SW: My name is Susan Ayers Walker. I'm married to Rob Walker, the subject of our interview today. Rob has been interviewing semiconductor pioneers since 1995. Rob has been in the semiconductor industry since the early '60's, holding positions in the early days at Fairchild, Intel, and LSI Logic. Rob is a car enthusiast. He has a '57 Chevy, red of course. We live on a rural road in Atherton California, only 10 minutes from Stanford University. I've been after Rob for several years now to tell his story, so now it's his turn. Rob is also an eclectic art collector. Darling, tell us something about your art.
RW: We bought this place in '83 after the public offering and I love it. It's got all the trees outside and just gorgeous, gorgeous place. And as Susan indicated, I do like art. I had this piece done for the room and I call it "Going Public" and it's kind of soaring up to the sky and it's supposed to be a big bird. If you look at it in a certain way, you can probably see that's it's intended to be a big bird, like an eagle or something. And then up here we have Sue the T-Rex, no relation to my wife but the largest T-Rex ever found in North Dakota and it's named Sue for the woman that found him or her. This is my office in here and I have some family photographs up here. This is my mother and I and this is me and it's interesting because we come into the world and we go out of the world the same way. That is incontinent, bald, and fat. I have some more family photographs. This is my brother Jim who's California Highway Patrol and his K-9 Turbo, so you don't want to run into him, especially as he does drug interdiction. And this is my sister and her husband and her son and they're all three volunteer firemen and paramedics as well as their regular jobs as banking and so on and so forth. This is me and my '57 Chevy that you've seen before, couple patents. Here's the LSI founders, here I am in younger days, Wilf Corrigan, Bill Omera, Mick Bohn. My book, the share of LSI Logic stock. I went to the Bondurant School racing and here's Susan and I, this is our wedding picture from a year ago, and this is back from Northern California and here I am, I'm about I guess ten there and there's my brother the cop and my sister the firefighter and they were little enough then that I could really boss them around and now my brother is wearing his 9-millimeter, he was licensed to kill. He says, remember Bobby how you used to play king of the hill and throw me down and I'd say I don't remember that. Here's mom and dad before they got married and before they had us, looking very relaxed there and we changed that. And here's Susan from ten or fifteen years ago visiting out on the deck here.
SW: Rob, give us some of your personal history. Tell us about your early days.
RW: Well I was born in Hayward, California in 1935. Which is right across the bay from Silicon Valley and have a brother and a sister and we were a typical working class family. So I went to UC Berkeley because it was the closest, cheapest, school. And I had almost flunked out of high school Spanish, I am just terrible with language and so I went through the UC catalog looking for a major that did not require a foreign language and electrical engineering was the only major in this entire catalog of thousands of majors and so I took electrical engineering. So then I got into the courses and I got into my junior and senior year and you had to declare a specialty within electrical engineering. And most of the specialties, well they were either dull like power, there's was three phase, or they were very difficult like Fourier analysis with communications and things like that. But there was a new, they had just started a new specialty called digital, which is all ones and zeros and it was so easy so I took that. So the reason I'm a Berkeley EE majoring in logic design is because it was the closest, cheapest, easiest way to go. So then in '58 I got a job at Lawrence Rad Lab over in Livermore and in the lab there, it was interesting because you'd go in the lab and there would be nuclear weapons sitting there, there would be reentry bodies, there would be A-bombs and H-bombs, I mean this was the lab from hell. Unfortunately I did learn one lesson there that you need to have a major, you need to work for a company where the president of the company, has the same major as you had. Because the people that ran Lawrence were all PhD physicists and they thought that electrical engineers were, you know, get me the coffee, that was the extent of your expertise. I had one interesting example of that, I had just been out of school a year and they put me on a project with a numerically controlled three axis contouring lathe driven by a fifty horsepower electric motor. This is a big lathe designed for nuclear warheads as a matter of fact; machining the various portions of a nuclear weapon which tend to be hemispheres made of various materials, various metals. And this was to be built by Cincinnati Milicron who was at that time the world's preeminent machine tool manufacturer. So I went back to Cincinnati which is really a groovy town, a lot of action there, until about seven o'clock at night and then it shuts down. At any rate, this thing in action it was all driven by relays and stepping switches and vacuum tubes and it was clear that it could never run for more than a few minutes without making an error and any freshman student could run the numbers on the reliability of the various portions of this and see that the mean time between failure was in the order of seconds, if not minutes, and the plan was for this thing to machine these for an hour at a