myself, what I don't want to build things that people aren't going to use. And it really made a shift in my career in saying I want to somehow be involved in the whole process so that if I'm going to build something, it will be something that somebody wants to use.

Zilog then spun off or spun out of Ungerman-Bass, which was one of the first -

Bob: Charlie Bass.

Judy: Charlie Bass and Ralph Ungerman and one of the first of the LAN startups with Net1 and I ultimately left Zilog a couple of years later with Bill Carrico and we started Bridge Communications. So, about two years after Ungerman-Bass started, we started Bridge, which was in the local area network arena but what a lot of people don't know is that Bridge shipped the first commercial router. So Cisco takes claim to shipping the first commercial router. Actually at Bridge we shipped a router before Cisco actually existed.

Bob: So, tell me how the idea to do a bridge or a router came about; I mean the network adapters had just sort of appeared.

Judy: Right.

Bob: There was Ethernet and there was Arcnet and there were a variety - like your ZNet and there were -

Judy: Right, right.

Bob: How do you decide what to support?

Judy: I can actually remember when Bill and I decided that we were going to start a company; Ungerman-Bass had started a couple of years earlier. We stayed at Zilog for a couple of years brought ZNet to market. Were unhappy for whatever reason and when were starting to think about okay, let's start a company, it came from there were a number of different networks out there. Ethernet wasn't really there yet. There was something called Net1 from Ungerman-Bass. There was ArcNet. There was X25. There were broadband networks like from Wang and so the original idea behind Bridge was let's build something that can interconnect all of these networks together.

And that's why the name Bridge because we were bridging between different technologies. Well we started to write our business plan and we suddenly realized there weren't enough networks out there to interconnect. We wouldn't be able to sell enough of these boxes. And so what we did first was develop products that were these communication servers that actually allowed us, that competed with Ungerman-Bass, to get more networks installed and then we also sold internetworking products that connected those networks to X25 and later to broadband and other technologies.

Bob: So you had the ambition before the geography was really set.

Judy: Right.

Bob: To do what you really wanted to do.

Judy: It was very early.

Bob: But you saw that the opportunity would eventually appear, is that the way it worked?

Judy: Yeah, I think we really believed it was going to happen. And I think actually probably a hallmark of the companies that we started was we were always a little ahead of our time. Sometimes too much ahead of our times, sometimes the right amount ahead of the time and at Bridge we saw that that would happen or believed in it. When we were raising money for Bridge, I remember we almost didn't close the funding because somebody came out with a market research report that said that Ethernet was dead; that Wang was going to win over Xerox and so Ethernet was not going to happen. It was going to be broadband. Our business plan had a lot to do with Ethernet and the people who were about to close the funding, got nervous. We went back up there and we re-convinced them that Ethernet was not going to not happen. Closed the funding and -

Bob: The rest is history.

Judy: Yeah.

Bob: Wow. Now I always thought of Bridge Communications as doing bridges.

Judy: Right.

Bob: And there's a difference between a bridge and a router.

Judy: Right.

Bob: But you did routers as well.

Judy: Yes, we did communication servers, bridges, routers and gateways.

Bob: Did you do brouters? Remember brouters?

Judy: You know what, I think after Bridge merged with 3Com, I think 3Com used that term. But at Bridge we didn't called them brouters.

Bob: Good.

Judy: I don't think I'd call something a brouter.

Bob: At Bridge, what were the technical challenges that you had to overcome?

Judy: The technical challenges in the early days were about no one had ever built networks of this size but some of the more interesting ones were the applications we were selling into. PCs weren't around. So, as we were first installing these networks, we were replacing terminal switches with LANs, with communication servers and nobody believed that you could have the echo response time on a local area network that you could with a switch. They simply didn't believe that a multi-access network like Ethernet was predictable enough to give you the echo response time. So, that if you typed on the terminal and the computer echoed that character that it would be fast enough.

And so the delay in performance characteristics of implementing a robust protocol stack and being able to support a lot of terminals on the network and have it reliable and cost effective were the real technical challenges. They weren't about ASIC development or making something really small because in the very early days, it was just about making it work.

Bob: Sure and was that a problem, a true technical problem, or a problem with perception?

