Bernie Marren has a long and notable career in semiconductors. He began at Fairchild Semiconductor and later served as CEO of distributor Western Micro Technology and MOS pioneer American Microtechnology. He was a founder of the Semiconductor Industry Association (SIA) and served as it's first president. At the time of this interview he was president of OPTI, a chip set and technology licensing firm. In the interview Bernie discusses the role of distribution in semiconductor sales and the increasing importance of intellectual property protection.
RW: Tell me a little bit about growing up.
BM: Okay. Well I grew up in Chicago , Illinois , in the city. I finished high school there. And the family had limited funds so my options in going to college were limited college was relatively easy to get into in 1953 because there was a shortage of students. The GI's had all gone through college after the GI bill and now we were left with the depression era kids. And there was a shortage of them, so most colleges were recruiting students. There was no SAT test so almost any college was willing to take you if you had decent grades. I had decent grades. I probably would have been accepted at most any college in the United States , as could most young men of that era. Of course I chose electrical engineering as my major. I went to IIT, Illinois Institute of Technology which was a Chicago based school and I would not have to pay room and board. So that's why I chose that school. And electrical engineering I think I chose that because I went to a high school that was very much into liberal arts (I had eight years of foreign language). So I didn't have as much science as most of the other freshman but it did not take long before I caught up with everyone and did quite well and finished. I was going to go into the service but I flunked my flight physical for the Air Force so they told me I could only be a weather officer. I wasn't interested in doing that so I decided to go to grad school. I went to the University of Cincinnati and got a job with a company called AVCO, which was an aerospace company. Today they're a financial company a bank and loan company. They hade two divisions Crosley and Chance Vought in Electronics and Aerospace. They were also the prime contractor on the Polaris Missile nuclear fuse. I was assigned as a young design engineer to work on that radar fuse.
RW: Well now, what did your folks do?
BM: Well, my father had an interesting background. He was born and raised in England . He came over here and went to work for this uncle who was superintendent of construction for railroad bridges for the Illinois Central Railroad. While working for his uncle he had a serious accident and fell down a well that had been dug to support a bridge. He made a lot of money in an insurance settlement and went into investments. He purchased a bank in Chicago and was the president until it failed in 1933. My mother was born in New Jersey and had moved to Chicago as a child. My parents were married when my father was about thirty-five and I was the oldest of the family of four. After his bank failed, my father had lost most of his money and the only job he could find was as a bus driver. He stayed a bus driver until he was sixty some years old and then went to work for Mayor Richard J. Daley as the head of the Bureau of Assumed Names. This Bureau was set up to review any business to see that it had not gone bankrupt under another name. It also issued Notary Public licenses. He ran that Bureau from the time he was sixty-four, I think, until he retired at seventy-nine. He died at eighty-one. So that was what he was doing.
RW: So why did you pick electrical engineering?
BM: Well my dad had a friend who worked for the bus line who was an electrician. He was a young man from Ireland . And I believe he had a limited education but he thought the grandest thing he could ever become in the whole world was an electrical engineer. I had worked for him during my summers in High school. So when I entered electrical engineering, I truly did not know much about it and didn't know anyone who was an electrical engineer but I knew several electricians. He had me work during summers in the Midwest they have what's called detached garages. So you'd build a house and you wouldn't have a garage when you first built your house and then as you saved up a little money, you'd build a garage and it would be a the back end of the lot. Chicago has alleys behind every house. So the garage would be facing on the alley. These garages were typically built by carpenters and they'd have to hire an electrician because the city building code required them to have a light and a duplex electrical outlet. My job for the summers was to dig a trench from the house to the garage, wire the garage, but not connect it. Then on Saturday one of the electricians would come by and he'd wire several garages by just connecting the Edison 220 Volt to the cable that I had trenched through into the basement. That was the total exposure I ever had to electricity or electrical engineering. So in my first year of college the math and science were a bit beyond my background but I was able to pick it up quickly.