time, two hours at a time, to machine a huge hemisphere out of beryllium or something. Anyway, so I came back and I wrote this trip report and I said it can't work, it will never work, it's throwing taxpayer's money away. Well, that was squashed immediately and I was told I didn't know what I was doing and just this kid out of school the usual stuff and indeed they brought it in and they thought they would start off with aluminum because that's easy to machine. And they never really got passed that portion, it used paper tape as an input if you can imagine that "Mr. Reliable. " And if it read it wrong, if it didn't make a connection with these little pins that dropped down, it might try to dive into the material an inch and a fifty horsepower motor driving this huge hunk of aluminum, the tool drives in a inch and it's like a bomb goes off and there's just shrapnel all over. And they didn't kill anybody the first time, they built a steel room which they put it in. Of course it never worked and they ended up operating it with the machinist going like this. So I decided that wasn't for me so I went to Ford Aerospace in Palo Alto, in the meantime I got a master's degree at San Jose State. So I went to Ford Aerospace and worked on the manned spacecraft center down in Houston. So when you see those big rooms with all those people setting up those consoles and punching buttons and things, I worked on some of those things, designed some of them. And then we went off and we, there was a little R and D group there and there was a little company down the street called Fairchild and they had just developed the integrated circuit and the second generation of integrated circuits were called Diode Transistor Logic, or DTL and we were the first beta site for this. So we had prototype circuits, logic circuits, gates and flip-flops, and the supply voltage at that time wasn't even set. Now the MIL standard at the time was six volts. The first line of logic IC's were called micro logic and they were four volts and we in going through the analysis and looking at speed power decided that five volts would be the best and so we were the first person to ever specify a five volt supply. We had to get our power supplies custom built because they weren't available in five volts, and so we set the standard that was then for like twenty years the standard supply voltage for a logic even in MOS days, when MOS came to the forefront, was plus five volts. That was interesting. We also had the first line of integrated circuit modules and I've got one here and these plugged into a back plain and in the back of the back plain you could either use a big PC board, or you could use wire wrap and then inside this thing you would mount the integrated circuits and you could have various boards in here, these could be at various sizes as well and then the tops of these pins came out and were test points. And then you marked on here the, what it was. This case this is a power NAND gate, so this is like a dual quad gate, not very high levels of integration, but this was the first of the what were called logic modules, that's how DEC started. Except DEC, they used a minus three volt supply, they used germanium, and they used discrete IC so that we and DEC could have been Digital except we had Ford expertise and they managed to destroy the entire operation, but that was a first. So anyway that was, that was through the early days and then in the mid '60's I went to Fairchild.
SW: Rob, you got started at Fairchild in the early days, tell us about that time.
RW: Well, it's interesting, Fairchild of course is the mother, Fairchild and TI, are the mother of all of the semiconductor industry, which is now in the many billions of dollars, and with TI in Texas and Fairchild in Mountain View, California. And the story of how Fairchild got started is very interesting with Bob Noyce, Gordon Moore, and a number of other people and, as a matter of fact, it's covered fairly extensively in this Silicon Genesis series of interviews. I've got an interview with Gordon Moore, Richard Hodgeson, Harry Sello, and others that talk about those early days and Bill Shockley at Shockley Labs. What a jerk he was or brilliant jerk, his group couldn't stand him, they left, they went around, there were no venture capitalist in those days. They were finally able to get money from Fairchild Camera and Instrument located in Syosset, New York. And they formed Fairchild Semiconductor, which went on to become the mother of all of Silicon Valley, all the Silicon Valley companies can trace back through two or three degrees of separation back to the mother ship of Fairchild Camera and Instrument. So I went there in the early '60's to design the integrated circuits, digital integrated circuits.
SW: Rob, at Fairchild, what products were you involved with?
RW: Well there was first something called Medium Scale Integration, which is so-called MSI and consists of up to a hundred-gate complexity. And we defined those things like four bit counters and Arithmetic Logic Units and things of that nature and we did such a neat job of it that today, 1998, those functions are still available, same pin outs, same basic function as we came up with way back when. But my real love was Large Scale Integration which again in those days was anything over a hundred gates and so we started and did the seminal work in gate arrays and standard cell and computer-aided design. And just as Fairchild was the mother of semiconductors, so was Fairchild Micro-mosaic, Micro-matrix, the two custom capabilities, they were the mothers of today's Computer Aided Design and of the so-called ASIC business.
