March 23, 2016
RB: Hello again. This is Robert Blair for Silicon Genesis at Stanford University Libraries. Today, March the 23rd, 2016 we will talk to William Davidow, known as Bill to all his friends and to all of us in the Silicon Valley. Bill was born originally in Pennsylvania and grew up in Chicago, Illinois. He earned a bachelor of science in electrical engineering from Dartmouth College. Interestingly, it was the Russian launch of Sputnik that led Bill to earn a PhD from Stanford University. Bill stayed in Silicon Valley following his PhD and joined General Electric R&D labs. Following General Electric, Bill moved to Hewlett-Packard where he worked on computer architecture. While he was there he met Tom Perkins, also a well-known venture capitalist in Silicon Valley. Bill was enticed to Signetics following that, where he defined some of the early concepts of a microprocessor, at that time in bipolar technology. That turned out to be important as his next move was to Intel Corporation, who was in the process of migrating from principally being a memory company to a logic and microprocessor company. There he worked with Noyce, Moore, Grove and other well-known Silicon Valley semiconductor staff. Bill spent a decade at Intel defining their marketing and products. He wanted to move into business, so following his decade at Intel decided to become a venture capitalist and joined with a partner, Mohr, to form Mohr Davidow Ventures. The partnership funded multiple semiconductor companies in the ‘80s and ‘90s up through the bubble in 2000 and beyond. Today Bill continues as an advisor to Mohr Davidow where his name remains over the door at 3000 Sand Hill Road. Bill’s written four books in his career, all around the concepts of technical marketing and the impact of the internet on marketing over the last ten to twenty years. Bill serves on a number of prominent board institutions in the valley, and remains fit at his eighty-plus years with tennis being the primary sport of interest. I caught up with Bill at Mohr Davidow Ventures’ offices recently on Sand Hill Road, the center of gravity for the venture community. Let’s talk to Bill now about his story of the semiconductor industry, his time in Silicon Valley, and his vision of where this is going.
RB: Well, hello Bill, and good to see you again.
BD: Good seeing you, Bob.
RB: Thank you for taking the time today to talk to Silicon Genesis at Stanford Library. I know you’re a Stanford PhD, so it’s obviously very fitting that your semiconductor story is captured for future generations to review. And we’re pleased to be here today at Mohr Davidow Ventures to talk to you. We’ve known each other for about twenty years. May –
BD: Oh, I think – I think –
RB: Maybe more. Maybe more.
BD: – considerably longer, Bob.
RB: Okay. Okay.
BD: I’d put us at probably at least the thirty-year mark.
RB: Okay. But this time I’m not here looking for money, Bill. So –
RB: – I’m here to talk about your history in the semiconductor industry, which is –
RB: – you are one of the Silicon Valley icons. But before we start, as you know, this week there was the sad passing of Andy Grove.
RB: So I thought we would just acknowledge Andy for his history in the industry. And I know he was a mentor of yours and you worked for him, so we’ll talk about Intel so we’ll come back to that. That certainly –
BD: And Andy as well?
BD: All right.
RB: So – but I’m sure the – the world will miss Andy Grove as a semiconductor person. Let’s start at the beginning, Bill, if we may. And you’re not a California guy. Tell us where you were born and what your parents did and what were the first fifteen, sixteen years of your life actually like?
BD: Well, I was born in Reading, Pennsylvania, but grew up in Chicago, or a suburb of Chicago. And my father was a book publisher. He actually never graduated from high school but set out on his own in I think 1916 to sell candy. And then he ultimately sold carloads of candy. And then met Walter Annenberg and moved to Redding, Pennsylvania and bought a news agency and was distributing newspapers there, and published a cookbook. And ultimately, because he had to get to a higher volume printer, met a printer named John F. Cuneo in Chicago. And Cuneo always wanted to be a book publisher, and so my uneducated father said I’ll make you one. And he – it was right during the middle of the Depression, and my father realized that people were spending a lot of time reading because they had nothing else to do, and he knew that because newspaper circulation went up. And so he selected a hundred classics, bound them and used them as newspaper promotions, and I think sold 35 million books the first year and that established him in the publishing business. And so he was this very creative, sales-orientated person, and my mother was this very pragmatic, organized person. And I grew up in a wonderful household, and somehow became interested in mathematics and geometry. And I remember spending – my father telling me it was impossible to trisect a triangle, and I remember spending a lot of time trying to do that as a kid. And then I went on to Dartmouth. And I made the rational decision to go to Dartmouth because I wanted to learn how to ski. And they had an engineering school, which was ideal for me. And I was sitting in class and I – I’m going to say it was October 8th, 1957, but it may have been one day or two the other side. But it was the day that Sputnik was launched. And I remember the professor, Millett Morgan – it was an electromagnetic theory class – walking into class and announcing the launch of Sputnik pointing out that our satellite was going to weigh around three or four kilograms, and that Sputnik weighed around thirty – I think it was twenty-eight – telling us about the difference in rocket power required to launch an object that size, and announcing that we were hopelessly behind Russia. And I said to myself, if you have any technical ability, you’ve got to save the world from Russia. And I then made that decision to go on and study for a PhD. And there was a – that was my senior year at Dartmouth. I was going on to get a master’s degree and I noticed an eight and a half by eleven inch sheet of paper on the wall in – probably it was – would have been 1957, and it was announcing a master’s degree program at Caltech. And I investigated Caltech and was told that it was the hardest graduate school in the United States. And I decided I would go there. And so I got a master’s degree from Caltech, and then ultimately came to Stanford and got my PhD in 1961.