Judy: Both. It was a true technical problem. It was implementing a fairly complex system in terms of the protocol stack and the managability that you needed to have a distributed system work like that and reliably. So from a technical development perspective, it was significant. Nobody had ever done it before. There wasn't something to copy but it was a huge perception issue and every - I was VP of Engineering when we started Bridge. There wasn't a sales call that I wasn't on and a deal that I didn't close because nobody believed it and you just had to go out there and be able to convince people that it could work.

Bob: Now Bridge went public.

Judy: Bridge went public. So we started in 81. We went public in 85. And it was very successful public offering and then we merged with 3Com in 87.

Bob: How did that come about?

Judy: We were looking at the market; looking forward. Bridge was doing very well. But there was a shift going on in the market from terminals to PCs. We were certainly a leader in the internetworking space and in terminal connctivity. 3Com was the leader in the PC space and we decided we had two choices. Add the ability to attach PCs onto the network and compete with 3Com or we could look to merge with them and we felt that the right thing to do was to see if we could bring the two companies together and that's what we did.

Bob: Had you known Bob Metcalfe before then?

Judy: Yeah. I took - I'll tell you a funny story, which is I took a seminar from Bob Metcalf at Stanford and Bob and I laughed after it because my 21st birthday, I was at Stanford and a bunch of people had taken me out to lunch. It was the first time I could have a drink and we all went back to Bob's seminar afterwards and we're in the back of the room, having had a bit too much to drink at lunch. The only good news is Bob had been at lunch with us so it was the only reason he wasn't completely insulted that. So I remember that. So, Yogan Dalal, John Schock, a lot of the people who are names in networking today were all at that lunch. Yeah, I had known Bob for a while.

Bob: What a character.

Judy: Yes.

Bob: So, you merged with Bob's company. Of course, he wasn't the only person who was there. But it didn't go so well.

Judy: No. It went well financially. It went well in terms of the integration but the idea was that there would be an office of the president as these two - there was Bill Carrico and myself on one hand. There was Bill Krause and Bob Metcalfe on the other hand. Two fairly strong, opinionated management teams and we came together and didn't have the same opinion as to where the company was going to go.

Krause really wanted to turn the company more into a computer company. We wanted it to be a communications company and after about nine months, Bill Carrico and I just decided it was better rather than to have a big fight about it, it was better to leave and we did.

Bob: Did you have an idea of where you were going?

Judy: Our plan at the time was to take some time off. But the day it was in the paper that we were leaving 3Com, got a phone call and a group of five people who had been working on something called an X-Terminal, it was the first X-Terminal, a group of people had been building a prototype. Did not have financing. Came knocking on our door and said would you guys come join us?

And our first reaction is we need a vacation. And we got so excited about what they were doing and the team that we ended up joining as CEO and Executive Vice President and within a month had closed financing and was off and running with what was Network Computing Devices.

Bob: Wow.

Judy: NCD was exciting. We actually -- our first year of sales was something like $13 million -- grew very fast but that was before Windows. And X Terminals were very dependent on the notion of Unix as a server and when the Unix vendors couldn't get - I'll say- couldn't get their act together in terms of application compatibility across different Unixes became more divisive, allowed Windows to really get a foothold in terms of applications. That impacted the X Terminal market tremendously, which really hurt NCD.

And we tried to diversify, ended up buying an e-mail company. We had PC-based X Windows software and NCD also went public. It was doing reasonably well but in 1994, we just decided we needed a break and so at that point, I was CEO. We hired someone to replace me and left. And unfortunately NCD afterwards was not a real pretty story. So it ended up not being a success.

Bob: What was the lesson learned from that experience?

Judy: The lesson learned? There's a lot of industry lessons learned just in terms of the way the Unix vendors didn't work together in terms of applications compatibility; really hurt Unix at that time. There were lessons learned. It humbled me to go out and take a company public and sell stock at a certain price and then have it not be worth that a couple years later. It just really is a humbling experience. But I think one of the big lessons I learned is I think there were times when the company was successful when it could have been sold or partnered. Even after we left and the new management came into place, there clearly were times when it could have been bought and ego got in the way of that happening.