SW: So Rob, what lessons did you learn?
RW: Well, at least here, at least at Fairchild, the president of the company had a similar degree to mine and a similar expertise, so I learned that that was, I was finally on the right track there. But I learned much as I had at Ford, that you can have the best product in the world and if you don't market it correctly, or if you don't deliver, you can't manufacture it, and that was our problem. We had these, this fantastic products and couldn't manufacture them reliably and we reached a point where in the mid '70's, in the early '70's, where customers would not buy our proprietary better circuits, they instead went with industry standard multi-sourced products and I learned that better isn't always better. We had CMOS line of logic that was better than RCA had. And we managed to poison the whole CMOS market with this big battle between ourselves and RCA over which was better, buffered, non-buffered, oscillation, non-oscillation, to the degree that people, customers threw up their hands and said, I think I'll continue to use TTL. And then similarly with a line of logic called ECL we had this battle with Motorola with ads going back and forth and showing our units which were incompatible to that of Motorola's, ours were better, Motorola said their's are no good, they'll destroy your system, they're no good, and so we scared people away from ECL. So I managed to destroy two whole markets, the ECL and the CMOS market, well not destroy them but certainly to stunt them through coming up with something that's better. In the meantime, customers could use TTL, Schottky TTL, and those products without multiply sourced and without all this confusion over pinouts and speeds and power and so on. So much for my days of destruction.
SW: So Rob, you went to Intel in 1975, tell us what you did there.
RW: Well, it was clear that Fairchild was a sinking ship and indeed that did happen. It was finally taken over by Schlumberger that finally sunk it and it was absorbed by National Semiconductor and the buildings torn down and all traces of Fairchild removed. However, I have made some other of the videotapes that kind of chronicle those days, Harry Sello being one of the better, Gordon Moore, talking about Fairchild and what went wrong with it. So any rate, I wanted to go to Intel and it was clear to me that they were the, I knew Noyce and Moore, I knew they were good guys. I knew Andy Grove, I didn't think he was really, I knew he was kind of nuts, but he was really a hard worker and very smart. So I wanted to go there but in those days Intel did almost no logic, they were almost all a memory company. And so myself as a design engineer, there really wasn't a job for me there as a senior design engineer. But they offered me, a guy by the name of Jack Carsten had just been brought in from TI to head up marketing and he offered me a job being in charge of marketing communications and I had the advantage, it was an area that I knew nothing about all right, I was completely, never done it, never had a class in it, didn't know anything about it. But I did know that what Fairchild did was absolutely wrong in that regard. And I had written a great deal already, I had a number of articles. So I was a writer, and I knew what not to do but that was the extent of what I knew. So I went to Intel when it was about, it was doing about, a hundred and twenty million dollars a year when I started and I had one clerk that worked for me and that was my marketing communications empire which ended up five years later, I had a hundred people in three states and we were doing a pretty good job, I think. I think I completely covered, our largest costs were that of printing and distribution. In, as soon as microprocessors started to come out instead of a little data sheet, you needed a data book that weighed a pound or so and cost a dollar to mail. And maybe a dollar, a dollar and a half to print. And I was able to sell enough of those to pay for the printing and distribution costs, which turned out to be in the end a several million dollar a year business that was, we would break even, which allowed me to print these things and helped, I believe, Intel to become the standard that it is today in microprocessors. I remember we had a photograph showing the documentation, Motorola, Zilog and ourselves. And here's Zilog, and here's Motorola and here's Intel, I mean we had far more books than they did and better documentation which is really important in the computer world. So anyway I was there until 1980 and I thought did a good job, Andy Grove didn't think so, and so rather than be fired, I left in '80.
SW: I understand after Intel you became a consultant.