RW: So, you're working at AVCO as a design engineer
BM: As a design engineer I was designing radar circuits using tubes. This was 1958 and I started grad school that fall. I switched to solid state physics. I just picked semiconductors as they were of interest to me and I didn't know much about them. When I was studying electrical engineering, even though I was in the electronics option, we were not exposed to semiconductors. There was only a chapter or two on semiconductors because the faculty wasn't too versed on it either. And the particular circuits I was designing were tubes because the government would not let us use semiconductors for the nuclear fuse because they were afraid the radiation would conk them out. So the first circuits I designed were using miniature vacuum tubes. As I went on in the graduate school I was going nights and on weekends and in the summers, I had time off to take a class or two during the day. We were working in Cincinnati , Ohio , and the company decided to move headquarters for that division to Richmond , Indiana . Now Richmond is on Interstate 40, right on the Indiana, Ohio border but it would've made it very difficult for me to finish grad school at night because it was a sixty some mile drive each way. So about that time, I had an offer to work for Philco Ford as an engineer. They were in Blue Bell, Pennsylvania, making semiconductors. I was very interested and I had negotiated with them for a long time because they are a very, very careful company is about the best thing I could say. They interviewed me and interviewed me; I had probably thirty interviews with them. They told me the job I was most interested in I wasn't qualified for so they wanted to give me a different job I turned them down. Then I'd hear from them a couple months later and they wouyld go through the same routine. To make a long story short, after nine months, they never filled the position I was interested in and they finally offered it to me. About this time, I got a call from Fairchild and they asked me to talk to Dr. Victor Grinich (a founder of Fairchild Semiconductor). Grinich was going to be on the East Coast so he came to Cincinnati . He asked me how much money I needed and I told him the offer I had received from Philco Ford. He said, well, we really can't do that well and he asks would I consider a hundred dollars less than that a month. I said I was sure that would not be the determining factor. So he immediately goes to the punch line and asks if I would consider two hundred dollars a month less. So, sure enough, I get a letter from Fairchild a week later with an offer of one hundred dollars less that I was going to make at Philco Ford. I took the job and I joined Fairchild in 1960. And I was assigned to Vic Grinich. Vic told me he was going to give me a special assignment but it was a secret. It was so secret
he couldn't tell me what it was. So I finally weaseled out of him what the job entailed. He said it's to train you so that you can train the salesmen on how this new product works but we can't tell you what the product is yet. It turned out later to be (they never did tell me) the first integrated circuit. I was supposed to go to the Midwest or the East Coast and be based there and train the East Coast and Midwest salesmen on how the integrated circuit worked. The people on the West Coast would be trained at the factory and I would do the training in the field. Well, to make a long story short, there was no product the product was delayed and delayed. In the meantime, they asked me if I would move to Minnesota and call on some customers there. They had a customer called the Aero Division of Minneapolis Honeywell. So I went up there and I met one of my fellow graduates from IIT who was a manager at Control Data. He introduced me to Seymour Cray. Seymour was building the world's fastest and largest computer at the time called the CDC 6600. We convinced him to design in our 2394 core driver transistor. Each machine used a half million of these transistors so it was quite a large volume order five million transistors. Shortly thereafter, Fairchild was afraid I was going to leave to go to work for some other company and they offered me a job on the West Coast. They assumed I would not want to stay in Minneapolis long term. They found someone they wanted to hire who was a native Minnesotan. The individual they picked, Earl Chiswell, who was working for Rheem Semiconductor. Tom Bay, Fairchild's Director of Marketing, and Don Rogers, Fairchild's Sales Manager, came to interview him and found out he was making more money as a regional manager for Rheem than they were making as executives at Fairchild. That was the end of that plan!
Shortly thereafter, I was promoted to be the regional manager in Chicago and the manager of the Consumer Products Operation which was a fledgling start-up in Mt View. I stayed with Fairchild in various positions primarily in sales but also managing engineering in consumer products.
We designed the first solid state TV sets for customers like RCA, Admiral and Zenith as demonstrations of what you could do with a solid state chassis.
We did not understand the marketing dynamics. One of the key engineers at Zenith told me you are competing with free. What do you mean by free? He told me we do not pay for the vacuum tubes we use. They are given to us by the vacuum tube manufacturers so that they would get the replacement business when the tubes failed. He told me your product doesn't fail so you are not interested in giving it away.