And it didn't and I think understanding what is the right time for a company. Sometimes it's when things are going really well and what we did very well at Bridge is when things were going very well, we looked out a couple of years and said where's the market going. Can we do this by ourselves? I don't think that was done so much at NCD and the other lesson is sometimes you can be too ahead of your time. And we were just thin clients before people were ready for thin clients.

Bob: I guess I hadn't thought of it that way but sure.

Judy: Right.

Bob: And of course we've had thin clients come and go a couple of times but now they are back I think with some gusto.

Judy: And later when I served on the Board at Sun and a couple of times Scott McNealy would get up and draw a picture of thin clients and one Board meeting, I just looked at him and I said I can't help it, Scott, where were you when I needed you? It was back when we were selling X Terminals and you know at that time, it was just too ahead of where the market was.

Bob: I guess but I think they've got landfills full of Sun Rays too. The early ones didn't sell. So, you left another company and took how long a vacation?

Judy: We actually took six months off.

Bob: Okay.

Judy: We took six months off and after six months, started a company called Precept. And Precept was one of the very early video streaming software companies. Actually I laugh now when I hear people say IPTV because it's become -

Bob: That was your product.

Judy: That was our product and Cisco bought Precept and I think gave up the trademark of IPTV. So, now it is commonly used but Precept was an early player in the video streaming market and relied on, we made an assumption that IP multicast was going to be widely deployed. So, if you look at other companies at that time, like Real, still exists today. They did lower quality video but didn't make any assumptions on bandwidth or any assumptions on network infrastructure.

We decided to look forward and do very, very high quality video but making the assumption that the infrastructure would support multicast so that you could take very high bandwidth transmissions and send it in multiple directions - to multiple locations at the same time. Unfortunately multicast, I tease people at Cisco that I actually read their white papers and believed them but they were just that. They were white papers talking about a future that took a very, very long time to develop. Even today multicast is not as deployed as you would like it to be.

So Precept had a great product. We sold into the enterprise market but just didn't take off the way it might have if multicast infrastructure had been deployed. But because of its reliance on multicast, we had a very good relationship with Cisco. And a couple of years into the company's history, Cisco came to us and asked to acquire. They had been an investor and then asked to acquire us. Part of that was for our technology, part of that was Ed Kozel was leaving and Cisco needed a CTO. So, they asked whether or not we would sell them Precept and if I would become Cisco's CTO.

Bob: Wow. Sounds like they bought the company to get you.

Judy: Partly. Partly. And -

Bob: Did you play hard to get?

Judy: I did what was right for my investors, which was to agree to go be CTO.

Bob: Let's look at the IPTV area a little bit because of course this is a show that's carried over the Internet. And we use MPEG-4 but we're talking a number of years ago and what were the emerging standards and what were the challenges beyond multicast versus unicast at that time?

Judy: So, the challenges with Precept were a couple fold. One is the standardization around what the codec was actually going to be; the compression technology was going to look like was just emerging. So today MPEG-4 is a pretty clear standard. There were a lot of different technologies that people were using. Probably one of the hardest challenges and one of the big successes and features that IPTV, Precept's IPTV, that other products couldn't was synchronization of the audio and video.

And so to be able to have a talking head and/or watch Lion King as a movie or whatever and have actual synchronization of the lips moving and the sound, to do that over an IP network was something that is not trivial to do. And the other difference is when we were starting Precept, PCs didn't have the power they have today. And so today you assume a certain level of power from a PC that can do that decompression. We were talking about fairly low powered PCs. And so to be able to get the compression and decompression and to do that efficiently was much harder in those days.

Bob: So was this all software or did you have a hardware component?

Judy: It was all software. And we did it on standard PCs.

Bob: So and you were doing like meetings and things, so it was real time.

Judy: Yeah, real time. Actually, I've got to say, it was a wonderful product. It was a great product. Because it relied on multicast, it was very bandwidth efficient. And from a video server you didn't have to support 14,000 streams. You supported a small number of streams that were then multicast over the network.

Bob: Sure. But because of the lack of support for multicast on the Internet, it became primarily a LAN product.