RW: Yes I was, that's what we called unemployed in Silicon Valley, we call ourselves consultants and I did a variety of jobs. And by that time I wasn't doing engineering, I had been out of engineering for five years, but I was helping people with promotions and public relations and that sort of thing. And one day a fellow by the name of Bob Hartman called me and Bob had started a consulting business of his own in gate arrays, which are a form of Application Specific Integrated Circuits. And he knew that, A. I knew something about those because we had essentially invented them at Fairchild and that I knew how to do promotions and public relations and that sort of thing. So he came to me and he said, "I don't have any money, I work out of my bedroom at home, how do I get to be known as a consultant in ASICs? " I said, well look, write an article that will give you some publicity and you can then reprint that and that in effect becomes your brochure so that when you go in say look I'm the world's authority on this and here's my article to prove that. Well, after we talked about it, it became apparent that Bob didn't know that much about anything but Signetics' products, which he had worked at Signetics and he didn't know what Fairchild had and what TI had and what all the other people had. And so I said well look, this is really kind of interesting to me now, I used to do this. I'll tell you what, let's change the deal, we'll co-author an article, a survey article on gate arrays, I talked to my buddy, Gold is his last name, I forget his first name, at, I think he's still at Electronic Design and got a commitment out of him to publish the article. And so I started calling around to everybody in the business and I said, " I'm writing this survey article for Electronic Design on gate arrays and can you tell me all about your product?" So I went and visited everybody in the business and they told me here's the computer aides that we use and here's our technology and on and on and on, our business model. And indeed, Bob Hartman and I did write the article which was published in Electronic Design and along the way I got back up to speed on this. I knew what everybody had what everybody was doing and coincidently, Wilf Corrigan who had been CEO of Fairchild in the Schlumberger takeover, he was forced out, and he didn't know what to do with himself, he just, he sort of tried to play for a few months, gave up on that, and his wife told him, "You're a one horse pony and all you know is the semiconductor business, that's what you like so you need to get back into that." So he started scheming about how he was going to get back into that. And he talked to, I plan to have an interview with Wilf here in a month or so covering his experiences. But any rate he talked to a lot of the major customers which at that time were the main framed computer people, the Burroughs, the UNIVACs, the CDCs and he asked them what they needed and they said we don't have a reliable supplier of gate arrays, there are people like Motorola and Fairchild that have good technology, but they're not really interested in the business. The business is for them a big pain. And there's little suppliers like AMI, primarily in MOS, and they're flakes. So you got the competent ones are not interested in that sort of business, the small guys are flakes, we really need a competent supplier of gate arrays.
SW: Rob, so continue your story with Wilf Corrigan, I understand you were one of the founders of LSI Logic.
RW: Yeah. I hadn't done engineering for five years, but I really could, I knew what everybody was doing and I told Wilf, we can blow away these people in just a couple of years and these people being General Electric, IBM, Hughes Aircraft, United Technologies, Motorola, in other words, Fortune five hundred kind of companies and our little start up we could do better. So he offered me the job as VP of Engineering and sure enough we blew away those companies. Today those that I mentioned, none of those people are in the gate array business or in fact the ASIC business, or in most cases not even the integrated circuit business or semiconductor. And it just showed we were one of the very first to show that a small fast-moving start up company what it could do.
SW: In the early days of LSI Logic, tell us about some of your recruiting.
RW: Well, I went out, I had to get the best people because we didn't have many. And I made it a practice, even after we became a fairly big company, to ask questions, detailed technical questions of everybody, whether they were a new grad or somebody with a lot of experience. And so I had little questions I would ask them, technical questions, which sometimes I even did over the phone and in five minutes or so I could get a feeling of the technical competency of these people. It was amazing, there were PhDs that came and had fantastic resumes that I found were utterly incompetent and there were people who were not degreed, didn't know anything, supposedly didn't know anything, but in fact were really excellent. So I recruited on the basis of, strictly on the basis of what would be best for the company with no regard to gender or age or race and as a result I ended up with, at the time, I think unquestionably the best engineering group of its type in the world and we took on the big Japanese companies as well, and from an engineering standpoint we blew them away as well; companies like NEC and Toshiba, Fujitsu.
SW: Can you talk about some of the people you recruited?