We finally got our first semiconductor design in the UHF tuner. This design was dictated by government edict, which required all TV sets built after a certain date must have a UHF tuner in addition to the VHF tuner. This could be done quite easily with a vacuum tube! But the reason it was successfully converted to semiconductors was that the early TV sets used the UHF tuner as an add-on. The tuner was placed in the upper right hand corner of the cabinet above the VHF tuner. This was a very high temperature location in the cabinet. The vacuum tube had a high failure rate because of the heat due to its location. Although the tube itself was not expensive the problem was the service call. It was very difficult to replace the tube due to its location. So we provided them a solution that had no failure mechanism at the same price, seventy cents a device, and we converted the entire industry overnight to a solid state tuner for UHF operation. Then slowly with our design techniques, we showed people how you could build VHF tuners using semiconductors and that quickly started growing with companies like Zenith, Radio Oak Electronetics, General Instrument and Standard Kollsman to be the dominant source but never into the major chassis. Major chassis really became mostly Japanese. And
the reason for that, the Japanese had to come up with a solution that didn't require any replacement parts because they didn't have an infrastructure. You could get vacuum tubes at every drugstore in the Midwest or the East or West Coasts . You bring a bag of tubes down, you'd plug them in and find the bad one, replace it yourself. So it is relatively easy to repair your old vacuum tube TVs, where solid state TVs had to be very, very reliable coming from Japan because it would take months sometimes to et parts for them. So the Japanese pioneered both reliability and solid state and slowly eliminated all the US manufacturers.
RW: Yes, wasn't Zenith the last to go?
BM: Zenith was the last to go and it was bought by one of the Korean companies.
BM: Yeah, and Zenith had a very, very enviable reputation for quality but they were still vacuum tubes.
RW: Yeah. Hmm.
BM: After that, consumer products spread throughout the world because we started making plastic transistors in Hong Kong . And those plastic transistors were used in AM and AM/FM radios and car stereos and tape recorders and everything else. Then slowly we expanded the operation to Australia . Then people would say why would you move to Australia ? Well, the government passed a law in Australia that all semiconductors that were not made in Australia would have a duty of twenty percent or twenty-five cents, whichever is higher. We were selling these transistors for an average price of a nickel so we opened a new operation in Melbourne , Australia , and charged them a quarter and got a hundred percent of the business because no one could bring a part in without paying a duty.
Finally after several years, the Australians figured out this was a no-win deal and changed the law to eliminate the duty. The plant wasn't large enough to make economic sense and therefore was closed.
One of the first electronic products we made for the automotive division was the voltage regulator for the Ford Motor Company. We shipped the first ones which were wildly successful and then we ran into the margin problem. The prices that Ford wanted to pay versus our costs made the product very uneconomical. We weren't ready for the automotive business at that time due to how much we could sell products into the computer market and other innovative consumer products.
In the summer of 1968 I was sent to Italy to buy out one of our two Italian partners in a company called SGS. My job was to get one of the two Italian companies, who each owned 1/3 of SGS, to sell out to Fairchild. Fairchild already owned a 33% interest in SGS due to a earlier negotiation by Dr. Robert Noyce who convinced SGS to transfer 1/3 ownership to Fairchild in return for a license on our technology. The two Italian partners tended to make decisions without consulting Fairchild. As an example, they would build a new plant in Germany and not feel it necessary to consult us beforehand but would obligate us to 33% of the plant costs. The relationship became contentious. I was unsuccessful in getting Olivetti or Teletra (a large electrical cable company in Italy ) to sell us there interests. No matter what price I offered, I was told the price was too low, so at one meeting I made an offer that was unauthorized I juts wanted to see if they would take it and once again, they turned it down as insufficient. My counter was to tell them to buy us out at that price which they did. Fairchild accepted the offer and we used the funds to open the Fairchild in Weisbaden , Germany , and we hired as many of the SGS sales people as we could.
When I returned from Italy in November of 1968, there had been a lot of changes, with more to come. Jerry Sanders left the company the week I returned from Italy . He had a run-in with Joe Van Poppelen, who was the new Vice-President of Corporate Marketing. Joe had been hired that summer from a UK subsidiary of IT&T Semiconductor.
RW: Well, since you were very early at Fairchild, what's your impression of some of the Founders, Bob Noyce, Gordon Moore ..and people like that?