Judy: Right. And one of the things that slowed down the sales a little bit is even within enterprise networks; a lot of people had not turned on multicast because they didn't understand it. They didn't know how to diagnose it. There wasn't a reason for it and actually the reason Cisco bought Precept, besides my going into the CTO role, was because they wanted an application that would motivate their customers to deploy multicast and IPTV was that. And Cisco sold IPTV. I mean it became part of their product line and they were successful with it and I believe they still do today.

Bob: So now you are at Cisco and you are the CTO. This is different, isn't it?

Judy: Very different. And it was at the time was an amazing time to be Cisco's CTO.

Bob: When was this?

Judy: 98.

Bob: 98.

Judy: So, I was there from 98 to 2000. Cisco went from 18,000 to 36,000 employees during that time. That was the peak of the bubble, the peak of the frenzy. And I started out as CTO with a kind of funny CTO organization because at Cisco, legal and M&A, as well as kind of the office of the CTO were all reported into the CTO. And then a year in, I took over responsibility for all of the IOS software. So then it was all of the centralized software also reported to me. And then I left in April 2000.

Bob: So, was this a job that was even possible to do? I mean, you lasted 2 years so you must have done fine but there must have been some real challenges to it? And were there technical challenges, political challenges, business challenges, athlete's foot? I mean you know?

Judy: So, I think the challenge of being CTO at Cisco during that time was not technical. It was organizational. When I think of the role of a CTO, I like to think of what a CTO does is look forward and look across. And this was the peak of the bubble and you know what? Nobody wanted to do either because there was so much sales. Everything was about time to market. None of the business units cared about whether or not they were efficient working with another business unit. They just wanted to get their products out. So time to market was the most important thing and no one had time to look forward because it was growing so fast.

So during that time, the function of looking forward and looking across was tough. It was tough. Do I think I was effective? Yeah, to the extent that Cisco needed at the time. Now if you fast-forward two years, and as the market crashed, a lot of the things that needed to be done ended up getting done at Cisco in terms of looking across the business units, bringing more efficiency in. But when I was there, it was a fascinating but tough job.

Bob: Wow. Why did you leave?

Judy: I left because when I went I had committed to John that I would stay for a period of time and I said and after that we'll see you know how I like it. He knew that it was a risk going from having run my own company to being a CTO and I really felt I wanted to go back to running my own, to building my - I had never really had to exist in someone else's culture. I had always been able to build my own teams and my own cultures and I wanted to go back and run my own deal.

Bob: Wow. After being at a small company, a number of medium sized companies and then a small company in Precept and then you go to this monster thing that you have to get your arms around, I just, it must be hard to even know all the people who work for you.

Judy: Right. No that's true. That's true. But I must say, as I look back at it, it was often a very frustrating experience, but it was a great experience. So, for me to have kind of done that and experienced some of the large company issues and some of the trade offs of large companies was very beneficial to me both in terms of when I work with large companies, also in terms of one thing we haven't talked about is the parallel to my entrepreneurial career, I serve on large company Boards.

Bob: Yeah, how do you get those gigs? I want one of those.

Judy: You know what you don't get to apply. Somebody kind of finds you and it - and so both I serve on the FedEx Board and Disney Board, at one point I served on the Sun Board and so being at Cisco really taught me some things about what life was like in a big company and it's helped me a lot in my Board positions. And so, it was invaluable and very exciting experience. But by 2000, I was ready to go back to my roots.

Bob: Oh sure. Now we're at the part of the show where I have to ask the obligatory girl geeks questions, which are you were often the highest ranking woman in the organizations you were involved in, you were in key positions, was sexism ever a factor in your career?

Judy: You know, I used to say no. I used to just say no. There are point places where sure. But I have - you know and I have my cases where I know something was harder or that I had to prove something because I was a woman but for the most part, I think the industry being young, the fact that I had a lot of technical credibility, that I came up the technical ranks, I built my own companies, so you know if you build your own homes, there are no glass ceilings in it. I didn't climb up a corporate - so by the time I got to Cisco, I already had a career. And I think I just worked through it. I think my approach is if somebody is acting in a way that is sexist, I don't confront it. I find a way to work around it. And I find a way to make it work. So, it certainly has not stunted my career or hurt me in a way because I think it's a young industry.

But have I lived most of my career being one of the few, only or few women around? Yeah and kind of just got used to it.