RW: Well let's see, Jim Koford and Ed Jones, Ven Lee, Dan Wong, Tony Wong, Jenson Wang, Steve Chan, I can't think of all of them. Most of the people I have named today in '98 have their own companies and have moved on and they have become very, very successful and I think carried on many of the principles that I espoused; openness, communication, the classic things that you read about in all the business textbooks, but are almost never, in fact done. In fact, I used to teach, on a regular basis, management classes for my engineers. Because in engineering, if you become a good engineer, they say congratulations, we're going to make you a manager, you've never had a course in it, no knowledge at all. And as a result you sort of do what your boss does, you use him as kind of your example. And so I actually introduced them to the concept of why meetings are so expensive and should be conducted in just a certain way and why everybody needed to arrive on time and how you had to be honest in your reviews even with not good news, nobody likes to give bad reviews. So I taught a whole generation of managers how to do it right and believe me, having come from Fairchild, I knew how to do it wrong and at Intel I had learned pretty much how to do it right.
SW: You told us something about your hiring and your personnel principles, can you tell us about your engineering principles?
RW: Oh yeah. Well it's sort of back to basics with all of that. Worst case design, no trick circuits. I recall one line of, early in our LSI career we licensed a line of gate arrays from Bob Lipp and we built them the same way and however once I went through the specifications I specified the delays in the worst case basis while Lipp had used typicals. Ours appeared to be half the speed or twice as slow as Lipp's, and yet they were the same basic silicon and just because of using a worst case analysis rather than typical, ours always worked, his didn't always work. And in fact reworking them to try to get them to make the specs, really sunk their company.
SW: I think this was the infamous 3-2-3 gate arrays in California Devices.
RW: That's where we met, you were at Wang and you were our first major customer because Wang and it had also licensed this design and that's where you and I met. I should tell the viewers that you're an engineer as well, an electrical engineer, and so that was I think in '81 and Wang became our first major customer. Any rate, so no trick circuits, worst case design.
SW: I remember test tape before fab.
RW: Oh yes, that was another thing that I initiated. That was one had to have the production test tape in hand before we went and actually fabricated the prototypes. And we were unique in doing that, a lot of the customers were unhappy with it, I stuck to my guns and we were able to go directly into production, our competitor, VLSI, who didn't espouse that, they got jammed up when it came time for production, they didn't have test tapes, the customer engineers that were to do it, they had gone on to other projects, other companies, and they got really, really jammed up with that so that worked out.
SW: Tell us about the events and milestones leading up to LSI Logic going public. That was an exciting time for you.
RW: Well it really was. We started out, we had nothing and we had to recruit the people, develop the technology, and get the customers. And so we did the first two and then it was time to book our first customer. Well, a custom integrated circuit is the heart of somebody's system. And it's like going to a heart surgeon for a bypass and you ask them, well, have you done many of these? And he says well I've, you know, taken classes in it and I've studied it, but actually I've never done one yet, you'll be my first. So it was a similar deal when we tried to get people to give us a contract and then finally this friend of Wilf's who had a power supply tester company, they built testers for power supplies, which was a very, very niche marketing, very small market. He came in and he said I'm going to give you your first three designs and here they are and a terrifically low volume and terrifically simple and I looked at him and said gee these really aren't candidates for LSI, you should do these with standard MSI, that's obviously the way to go. So Wilf took me aside and said Rob, "Shut up and take the business." So we took them and in fact struggled to get those out, those were our first three, they all worked the first time out, and so for some time Halfon and his VP marketing, they were our only people that we could refer other customers to, so customers would tell us, would ask us well, have you done many of these before, we'd say oh yeah, we've done many, many of these, well could you give us some names and references? Well we're under non-disclosure, it's very difficult, and anyhow, that allowed us to, those three dumb designs, I mean economically made no sense at all, allowed us to then book other customers including Wang and including DEC and Hughes and become successful. And once we really started turning the crank, we were able to crank out designs increasingly faster and we hit a peak in the mid '80's of ten designs a working day, we would crank out ten integrated circuits a working day, all of them working either the first time or by the second, by one integration. So we really had a design factory going. Anyway, that part really worked well. Let's see, another interesting design was one we did for Delco. Delco came in and said we want a light controller for our Camaros and we want it to do the following and it had all sorts of bizarre kinds of things it would do and if you left the turn indicator on for more than three minutes, it would decide that you had neglected to shut it off and it would shut it off automatically and if you parked the car it turned the key off and had the lights on, it would shut the lights off after a certain time. And Delco in its, one of the reasons that General Motors continues to sink, had decided they wanted all these times accurate to about one percent which is better than you can do if normal RC type of time constants, so I used a crystal control so that's probably the only light turn indicators accurate to a microsecond with a crystal control and I used color burst crystals so they're about a dollar and they were in high volume. Anyway, so that was very early design we turned out and I did it with a young engineer just out of school and he and I did the design and the customer who was about twenty-one as well, customer engineer, so there was me trying to be the VP of Engineering and working on this thing on the side and these two kids right out of school. Anyway, we did this design and unfortunately in the test tape we had neglected to check whether the left turn indicator pin worked or not, an oversight, so some percent of these that we shipped to Delco, perhaps one percent, didn't work when they put them in the module, so Delco started shipping back these light modules, power transistors and its assemblies and they're piling up and meantime our, the guy in charge of manufacturing, he took a very legalistic view and said well, this is your test tape, this is your spec, you didn't test it, and when I found out about it I said my god, we have to revise the testing, we're not testing it completely and we did that and we found indeed and now we started shipping good parts. Any rate, we still had boxes full of these modules and Delco never gave us another job after that. So we were right legalistically and it worked, most of them worked, but it didn't work. Any rate, this culminated in '83 and we went public and that was a great time.