BM: One of the things that impressed me the most, though I had a very limited career in industry prior to that, was the youth of the management. Everyone was so young. I was twenty-five years old but most of the people were not much more than five or six years older than I. And the intelligence of the people I met my first real contact working for somebody was Vic Grinich, a very brilliant engineering manager and a brilliant engineer. I had many occasions to talk to Robert Noyce. And Noyce had a great ability to eommunicte with employees. He'd sit down at your desk ask you what you're doing, what you're working on, see if he could be of some help. Some of the founders were fairly quiet, like I remember Gene Kleiner who I got to know quite well. He was probably the oldest of the founding employees of the company. But he didn't talk to you a lot but if you asked him something, he was extremely helpful. Everyone at Fairchild really would pull in together to get this thing off and running. And the thing that I noticed most was most of the equipment they needed, they had to build themselves. There was practically nothing you could buy. In fact, I remember Bob Noyce working personally on step and repeat aligners using commercial lenses that he'd bought in San Francisco . And he made his own step and repeat cameras. From there, almost everything that was done in the industry they had to generate themselves the equipment, that is. We used to buy chemicals in barrels, fifty-five gallon drums. They'd be spilled all over and the accuracy of the measurement and mixing ingredients was always questionable. And there was a guy by the name of Mike McNeilly who was a young salesman who was also a chemist. He started mixing chemicals very carefully and put them in two and a half gallon jars and delivered them to us. Now they cost a little more but they were already premixed for us. So the loss was less and we started switching to those and that became the company called Applied Materials. Mike founded Applied Materials based on mixing chemicals initially. Now it's one of the world's largest equipment manufacturers.
BM: There was a lot of that going on at Fairchild at the time. I stayed with Fairchild until 1972. One of my lat jobs was working in divestitures for Fairchild where we were selling off divisions on the East Coast primarily for cash. In 1978, I mean 1968, Fairchild hired Dr. Hogan, but by 1970, in the first quarter of 1970, we made five million dollar profit. For the year, we lost twenty million. So in the last three quarters of the year we lost twenty-five million dollars. The company was in jeopardy; it needed cash. It was decided to sell most of our East Coast operations. Most of these divisions were losing money. Sherman Fairchild had returned to the company as Chairman. He also hired Richard De J. Osbourne .Sherman brought him in from IBM. Very brilliant man but kind of a caustic personality but he was smart. My job was to work for him. We packaged land that Fairchild owned. Fairchild owned a lot of real estate in the New Jersey Meadowlands and other places in New Jersey . We would package these land parcels with businesses we wanted to unload. And whoever bought them had the problem of terminating whomever they wanted to in order to get the land. And that's how we got rid of all these properties at a price above the book value of the business.
After selling off these divisions, I came back to Mountain View and was involved in the Illiac Memory Program. Fairchild had a contract to build an integrated circuit bipolar memory system for the Illiac IV Computer. Our major problem we couldn't get any yield of 100% good bits on the devices we were manufacturing. We had a gentleman there by the name of Rex Rice, very smart fellow, kind of controversial, but very bright. Rex came up with this scheme that plotted every memory bit and created a matrix design so that the computer couldn't address a defective bit. We only needed to ship one memory system to the University of Illinois and we shipped it with no working parts. Rex came up with a scheme where he could plot every bit on every RAM.
RW: They were called partials.
RW: Right, partial memories. I then got an offer to go to work for Intel to head up the Memory Systems Division. Bob Noyce had started this company in 1968 when he left Fairchild. And he had developed a DRAM called the 1103 and the 1103 was a one thousand bit three power supply part. He took that part and he started making banks of memories to sell to people like banks for add-on memory. Well after my experience with memory and the bad bits, when Bob talked to me about that job, I couldn't think of a job I disliked more than being involved in memory systems. So instead of that, I took a job with American Microsystems in Cupertino, California. I replaced the founder Howard Bobb who I had worked for briefly when I was at Fairchild in 1960. He was the sales manager working for Tom Bay. He had left to found General Micro Electronics which later became Philco Ford Semiconductor. Then he founded American Microsystems with two other people. I took over as the CEO and ran that until 1976, at which time Bob Noyce asked me if I would be interested in starting a trade association to replace the American Electronics Association which Fairchild and most of the semiconductor industry on the West Coast were members.