RB: Did you have any sort of technical hobbies as a child, Bill, or did you just suddenly sort of go from home life to the engineering school in about a year or something?
BD: Well, I never was that much of a tinkerer. I had a group of us and we used to work on one car. We bought a car that we could work on and couldn’t break. And we got that running and we used to do all kinds of fine things with it. We decided that it needed a paint job and we painted it with basically watercolors, and it rained and so it was called the Vomit Comet after that. [laughter] And we used to take it to football games and things like that, and do all kinds of things that any normal person would do. Like we took the trunk off it, and it had a handle on it, and you could tie a rope between the trunk and the back bumper. And it snowed a lot in Chicago at that time, and you could haul – you could ride on the back bumper and – I mean you could ride on the trunk and go around corners and swing wide and things like that on the snowy streets. So it was a very useful thing. And then among the other useful things we did was – I forget – calcium carbonate is what they use in miners’ lamps. And if you put water on calcium carbonate – I believe I’ve got the right compounds. Anyhow, it releases methane and that’s the gas for [inaudible]. You can also blow up cans with it. So we were into important scientific advances.
RB: So some mechanical engineering and some chemistry.
BD: Yes. [laughter]
RB: So, okay. Interesting. So now you have a PhD from Stanford, also in electrical engineering.
RB: What was your thesis on, Bill?
BD: It was on logic design. And I – as I – after my year at Caltech, I realized I wasn’t probably cut out to be a great scientist, but I – and I was – had always been interested in business. But I thought, you know, if you start something you ought to finish it. So I finished my PhD and then kept reading about business and things like that. And I then, after Stanford, went to work for General Electric. General Electric had a computer lab here. And it – GE had designed the first transistorized computer for the Bank of America, and that design was done in Palo Alto. And it was the ERMA system. And General Electric then decided to get into the computer business based on the ERMA system, and they set up the computer department in Phoenix, Arizona and left some of the technical people here. And so I was recruited to work there by Joe Weizenbaum who was one of the founders of list processing, but also – or one of the fundamental scientists of list processing. Joe went back to become a professor at MIT. But these were the – basically the early tools and – of – that were used in what today we would call artificial intelligence. John McCarthy, who came to Stanford, was the original thinker behind list processing and Joe came up with a version of it. And he went on to MIT. And it was – one of the most famous things he did was he constructed a – sort of a system that enabled you to have a conversation with the computer, and the computer sort of served as a – if I’ve got it right, a psychoanalytic tool that kept asking you questions to discover things about you.
RB: So what year was this, Bill, roughly when you came out of Stanford?
BD: Well, I joined GE in ’61 and was there through ’65.
RB: So ’61. I mean that’s a pretty exciting time here in the valley. I mean did you – so with your PhD in your pocket, did you shop around in Silicon Valley, or did you somehow magically jump right into GE?