SW: I tried calling you from Wang that day, the phones were very busy, what were you doing?
RW: Well, we were drinking actually. We went out, we were in Milpitas and there's a winery near Milpitas up in the hills that has champagne, I think it's Weibel and we went up there for lunch and we started drinking champagne and we didn't get a whole lot done that afternoon but that was a crazy time because we had done several things. We had gone public with a huge public offering at the time, the biggest since Ford Motor Company. Based on, I don't recall our sales, but they were twenty million, thirty million dollars or something like that. We weren't very large, but it was a time of near hysteria in the market similar to 1998, now. So it was that, I made on paper, at least I was worth a lot of money and probably most importantly to me it vindicated all these principles, techniques, concepts that I had laid down, through the years going back literally fifteen years or longer and most of which were very unpopular, very much out of style, I was right and they were wrong. So that was really a good feeling.
SW: Subsequently you wrote a book about LSI Logic and ASIC design, can you tell us something about the book?
RW: Yeah, I just happen to have one here, called Silicon Destiny, and I named it that because it was my destiny to do this application specific integrated circuits and do them right. And looking back over the decisions I made in my career, to go to UC Berkeley, to take digital, to learn all at Lawrence Rad Lab, to learn at Ford Aerospace to learn about logic design and how to do it the right way. All the things we learned at Fairchild, how to manage and promote the wrong way, but how to do Computer Aided Design and even my days in Marcom at Intel I learned how to promote products the correct way and how to write ap notes and data sheets, which I did even though I was VP of Engineering. In the early days I did all our literature, our technical literature, at LSI Logic. So all those things had just worked out and all the timing having done that article back in 1980 with Bob Hartman, all of those things had added up to this success. If you look at the odds on each one of these happening and then you multiply all those odds together the way you do probabilities, the probabilities are down with winning the lottery, are probably worse than winning the lottery and getting struck by lightening or one of those things. It's just incredible that that all happened. So I decided it couldn't be just happenstance, it must be destiny. So I called it Silicon Destiny. And the book is a series of oral histories on the subject of Application Specific Integrated Circuits and Computer Aided Design. And then the companies that espouse those. Starting with LSI Logic but going into VLSI and going through Altera which is the company that Bob Hartman, after he and I wrote the article, he founded Altera, which were programmable ASICs and he went on to fame and fortune and then there was a series of losers as well that I picked to show the wrong way and unfortunately Susan, that's your chapter, Wang is one of the series of losers.
SW: Yes I remember, "Wang Self-destructs".
RW: It was unique because my group of engineers were primarily Chinese Americans and I don't know how it turned out that way but it turned out that way and they were the best in the world. And I went back to Wang, a very important customer I had visited there many times, and I was awe struck. Here are truly incompetent Chinese engineers, I didn't know that that was possible genetically. These guys were such clowns. Any rate, that was one, surprisingly another reason was Intel who attempted to get into this business, headed up by my old boss Jack Carsten, a brilliant guy, engineering being done by Don Soderman, a guy that used to work for me that learned all the LSI Logic ways of doing things and with Intel's fabulous marketing capability, fabulous process and so on, and we kicked their butt and they got out of the business. So it was very satisfying.