Bob: Now what could the audience of this show learn, that they ought to learn, about working with women in technical positions? Have we been just screwing up in some horrible way that we should do something different?

Judy: Why are you assuming that there is not a lot of women?

Bob: There actually are a fair number of women but I think we've figured out it's about 15%.

Judy: Okay.

Bob: So 85% men.

Judy: You know I think and this may - this statement sometimes makes women mad, but I'll say it anyway. I think what's important to acknowledge is that women are different than men. So, what I have found is when you try to act like you're a man, that's when you usually get in trouble. And so it's a generalization but I think there are complimentary strengths and weaknesses. I think that one should never assume that another person can't do something because they are a man or a woman. It has nothing to do with just one. And one should never assume that another person should be given the opportunity to do something just because they are a man or a woman.

So I think the question is how do you work with someone? Take them at face value and try to figure out what they have to bring and what qualities they have to bring to the party and try to keep the ego out of it. And don't focus on what sex they are, again, man or woman. Focus on what they have to bring and what skills. I think in general, women I have worked with have an easier time being collaborative than men. I think it's something that as women are raised, as generally women tend to be less political than men. But that's not true about engineers because I would say that engineers are nerds and are not political to begin with. That is one of the wonderful things about them. So, that may not apply as much in the engineering rank and file.

Bob: Now, you left Cisco presumably took a vacation?

Judy: No.

Bob: No taking a vacation. But then you leapt into this thing that was going to be like this continual revolution.

Judy: Right.

Bob: Packet Design was going to be the cultural revolution.

Judy: Packet Design was going to be the ideal company so we didn't think we needed to take a vacation. And I will say now and I'll explain why it did not turn out to be but what we did at Packet Design was we wanted to focus on medium term innovation. Not long term research the universities do and things like Xerox PARC. The industry was very short term focused at the time with the Internet bubble and nobody could look past their nose.

People were painting over problems as opposed to really looking structurally at what was going on in the Internet and saying what do we really need to do? So one of the people who worked for me at Cisco was Van Jacobson who is one of the leading minds in the Internet and I believe probably one of the smartest people I have ever met. And he joined us and we built a team around him of researchers and developers. So we took people from the research world and people who knew how to develop products. Sprinkled a little marketing expertise in there because we didn't want to be ivory tower and started a set of projects that were looking at solving some of the real problems we thought existed in the Internet infrastructure, working very closely with service providers and enterprise customers so that we could really understand the problems as opposed to again being ivory tower.

And the idea was to fund those projects through prototype and then either they would fail and we would learn something from the failure. My younger sister when we started Packet Designs said you can't call it research unless you are willing to fail. And so, we knew some projects might no go anywhere and that those that succeeded through prototype and it looked like had a market, we would spin off companies to productize them so that the parent company, Packet Design, would not end up getting distracted by productization. And we started it in 2000. We raised $29 million. We spun off our first company in 2001, which is Vernier Networks, which exists today and then in end of 2003, I don't need to tell anybody in the IT industry what happened between 2000 and 2003. And we anticipated that the bubble might not last forever. No one could have anticipated how hard everything would crash and the corporate scandals and 9/11.

And it came to 2003 and I had a choice, to raise more money for the R & D part or to, as I say, say uncle on the business model and say you know what, this business model for this time in the telecommunications marketplace and we took our two remaining strongest projects, spun them out as two additional companies, Packet Designs, Inc. and Precision I/O and decided to stop any further R&D. We took all the people and divided them between the companies and I decided to focus my time on trying to make the three companies successful. So we're no longer a perpetual startup. We're not an R&D organization anymore and I'm really focused on, I call myself, I guess I'm a professional Board member. So, large company Boards and the small company Boards.

Bob: How are your companies doing?

Judy: Vernier is in the security space, shipping product, is doing reasonably well. Packet Designs in the route analytics space, created a new market, also doing well. Precision I/O is the baby and it has yet to ship a product. So, it's the youngest of them. I tend to refer to the companies in different stages of -

Bob: They are your children.