BD: Well, I knew that I wanted to be in the computer business. I had a friend of mine, Bob Rosen when I was growing up in Chicago, and he went on to Michigan. And it’s interesting. Caltech was not particularly strong in digital computation. They were very strong in analogue computation. And after my year at Caltech, I used to have discussions with Bob about whether analogue computation was going to be more important than digital computation. And I tried to convince him of the importance of analogue computation. And, you know, it’s hard to imagine these things, but at that time one of the most important – one of the – of the very important problems had to do with the stability of airplane wings. And it turns out that if airplane wings aren’t properly designed, when they get buffeted by the wind they have harmonic vibrations, and if they aren’t dampened properly, the wings fall off the plane. And, as a matter of fact, when I moved to California, there used to be Lockheed Electras that would fly back and forth to Southern California and one of them lost its – a wing as a result of this. And Caltech had the analogue means of computation at that time that was used because the calculations were too –
BD: – complex. And we also had two digital computers at Caltech. We had an LGP-30 that was invented – or developed by Stanley Frankel. And it had fourteen vacuum tube flip-flops, and 768 germanium diodes, and 4,096 words of drum memory, which meant that it could perform – the drum was 60 RPMs, so it was 60 instructions a second. And then we had a Burroughs 205 which was a larger computer, and it also had a drum memory. And it had a short loop on it which meant it ran about ten times faster than the LGP-30. And Caltech, at that time, did not consider programming to be a legitimate academic discipline. So if you learned how to program, you kind of learned how by interacting with people. And so I learned a little bit of programming. And I remember learning every logic equation in the LGP-30. So that was my logic design training. And then there was a wonderful professor at Caltech named Joel Franklin who was a numerical – not numerical theorian. I’m searching for the word. But he was very strong on numerical computation, applied mathematics, and so he taught me a lot about problems with calculation and the errors that crept through calculation. And interestingly, the floating point standard that ultimately became the IEEE floating point standard was a direct result of what I learned about the propagation of round-off errors at Caltech in 1958 and ’59. And years later, I went back and told Franklin that that was one of the practical fallouts of his teaching.
RB: So after your GE experience, which of course was a huge brand name at the time, and today one tends to think of it as maybe aero-engines and washing machines –
BD: Right. [laughter]
RB: – rather than computation. And then you found your way to HP. Did you find HP or did they find you, Bill?
BD: Well, I had a friend of mine that I went to Stanford with who was at HP, and he said HP’s getting into the computer business. Well, I wanted to be in the computer business. And at that time – it’s hard to imagine. If you wanted to be in the computer business in Silicon Valley, as far as I can tell, at that time you either worked for GE or for IBM, which was in San Jose. And so HP had bought a small mini computer company in Michigan and moved the people to Palo Alto. And then they put the development under a person named Kay Magleby. And I joined that group. And that was to help HP get into the computer business.
RB: Okay. So was that more digital or more analogue?
BD: Oh, that was digital for sure.
RB: That was digital. Okay.
BD: And I was a digital person by that time.
RB: Okay. Okay. Okay. Okay. Now, I think that our mutual good friend, Tom Perkins shows up at HP about that time.
BD: Right. Right. Well, Tom – I’m not quite sure when Tom arrived there. But what happened was really, HP entered – the HP architecture was horrible. It was about as bad a computer architecture as you could choose. And I was just shocked when I saw it. And I ended up – and then HP didn’t know anything about marketing. And meanwhile I had been doing all this reading about marketing and sales and what have you. And so I went from engineering into marketing to help them sell the computer. And Bob Grimm was running the area. Then Jack Melchor took over from Bob. And then ultimately Tom Perkins came. Tom and I became very good friends, and I ultimately ended up running marketing for the HP computers.
RB: So that was the beginning of the Bill Davidow marketing era –
BD: Right. [laughter] Well, I –
RB: – which you obviously have got a trademark stamp on by now. So, okay. Well, that’s really interesting. So how long did you stay at HP with Tom?
BD: I was with HP for four years. And then Orville Baker found me and – well, it was really interesting. I had been playing around with HP and I came up with an idea for something that today we call a microprocessor. And I was very interested in it. And I – let’s see. HP would have been – I left HP in 19 – around 1970, and I – and Orville was – they were forming a spinout from Signetics called Signetics Memory Systems. And I said, well, what you really should do is build this microprocessor. And we came up with a – and I had had a friend of mine at GE named Glenn Oliver who was a really good computer architect. And so I joined Signetics Memory Systems and Glenn – we hired Glenn. And the idea was to build a bipolar microprocessor that could be used to replace logic in systems. So the idea was to do programmable logic at that time. And Glenn came up with – you know, since transistors and logics were very expensive at that time, a very neat, minimal system for a way to compute logic equations. And that became the basis for the Signetics – ultimately Signetics Memory Systems failed and Signetics took over this as their bipolar microprocessor. The problem we had at Signetics Memory Systems was that I thought the semiconductor part people were going to take care of. And I knew nothing about semiconductors. And we couldn’t make a semiconductor yield. I mean it was – the group that was responsible for that would process wafers. And they weren’t processing them correctly and the wafers would come out all warped like potato chips. And the yields were effectively zero. And – but at that time we came up with all these programming aids so that they would facilitate the ability of engineers to program logic. Engineers were doing logic design. They hadn’t thought that you could use computers to do logic design. And our insight was that you could use read-only memories and very simple computation methods to compute the logic equations and implement all these logic circuits and that it was very economical to do that. But in order to do that, you needed programming aids. And so we came up with a simple assembler, and we came up with a read-only memory simulator that it used bipolar memories. And what you could do is you could take what was going to be the read-only memory to go with the microprocessor, pull it out, plug in the read-only memory simulator, and you could then program the logic and make logic changes and debug things. And so we came up with a set of programming tools for that application.