SW: Well aside from the book Rob, you've written a lot. Can you tell us about some of your past articles and publications?
RW: Well, I'm glad you asked me that, Susan. I just happen to have here this collection of articles that I've written over the years and this too is going to Stanford. I wrote the first one back in 1959, the first year I was out of college and it was about vacuum tubes and here's one called Designing a Boot Strap Emiter-Follower Amplifier that was the result of a homework assignment I had at San Jose State. And here's one about a 945, way to use a 945 DTL flip-flop. Here's Digital to Analog Converters using discrete IC's, using bipolar transistors operating them backwards. So any rate, I've got a bunch of these and I'm going to send these to Stanford for the archives for those of you that want to learn how to do vacuum tube designs. But from a career standpoint, I think it really is an excellent way to get yourself known, get your name out there, and it publicizes your product. In some cases, it was the only thing we could afford to do. And as you know Susan, this year, '98, we just published in Electronic Design an article that tells about our consulting business. So I've been writing these things since from '59 to '89.
SW: I suspect you have a few more articles in you yet Mr. Walker.
RW: Well, we'll see.
SW: In retrospect, how important have semiconductors been to our society?
RW: Well, I think absolutely crucial. We're in something called the Information Age today, which is fundamental as the Industrial Revolution was; it has changed the lives of people everywhere. In China I read that ninety percent of the people in China have television sets. And of course the world is wired together now, you can talk to anybody, anywhere, anytime. The average American touches forty microprocessors before they have lunch; in their VCRs, in their microwave ovens, a number of them in their car, and they're ubiquitous. Somebody has said, and I don't know how accurate it is, but there's now something like a million transistors for every man, woman, and child in the world that are manufactured and used. And that sounds kind of crazy, but a modern microprocessor may have five million transistors in it. And so they add up. Anyway, it has been just a fundamental change in our society and you wouldn't have inexpensive computers without that, you wouldn't have satellites able to transmit these TV pictures all over the world, you wouldn't have cell phones. In other words, if we were stuck with vacuum tubes of the earlier technologies, these things would not be possible either on a cost, reliability, or size, or power standpoint. So it is really fundamental and the reason that I am writing books and doing these interviews is because most of the pioneers in this field are still alive and they're available and they're willing to talk about what they've done. That's not always going to be the case. So I'm just trying to go out and talk to these people and where possible I try to go to their homes or I go to their businesses, interview them there, where possible talk to their partner, their wife, and not only hear about the kinds of things that you would get out of a history textbook about what they accomplished and so on, but to get a little bit of an idea on how they live.
SW: Rob, do you have a name for this project?
RW: Yes, the Silicon Genesis Project. I am doing it in cooperation with Stanford. There is a video, which is called Stanford and the Silicon Valley Project, which talks about Stanford's efforts along these same lines. They are thinking the same way I am that they want to document as much of the pioneering high-tech aspects of Silicon Valley as I do and they are going about it by getting all the records, for example that they could of Fairchild, they just got a huge group from Apple Computer and so we're headed in the same direction. In my case, I am doing on videotape and I'm only doing semiconductors.
SW: Rob, what have you learned from the Silicon Genesis Project?
RW: Well, it's been fascinating visiting these people in their homes and talking to them about their careers. Number one, nobody started out to make a lot of money, most of the people we talked to have, but nobody, instead they wanted to do something neat and groovy, and fulfill a need. Uh, secondly literally none of these people came from a wealthy family. Bill Gates is really the exception and not the rule. Everybody that I've talked to came from a middle class or a working class kind of family. And along the way, there's been a number of interesting disclosures on the microprocessor, Ted Hoff and Gordon Moore both disclosed that Intel never bothered to get a patent on the microprocessor. Because they thought it's just a mini-computer on a chip and so there's really nothing new. So they didn't bother to get a patent and in the meantime some yo-yo who never built one actually did get the patent and that court battle is still going on today. Learned that Les Hogan who I interviewed who is in charge of Fairchild for the years, a number of years following Noyce and Moore's leaving to start Intel that, and started the decline of Fairchild, he disclosed he had never been well during that period. That he'd had an operation just prior to leaving Motorola, they botched it, and he was never a hundred percent until he had another operation, but by then things had really kind of gone to hell. Its really been fascinating, learned from Regis McKenna who talked about so many of the promotions, he's sort of the master of public relations and now he's in the venture capital business as well and some of the machinations at Intel that resulted in Motorola being pushed out and Intel becoming the preeminent supplier of microprocessors. That was a very near thing that could have gone either way. Motorola in fact had a better product and in fact so did Zilog and learned why Federico Faggin founded Zilog and the Z-80, Z-8000 which again were better products than Intel had but were not able to capitalize on it. And so Federico was pleased that Zilog was still as of 1998 still actually in business, albeit much smaller company, they were the only company that Exxon didn't manage to destroy along the way out of like fifteen or so. So many many interesting stories.