Judy: Yeah, they are kind of like my children. So, and as I mentioned to you earlier, Packet Design was a phenomenal idea in terms of producing technology and producing ideas. I'm really happy when I look back at what we've done and I look at the companies and the products that have been developed and the customers that we have. As a business model, it wasn't a success and because the venture environment has changed and it's harder to be successful as a Series A investor that then can't invest in further rounds, which Packet Designs did not. In terms of making a lot of money for my investors, still not clear and was not as successful. In terms of developing some technology and some great products, Packet Designs was great and continues to be.

Bob: So what excites you now?

Judy: It's interesting. When I think about what I'm most passionate now and although I'm very excited about the companies because they are like my children but it's not the area of the market that I'm most passionate about today because the enterprise, the networking market to me has reached a level of maturity and I tend to look a little bit more forward as you can tell.

Bob: So what do you see?

Judy: So I think the places that I get most excited about and when I look at spending my time, I spend a lot of time on Disney even though I'm a Board member and I realize one of the reasons I'm so excited about that is because to me the consumer market is really where so much of the excitement is today. And it's hard to be a start up in the consumer market. It's just a hard place because brand and critical mass and alliances are so important. And so my Board seat at Disney has been a way for me to participate in the convergence of technology in the entertainment market, which to me is clearly going on and a very, very exciting place to be.

I look beyond that another area and that I am involved in from a Board seat perspective that I think is very exciting is the whole area of sensor nets. And if you think about the evolution of the Internet, it started out connecting people to information, data, file transfer, e-mail. Then all of sudden we added commerce in to that equation, then we have added and are adding entertainment into that equation. One way to think about what sensor nets are all about is adding awareness into that equation. So suddenly you can bring information about the physical world into your information systems in a new way because you can suddenly have little devices that have sensors and wireless and processing on them.

And so I think when I look out there in terms of trends and areas that will bring a lot of really exciting new companies and new technologies, that's an area that I'm really excited about today.

Bob: So sensor nets, look two years out and give me an example of what we'll do with them.

Judy: You know I think it's everything from environmental - you know think about if whether it's the tsunami or the levee that broke, think about if you had early warnings to changes going on. Think about oil exploration and what can be done in terms of oil fields if you have sensors. Think about packages and the transportation industry and if you knew things about packages and where those packages were going. There's kind of unlimited different applications of - there are medical applications. There are home applications that if suddenly you had sensors in your home that were not just - that you actually could reduce that information and do something with it as opposed to just turning on and off a light.

So I think there is a very broad set and I think like most technologies, we don't even know yet because we haven't yet - the technology hasn't yet gotten to the place where people are using it enough and the most exciting applications are going to be the ones that your audience is going to figure out once that they see the technology is there.

Bob: What about the evolution of entertainment? What comes to my mind is the fact that so much of the value in these companies is in archival material, libraries, things that aren't three dimensional, aren't necessarily digitized and I mean, I Love Lucy will be around for a long time but we'll adapt it to new delivery mechanisms somehow. Is that part of what you see or is it something else that is going to enthrall us five years from now?

Judy: You know I think it's a lot of things. I think that there's the content that we know and love and have known and loved delivered in new ways. That is clearly one really interesting intersection of technology and entertainment. There is new forms of entertainment and so video games didn't really exist 10 years ago and today if you look at the number of people who spend their time being entertained with video games, there's a lot of exciting opportunities there. There are new types of content and you know, or new ways of making content. Think about the move from 2D animation to computer-generated animation to 3D. So, there's a continual evolution of how you make the content itself and how you make it more entertaining, more pleasing visually.

And then I think there's the other area that's going on of social networking and kind of the interaction of social networking and entertainment coming together that I think will provide a lot of really interesting opportunities. So, I think it's in the content itself, the delivery of the content and I think both of those are changing. So there's a matrix and you know, one of the reasons why and obviously I'm biased towards Disney, but one of the reasons I'm so excited about Disney is I think it's a company that gets it and I think that the companies that will win - Everybody is so nervous about piracy and you know TIVOs and all of the changes in technology, yeah. All of those things will change the industry. Those companies that will emerge out the other side, though, are the ones that know how to produce great content.

Because in the end, ten years from now, we and your kids will still going to want to be entertained and it still is going to take great content to entertain us. It's not just somebody with a video camera who doesn't put the time into telling a great story.

Bob: Well that would be me.