RB: So this is around ’70, ’71.
BD: Well, it would have been ’71 through ’73.
RB: Okay. So at that time, I mean Intel was obviously launched –
RB: – existed. Were you sort of watching what they did, or –
BD: Well [laughter] –
RB: – not really paying attention, or –
BD: So I happened – I was in a bar in Las Vegas and ran into Bob Noyce and Ted Hoff and started –
BD: – talking to them about microprocessors, and, you know, how important they were going to be. And Noyce was fascinated and he said, “Why don’t you come and make a presentation to the Intel board.” And I did. No. I’m trying to remember whether I – I’m trying to get the timing on this correct. Maybe that was – actually, that was before I joined Signetics Memory Systems. I was – I was still working for HP at the time.
BD: Because Intel had started in ’68, and I joined in ’73. And I was still working for HP when I ran into Noyce and Hoff. And I went to Intel and I said, “You ought to consider getting into the microprocessor business.” And I remember Tom Perkins saying, “You better be careful if you’re still working for HP doing stuff like that.” And HP has never sued me about that. I don’t know as I’ve ever openly told this story for the history of the world. And so then I went off to Signetics – never heard much back from Intel. Then I went off to Signetics Memory Systems to develop my own version of the microprocessor. And Intel, of course, announced the 4004 I think it was in ’71, if I’m not mistaken.
RB: Around then.
BD: And so it was highly likely that Hoff was working on that project [talking over each other].
RB: So they might have benefited from that bar conversation in Vegas.
BD: I doubt – I doubt whether he did.
RB: No? [laughter]
BD: But I think they probably found it of interest. And then – I wasn’t getting along with my boss at Signetics Memory Systems. Signetics had replaced Orville with a person named Hugh Kern, and he and I just didn’t get along very well. And so I walked into his office and said, “I’m quitting,” and walked back to my office and the phone rang, and it was Ed Gelbach saying, “I’d like to talk with you.” Ed was head of sales at Intel. And I went over and I talked with Ed. And Hank Smith was at Intel at that time and he had been running the microprocessor area. And he was leaving. And I – apparently – well, I know – I’m quite sure that Noyce had given my name to Ed. And they offered me Hank Smith’s job which was kind of interesting. The engineering of the microprocessor was going on in engineering, but sort of the marketing and software development was going on under Ed in marketing. And I took – and they were – they had some very crude development boards that were being sold, MCS boards. And they were horrible. And I remember looking at the electrical design of the system and thinking somebody’s going to electrocute themselves. It was just not well [talking over each other] –
RB: Not well designed.
BD: – at all. I mean Intel had no idea. And so I came to Intel and took over the group. And at the time the manuals were costing us more to make, or to print, than the cost of the silicon we were selling because we were selling so little silicon and everybody wanted to read the manuals. And so I – and what I brought with me were a lot of the ideas on doing computer support for the microprocessor. And my feeling was that we had to get engineers comfortable with the idea of using computers instead of logic. And so I had had my experience with the ROM simulator, and there was a individual at Intel named Hap Walker, and there was somebody running software development named Terry Opdendyk. And what – I kept talking to people about the ROM simulator and how we could facilitate things. And Hap Walker came up with the idea of the in-circuit emulator. And he was the only person that was working for me who would listen to me about this idea because Terry wanted to do it all in software and I said you can’t build a fast enough software emulator so that it’s meaningful. So that – Hap came up with the idea of putting a plug in the socket for the microprocessor, and we would pull the microprocessor out of the system and imbed it in a hardware box, which we called the microcomputer development system. And we had all the software tools, and you could put in break points, and this and that and the other thing. And the idea was that we would sell the box to engineers, and then within a couple years they would be buying, you know, twenty-five thousand dollars’ a year worth of Intel microcomputers. And those systems ultimately ended up selling for – I don’t know – ten, twenty thousand dollars a year. And we created a hundred million dollar business selling the design aids for the microprocessors, and the idea was to populate the world with Intel instruments that could be used to design the microprocessors into the system. So – and then we had – I hired Jim Lally and somebody named John Pavone, who had worked with me at Hewlett-Packard, and we put in training for engineers much like IBM trained computer programmers at the installation, so that we were training – I don’t know – a few thousand engineers a year on how to program Intel microprocessors. So we had the schools. And then we ended up getting application support into the field. And we put a support network in behind the microprocessor at the time.
RB: So this is pretty radical. I mean this really – this cast a methodology and a direction for Intel that was brought by you to the company in the early ‘70s and set them on that track of being – using design systems for microprocessors to populate potential silicon business.