SW: When do you think the Silicon Genesis Project will be complete Rob?
RW: Well I hope never. I've been trying to get others to involve themselves as well, and after I run out of ideas, hopefully other people will take it over. Certainly there is no limitation on what people can do and give to Stanford.
This is Tiger of Tiger Productions, he produced a lot of the early Silicon Genesis work. Well okay, I think now it's time you ought to come over here and let the people see you. So why don't you come over Susan. We have the whole family together, here's Tiger.
SW: Hello Rob.
RW: Hi there, so here's Susan and Tiger and my goodness he's right on cue. Well any rate, we've been married a year now and we met in '81 and our meeting caused you to be grounded for a year. Why don't you tell people about that?
SW: Well that's true. At the time, back in 1980, '81, Wang was very aggressively going after developing an internal fab process for designing proprietary integrated circuits within their own facility; captive fabs back then were very popular. I was in charge of design automation for Wang Laboratories, their new start up in semiconductor design and was sent to the Valley to make sure that I could interface between Wang's design automation tools and the companies that we were working with like LSI Logic, National Semiconductor, AMI, the infamous California Devices and several others. That's when we met back when LSI Logic was a very small company. I remember the business plan, the ink was pretty wet on the business plan at that time and you really didn't have any products, however you had an idea, you had a dream, and you had enthusiasm and I really believed that you guys could do it. I had come out to the valley to visit your first facility in Milpitas and I remember walking in the door and you didn't even have any furniture, you were sitting on boxes, you and Wilf and Bill O'Meara, and Mick in the middle of this huge building in Milpitas and I told you guys, the Valley is five miles away, what are you doing out in the boonies here? To make a long story short, we went to dinner that evening and it was Wilf and myself and you Rob. And during that dinner what we didn't know was there was somebody having a meal also at Ming's of Palo Alto from Don Hoefler's magazine, Microelectronic News and this at the time was the National Inquirer of Silicon Valley; it was basically who was sleeping with who, who was screwing who at the time in business. I returned back to Massachusetts to Wang to my office and with the next edition was surprised when my vice president walked into my office and said Susan we just got Microelectronic News, are you the shady lady from the East Coast Chinese company who was having dinner with the infamous Wilf Corrigan? And I said well, I did have dinner with Wilf and Rob on my way to the airport. And actually reading the article it looked like I was having an affair with Wilf and Rob was the beard that evening. Needless to say, my flying wings were clipped for quite a long time. I ended up going to Pocatello, Idaho and Motorola in Phoenix and National Semiconductor in Salt Lake City, but didn't see the Valley for almost six to eight months.
RW: Well now you've lived here for a year, you've always been in the past working in the New England, greater Boston area, Route 128, so what are the differences you see here in the Silicon Valley?
SW: Well of course the most outstanding thing is the weather here, not to have to worry about snow. Certainly the sunshine is a great uplifter for all of us. But what I notice here is that both the entrepreneurial community and the venture capital community are very intertwined. There's a lot of networking that goes on here. People are very free to speak with each other and to share information and this enthusiasm is electric. Back east, we tend to huddle in small corners, be quiet about our information, stay much more secretive and that doesn't, I don't believe that leads to a lot of good cross pollination and information sharing that allows some of these companies to be given birth out here. So there's an incredible enthusiasm for people who want to start companies, they're really, really encouraged.
RW: Well thanks Susan for talking me into doing this, I've hated to do it, I like to be behind the camera, not in front of it.
SW: So I've noticed.
RW: But we should make one thing clear to our spouses at the time, back in
'81. We were not screwing around, that's only something reasonably recent. So
please don't think we were doing that at that time, it was perfectly innocent.