Judy: What? It took a lot of time to set this up. You're not just a guy with a video camera. So, it is -

Bob: Well, it's interesting what you say. Even if you look at NerdTV, which couldn't have existed even five years ago and it really allows, you mentioned social networking. It allows a sort of self-aggregation of an audience that is potentially very influential and certainly people will be watching this show for years, and years, and years and they will be thinking about some of the things that we've said here. And I'm excited about it. I just wish I knew a little better where it's going. I feel like one of those guys who is just trying to get the product out the door and I don't have your advantage of the ability to look across and down. So, if you have any advice for me, I'm eager.

Judy: But I think you are doing exactly what needs to be done. Meaning I think that when you are talking about consumers and consumer behaviors, it's iterative because you can't look out there and say, people are going to like XYZ. You can't ask people what they are going to like if they don't know what's possible. And so a lot of the really interesting things that have evolved have started out with someone having an idea and trying it and then realizing that this is something that really takes. Look at things like MySpace. That is something that becomes, that is iterative in terms of how it develops and to me the most exciting things about kind of looking at the technology industry has watched the times when someone has come out with a new technology, not anticipating - knowing it was good for one thing but never anticipating the breadth of things that it could be used for. But then it's the people who are adopting that technology who are continually coming up with new ways to use it and the technologists react by adapting the technology to it.

Bob: One of the things that amazes me is the resilience of some core technologies, like TCP/IP and Ethernet. Who would have thunk? I mean how did that come to be?

Judy: Well, I think it is because a couple of things. One is that if you have a good technology that has a good base and that gets the critical mass from a cost perspective, it will win as long as it can carry you forward until the next iteration of the technology. With both Ethernet and IP, there were decisions made early on that it was architected well. These were not quick and dirty things that were developed. And the people that developed the technologies thought architecturally and not just from an implementation perspective. So whether it is the 48-bit address of an Ethernet or whether it is the inherent architecture of IP and the layering and the separating out of the end to end TCP from the IP header, those are some basic decision that were made that were well thought out that then provided an architecture that the technology could evolve.

And what's happened with both IP and Ethernet, they are not exactly the same as they were but the developers of them thought about it enough that it could survive and it could evolve. And what makes me nervous about the bubble and the crash after the bubble and the lack of funding for R&D research today, what I see is a lack of real innovation in thinking is we don't develop like that anymore. And most products that are being developed are time to market, time to market. And we are riding off of the architectural work that was done 20 years ago. Plus that work is not being done in a lot of places today. And so it is a concern and I think one of the issues the industry has is when we need to get back to figuring out how to do the R in R&D so that fifty years from now we can have this conversation and be excited about a piece of technology.

Bob: I remember Metcalfe said that you know this was 70 - whenever it was 73, 74 and doing Ethernet, they needed a megabit per second and they can do that and then they thought well how much are we going to need for like the end of the 80s. And they said that would go to three, because you remember three megabit was the - and then they said well, you know what about another ten years. Then they thought ten! So they designed for like 20 years ahead. And we design now for like 20 months ahead and you're right. I think this is a terrible concern and even basic research is becoming more like that, despite your sister.

Judy: Right. Right. There are few pockets of places where people are doing real good research in the computer science industry. But there is a real problem in terms of R&D funding. A lot of the funding for research in universities, if it comes from DoD, it's classified. The other government funding agencies really kind of backed down on their funding of telecommunications as the Internet started to grow and people felt that industry should pick up their load. But industry hasn't so it's a problem today. It's a problem today.

Bob: Has the U.S. lost its lead? Is it in danger?

Judy: It hasn't lost its lead yet but it's in danger. Yes. Yes.

Bob: Is that something we should worry about?

Judy: Yes. Yeah and I know that one of the things that I certainly spend time when I talk to people is try to build awareness about this because I think it's one of the fundamental core competencies of the U.S. and has been innovation and R&D in terms of certainly the computer industry and I think we are at risk of losing that lead if we don't invest in it.

Bob: Judy, thank you very much.

Judy: You're welcome.

Be more connected. Help support your PBS station and projects like NerdTV...pledge online.

© 2005 NerdTV LLC and PBS. Offered under the Creative Commons license.

Privacy Policy | Terms of Use | Published on September 6, 2005