BD: Yeah. And the significant difference there, and why we could do it is that I had this hundred million dollar business that – which was a system business. And so [laughter] we had maintenance people, we had this and that and the other thing, and the other semiconductor companies didn’t have this stuff. And so, you know – and I – in a sense, that business could finance the field support – I mean it was just like having field support for mini computers.
BD: And so I had a little mini computer business going.
RB: So – I mean at that time the semiconductor industry was basically shipping hardware.
BD: That – chips.
RB: Chips. And so this was pretty radical for, you know, the industry and for design methodology in general.
BD: Yeah. I – well, yes. Yeah.
RB: So that’s very interesting. So Intel became your home for a decade.
RB: And the Bill Davidow stamp on Intel I think is become legendary in that that’s where you emerged as a marketing guy in the true sense of the word. I mean, you know, like you I’ve spent years and years in product marketing, which I think was a term that the semiconductor industry invented.
RB: And it was a very important and valuable function in the ‘70s and ‘80s. And so you spent a decade with essentially the best guys in the business.
RB: Noyce, Grove, Gelbach, House. And you were right in the middle of all these guys. And so it was a hell of a ride ten years at Intel for that decade.
BD: Yeah, it was. And sort of the way – you know, I had always had this image that I wanted to become a general manager and president of something, and what have you. And so Les Vadász and I co-managed the microprocessor area and we both were reporting to Andy. And then the microprocessor system business was something that I was – the boxes that I was really interested in, and I said to Andy I’d like to go run that business. So I was running that business and Les was running the microprocessor business. And so – and we were both working for Andy, and Jack Carsten was running – it was either Jack – I – was – Gelbach was running the memory business at that time, I think. And he hired Carsten, who was running sales and corporate marketing for the company. I think that’s the way it worked. And so I was not really doing marketing, I was a big deal general manager. Right? And I walked – we – I – we had the Intel 86 architecture. Motorola had the 6800 architecture going to the 68000. And Federico Faggin had left to start Zilog with Ralph Ungermann. And it turns out that we were moving to the 8086 architecture, and the 8086 architecture was a architectural disaster. It was built to remain compatible with the 8008. We wanted to maintain some kind of compatibility because the 8008 had been the first microprocessor. And, you know, I was all in favor of trying to preserve the programming lineage. And – but then Motorola came up with this much cleaner, better architecture and Zilog did a better version of the 8086. And suddenly I’m sitting in my office and I get this five-page telegram from Casey Powell and Don Buckout, who were selling in the Long Island area, about how we were getting ourselves just destroyed in the market. And so I went to the executive staff meeting with the telegram. I made copies of it for everybody. And I said – and I made the statement that – quite forcefully, that we were losing in the market, that if the semiconductor guys didn’t start getting designs, my business was going to go right through the tubes, and that they were acting like a bunch of whipped dogs. And I had had an experience of selling one of the worst architectures in the computer industry before at Hewlett-Packard. And what I learned at Hewlett-Packard was that there was always a market segment that you could serve better than anybody else. And that if you focused on that you could win. And that’s what we did at Hewlett-Packard, and ultimately Hewlett-Packard overcame a lot of the architectural problems and went on to become a tremendous success in the computer business. But my lesson from Hewlett-Packard was that you didn’t need the best technology to necessarily win. And Andy Grove looked at me at that meeting and he went, “You fix the problem.” And so I was running a division, and I became – also then had a staff job of running the marketing program at Intel which became Operation Crush. It was Jim Lally who suggested the name. And the Denver Broncos had a frontline which was called the Orange Crush, and so the idea was to crush the competition which was – it was not the kind of thing that you would publicize to the Antitrust Department.
RB: But that was a very high profile program at the time. I mean it raised a lot of eyebrows and got a lot of attention.
BD: Well, what happened was that I had a lot of authority, and I had a lot of authority because Andy said he wanted it done, and Bob Noyce supported the idea. And I don’t know how many divisions Intel had at that time, but it enabled this group that was Rich Bader, Jim Lally, Regis McKenna was helping us, to come up with a marketing strategy where the argument was that the microprocessor architectural decision was the most important technology decision your company is going to make over the next decade. And that forget about the architecture of the microprocessor, what was important was to get you product to the market successfully. And that guess what? Intel had the software programming aids, we had the peripheral circuits, we had the design aids, and the microprocessor architecture was incidental. And we got the company revved up around that idea and we went out and we had this big marketing program where we – out of the air – right – we picked a goal of two thousand design wins. And I mean I – you know, it was just like –
RB: A number.
BD: It was a number. And we got everybody revved up. And, you know, Andy was beating the drums and this and that and the other thing, and we had a coherent positioning. And it was very, very hard for other people to duplicate that. We had the application engineering infrastructure in place. And people didn’t have the application infrastructure in place. So we went out and, you know, we got the important design wins and that was what built the momentum behind the microprocessor.
RB: So your important vision here of having the complete system support, you know, overcame another piece of superior core hardware –
BD: Well, it –
RB: – as a processor.
BD: The [laughter] – it turns out that the pieces were all there, it was just a matter of then pulling together a coherent marketing story, and we were uniquely positioned from that point of view.
RB: Classic product marketing, as it became -
RB: – to take the product you’ve got and package it and position it in the right manner to, you know, win in the market –
RB: – even though it might not be two nanoseconds better than the other guy.
BD: No. It turns out that it was what you define “better” to mean, and we chose to define better to mean that you would end up with a system that was more economical, that you could get to the market predictably, and – you know, and we would make the project successful –
BD: – rather than worrying about whether this one benchmarked better against that one.
RB: Yeah. (?).
BD: So we sold what I would call project success against benchmarks.
RB: Yeah. Sort of a little bit like, you know, some of the upcoming activities for maybe Microsoft and, you know, Windows versus others. But we’ll come back to that. So you had a very interesting decade then at Intel, Bill and obviously spent a lot of time with Andy over – over that period. And unfortunately is he’s no longer with us. Maybe it’s appropriate just – you know, give us your opinions of working for Andy over that time. You obviously got to know him extremely well in both directions.
BD: Sure. Well, Andy was very smart, incredibly organized, extremely driven. And I’ve been reflecting on this over the last couple of days, but he had an uncanny ability to get people to perform better. I used to say about Andy, he took B’s and made them A, C’s and made them B’s, and B’s, then made them A’s. And that was a lot of result of some of the Intel systems he put in place. But he had another tremendous ability, which was to sort of talk very directly with people and get them off in the right direction. And, you know, I don’t know in the past forty-eight hours how many people have said, “Andy shaped my career.” I was sitting next to somebody (?) the other night, you know, Andy was my mentor, and this and that and the other thing. And I’ve always said that Andy shaped my career. So I was forty-five years old. I had just run this successful marketing campaign at Intel. And I walked into Andy’s office and said I’m going to leave Intel. I had gotten a job offer from Kleiner Perkins to become a general partner. And Andy launched into his sales mode. He looked me right in the eye and he said, you’re an average general manager. [laughter] And I said, what does that mean, Andy? And he said, “Well, so that means you can be a general manager for me as long as you want but – and that’s okay. And when I ask you a question about a general management problem, you say, well, let me think about it, and a week later you come back to me and say this is what we should be doing. And said you’re right. And that’s great. He said, but you’re the best marketing guy in the company. And I said, why don’t you go run sales and marketing? And he said, when I ask you a marketing question you always say this is what you should do, and it’s right. So I went off and I ran sales and marketing for Intel for five years. And the reason I say that that shaped my career the rest of my life was that ever since then I’ve said, hey, you better do what you’re best at. And so when I got into venture capital I said I’ve got to position myself in venture capital and nobody wants to talk to me because I was a general manager. And I thought a pretty good one. But Andy –
RB: Had other opinions.
BD: Other opinions. And I now know I’m – was a lot better than Andy gave me credit for, but that’s only a personal evaluation. But everybody – I had the reputation as a marketing guy, so I – then I wrote the book Marketing High Technology, which I spent a year, since I knew I was leaving Intel at that time, outlining it on an airplane, and I wrote the book a chapter a week from those outlines. And, you know, it was a dumb idea. And when I say dumb idea, I realize how hard it is to get things published these days. But I had met Ted Levitt through Dave Norman and I showed the manuscript to Ted. Ted was sort of the dean of industrial marketing at Harvard, and he said, boy, I really like this. And so I’ll introduce you to my editor at the Free Press. And the Free Press liked the book. And so we – the book got published, and I had some wonderful marketing title for it. And Bob Wallace at the Free Press said the title should be Marketing High Technology. And I said, well, that’s pretty limited. And he said, that’s what the book’s about. [laughter] And so the book became really successful and it sold in Silicon Valley one twenty-eight and around – in Texas, and there was never a bad review written of the book. It was absolutely incredible. I forget the name of the person, the sort of the general book review thing. But they – there’s somebody that has [inaudible] – Kirkus reviews. Wrote a great review. The Economist picked it up and wrote a great review of this crummy little focused book. It was a terrific review. And I decided that one of the reasons why people really liked the book was that it was about my experiences. And I would say this is something dumb I did. And as long as I was willing to admit that I did dumb things, they didn’t have to say that I did dumb things. [laughter] So anyhow, the book became a success. And ever since then I’ve basically sort of branded myself as a marketing person. And I now do a lot of philanthropic work, and I work on the marketing problems of philanthropies because philanthropies are usually bad at marketing themselves. So that discussion I had with Andy when I was forty-five years old really shaped the next thirty-five years of my career. And that was the kind of effect that Andy had on people. And I presume other people have stories as well that are similar. But it was Andy’s ability to be direct with people and then present an alternative, and it – I think that was one of his greatest skills.
RB: Was certainly spotting the strengths and weaknesses of people is fundamental to getting the best out of people –
RB: – and helping them, you know, do what they’re best at. So towards the end of Intel then, so the bell was ringing from Kleiner Perkins. You had venture capital in your head. And I guess Tom was saying come over and join KP.
RB: But you didn’t.
BD: No, I stayed at Intel.
RB: And ultimately when you left Intel and moved to venture capital successfully, you still didn’t join KP. It became you and a partner, Larry Mohr.
RB: And Mohr Davidow was formed. Why did you do your own deal rather than joining Tom?
BD: Well, you know, I was fifty at the time. And, you know, career transitions when you’re fifty, even thirty years ago, were not ideal. I remember when I went to work at Intel, I was thirty-eight, and there was a discussion within the company as to whether I was too old to go to work at Intel. I know that.
RB: Today it would be twenty-eight. [laughter]
BD: You’re right. [laughter] And I had worked with Sequoia, and I had worked with Kleiner Perkins, and I had been positioning myself to do these things, and I had also been a venture investor. I brought Kleiner Perkins one of the first deals they did, which was a company called Antecna, and I had invested five thousand dollars in it. And that’s what really got me interested in venture capital because I made forty-five thousand – I made fifty thousand dollars on the deal. I thought ten to one on your money was pretty good. It wasn’t a great deal. I mean it was a great deal for me, but – and then Tom got me involved in Tandem Computers and all those things. And at that time there was – then there was a question of, you know – and I thought I really only want to work three days a week. I want to do other things. And so that wasn’t too appealing. And so I was walking around talking to people like that and Art Rock introduced me to Larry. And Larry said, oh, yeah. Come on, work three days a week. That lasted for the first week. And so –
RB: Became six days after that.
BD: Right. [laughter]
RB: [laughter] So that became a pretty successful venture capital company. I mean I – for me, looking in from the outside, I mean I’ve come to talk to you about money over the years, and Mohr Davidow for me was always a high integrity place to go, and obviously a semiconductor knowledgeable VC with Bill Davidow there. So – and things turned out pretty well for you as partners.
BD: Yeah. I basically stayed with Mohr Davidow as a general partner from Funds One through Five. And we began to transition over to Nancy Schoendorf and John Feiber, and they joined us – I think John maybe joined us at the end of Fund Two, but it may have been Fund Three. But Nancy and John were in place in Fund Three. And then they took over basically – Larry left in Fund Four and I – when I got to be sixty-five, this was the year 2000, and the entrepreneurs were coming in and, you know, it was sort of bubble time, and I thought I don’t want to be investing at these prices and sticking with these companies for the next – I was always very early stage, so you lived with your companies for a long time. And it – on top of that, I didn’t like the attitude of the entrepreneurs. They’d come in and they’d say, well, I want to be a serial entrepreneur. And I thought, gee, the thing that made Hewlett-Packard great, the thing that made Intel great were that the entrepreneurs stuck with the things. And after they got them public, they tried to increase the value of their positions. And I didn’t like the entrepreneurs who were coming in saying I’m going to start this company, and after I get it running I’m going to go off and do something else. And – I mean that was the attitude. So I thought that would be a great time for me to step back and write more books and do all these wonderful things and do more philanthropic things. So that was the – sort of the progression there.
RB: Hm-mmm. So at Mohr Davidow, I think – I’ve noticed that you had – let me call it a general philosophy of put in fairly low amount of money, get the risk out, and then once you felt comfortable the risk was out, then open the faucet and back the company heavily. And did that work well for you?
BD: Well, I – you know, I talked with Tom. And Tom and I were extremely close friends for a number of years. Then he moved up to Belvedere and, you know, we didn’t see that much of one another. And I talked with him and I said, Tom, what’s the real secret? And he said put a little money in up front, get the risk out, and then if it’s going well, pile in. Now, what was happening in 2000 was that people were piling in with the money, before –
RB: Too soon.
BD: – before they got the risk out. And, you know, the other thing that happened in venture capital is it changed so. And, you know, we don’t even think about this, but I was involved in five funds at Mohr Davidow. I did twenty-five deals – roughly twenty-five deals over the period of 1985 to 2000. So one and a half deals a year. I invested a total of seventy-five million dollars roughly. And I returned – well, and today a venture capitalist puts that out in one deal or two deals. Right? And I returned seven hundred and fifty plus million dollars. So, you know, my goal and the way I worked was this mode. You know, now the – and the industry flipped. And I – you know, and – you know, I think that – I didn’t want to operate that way. And I – and so it –
RB: You had a more cautious approach –
BD: Well, [talking over each other] –
RB: – a lower deal rate –
RB: – but a higher confidence on the return.
BD: Well, I – maybe not – I mean it – if you look at the record, and I’ve got it on a spreadsheet, I – and I’ve looked at the record a number of times. But all of your returns come from the deals that you do better than five-to-one on. And I think – I’m going to throw out a number, but I think eighty to ninety percent of my returns came from deals that I did better than ten-to-one on. The remaining – you know, ninety-eight percent of the return came from deals that I did better than five-to-one on. I had – my largest – two largest investments, and deals where I lost everything. And so what I learned from looking at that was that the biggest mistake was to fall in love with things that weren’t succeeding and saying I could save them. And that the reason why the good deals didn’t require a tremendous amount of money was that they were going along and becoming successful and they didn’t have to keep coming back asking for more. Now, today that is so different. And it – and I think people are operating and doing quite well, but my goal as a venture capitalist was to earn five percent more than the S&P 500 to justify the risk, and to return – basically I used to tell people four times their money, and hopefully six times, for the funds – for the funds. Today you look at venture capital firms and they’re talking to investors about returning a couple times their money, maybe three times their money. And, you know, this can lead, still being very high return on investments depending on how long you’ve got the money out there. I know I once did an exercise for our partners, and I said, well, I showed them that if I had a venture firm that invested twenty percent of their money for five years, and in the first year returned eighty percent of their money and never returned a penny more to them, they would have gotten back eighty cents on the dollar. But if you did an ROI calculation, it turned out to be I think forty-six percent because you gave the money back [laughter] – money back early. And so the goal was to achieve a multiple for the partners of money and earn a respectable ROI. But ROI does not necessarily mean something good for a limited partner.
BD: I mean you want both.
RB: Maybe there’s another book there, Bill, sometime of the old versus the new (?) venture world, and I’m sure that would be extremely interesting. So now, I mean the semiconductor industry which you grew up with and helped to formulate has radically changed.
RB: I mean today huge consolidations, of course. Good ol’ names that we know well – Fairchild, National, Motorola are gone. Intel prevails. Intel prevails. So do you think it will – is Intel a forever, or is it something that Apple might be interested in?
BD: Well, I don’t know as – oh, I hadn’t even thought about Apple being interested in them. But, you know, I used to look at IBM, and IBM was unassailable. Nobody could knock off IBM. Nobody could knock off DEC. I remember growing up, nobody could ever conceive of knocking off GM. And if you look at it, I mean GM had over fifty percent of the domestic market. And these franchises all have difficulty remaining there forever. Microsoft, nobody could knock off Microsoft. And Intel had a monopoly position in a growth industry, and is – I’m going to say fifty billion in sales. I don’t know what the sales are – forty-five, forty-six. And that industry is not rapidly growing anymore, and people are going to chip away at their monopoly position. And the challenge for Intel, if you’re a fifty billion dollar company, is to find other ten and twenty billion dollar opportunities. And finding those opportunities are difficult. And they’re difficult if you’re Intel, they’re difficult if you’re Apple, and they’re going to be difficult for Facebook and for Google as well. And so I think that Intel will continue to be a great company, I don’t know as it’s going to be a rapidly – as rapidly growing, or have the influence that it used to have.
RB: It used to have. Yeah. Yeah, it will be interesting to see how that pans out. So Bill, I know you’re retired and enjoying your time now. You’re on a number of boards of directors pursuing your individual interests. I know you used to play tennis.
BD: Still do.
RB: Still do. That’s better than me. That’s – good job on that. Are there any other things, Bill, that you’d like to sort of get plugged into the little summary here that we’ve missed on that you think are really important for Bill Davidow or the industry?
BD: [laughter] I mean to me I think you’ve covered it very well.
RB: Okay. Okay. Great. Great. Great. Well, listen, Bill, it’s been a pleasure to know you over the years. And we’re both still here. We’re on the wrong side of seventy, but we’re both still here. That’s good.
BD: Maybe we’re on the right side of seventy.
RB: Well, you know [laughter] –
RB: You’ve left your footprint on Silicon Valley as one of the marketing gurus of the semiconductor industry. And today you’re enjoying yourself with your family. So I’d like to thank you for your time today. And this story will be posted on Silicon Genesis to Stanford Libraries. And I look forward to continuing our relationship.
RB: Thank you, Bill.
BD: Thank you very much, Bob. Appreciate it.
[End of Interview with Bill Davidow – March 23, 2016 – Silicon Genesis]