RW: They were interested primarily in representing the interests of the large equipment manufacturers. There were a lot of political issues facing the industry and a lot of dumping problems. So what we did, we formed this trade association and became the monitors of the statistical program that the AEA was doing. There was a lot of bad blood initially between the AEA and the SIA. I really was not interested in becoming a long-term employee of a trade association. So I told Bob and Jerry Sanders, Wilf Corrigan and Charlie Sporck who were the founders of the SIA, that I would help them find a permanent successor but I wasn't interested in doing it for the rest of my career. It became a very difficult job to find a successor because initially I did not realize each member of the board had veto power and could blackball any candidate. So it wasn't a consensus- building thing at all. The best candidates I found had somehow irritated one or more of these members in their past life and would immediately get voted down. There'd be no further discussion on them. We finally picked Tom Hinkleman because very few people knew Tom and he hadn't really scored many enemy points. Wilf Corrigan of course knew him and Jerry Sanders also remembered him from his days at Motorola. And, he didn't have any objections. Charlie Sporck didn't know him and Noyce didn't know him. So he passed muster and he got hired as the GM of the SIA. But the thing that I think was important about the SIA it had a very difficult start. The dues were two hundred fifty thousand dollars a year, plus you had to put up two hundred fifty thousand dollars in an interest free loan as a new member. Business was tough that year and people said they would do it but no one had done it yet. I went to Texas to meet J. Fred Bucy, the CEO & President of Texas Instruments, to convince J. Fred to join the SIA because before that we were really a Silicon Valley boys club (everyone was from Silicon Valley). J. Fred told me he'd see m in the morning. I arrived the night before, went to the lobby to meet J. Fred Bucy at the appointed hour and he was not available. I waited and waited he finally saw me at four o'clock in the afternoon to explain to me that he had no interest in joining any association of the West Coast Semiconductor companies. With the standoff I left and reported to the board and explained to them we better get a new member. The strategy was to hire John Welty, General Manager of Motorola Semiconductor to be the Chairman. John could not authorize membership himself because he was working for Bob Galvin, the President of Motorola. I then went to see Bob Galvin and convinced Bob that we needed John Welty to be the chairman for his leadership qualities. With Bob hearing that, he immediately said they we're joining and gave me a check on the spot for five hundred thousand dollars the first money we had ever seen. With that, I returned from Phoenix with the check and the promise that John Welty was going to be a new member and also the chairman of the board, and it was unanimously adopted by the board. We then hired Tom Hinkleman and I left to form a new company Western Micro Technology with Marshall Cox who had been the President of Intersil since late '68 after he left Fairchild. We started this company, Western Micro, and we were in a new field - high-tech distributor. We were going to specialize only in semiconductors leading edge semiconductors. The first line we asked for and received was from Dr. Noyce - he gave us a franchise for the Intel line. We didn't even have an office yet and we had received the okay from Bob, verbally, that we were going to be the new distributor. He went back to his office and explained to the marketing people that he had appointed us to be the distributor and all hell broke loose. About two days later Bob called me and said let's have lunch. He explained to me he wasn't going to franchise us for at least six months to a year because there was too much heat coming from the marketing people who had promised the franchise to other people. And here we were not even in business yet and we were going to get a franchise. We were struggling to get some franchises and we received several. One of the things we recognized was there was a new product coming out and that was the next generation of DRAMs. It was a 4K, 4096 bit DRAM. There were several companies making it. One was Mostek and they made a part that had 16 pins. Most of the industry was going to 16 pins. But there was a company in Japan that I was very familiar with because they were a big licensee of Fairchild, called NEC. NEC made a version that was 22 pins. The 22 pin version of the 4K was superior in many respects to the 16 pin version and became widely adopted. In fact, several of the companies using it at that time were Intel, National Semiconductor, as well as several of the big manufacturers around the world like Apple in their early PCs. We got a franchise to sell DRAMs for NEC. The Japanese had an approach where they burned in everything, they're very quality conscious. They had a certain specification they burn it for so many hours and to so many degrees Fahrenheit and under power. The American manufacturers that were using these also had their own specs. Their specs would call for different burn-ins; everyone had their own burn-in spec. The Japanese were very inflexible on their quality. So if you tried to burn their part in and it failed, you lost your warranty. We couldn't convince the Japanese that was fairly foolish of them, but they stuck to it. One of the new manufacturers besides NEC was Hitachi and Mitsubishi. Mitsubishi was interested in joining the Semiconductor Industry and they were also a licensee of Fairchild's in Japan. I got to know the Mitsubishi people personally. They were going to set up their own operation and start building some of those DRAMs in the United States. They wanted to do their burn-in and I convinced them that we could do the burn-in to the customer specifications. I was able to hire a gentleman by the name of George Lu, who was the head of the test lab at Hewlett-Packard in Palo Alto and Durham, North Caroling, to build DRAMs and we were the back end. We set up and bought equipment and we did all the burn-in testing, marketing and shipping for them. That gave us an edge in the DRAM business. So quickly we probably were a distributor of every Japanese DRAM manufacturer going. We had Hitachi, Toshiba, Mitsubishi and NEC just about everyone who made DRAMs.
RW: Was that controversial?
BM: It was extremely controversial because the Americans were trying to prevent the Japanese from getting into the marketplace. And how they did it was they told distributors like Avnet and Arrow Electronics if you take on the Japanese, you can't have our franchise. Se we were hard-pressed to get some of the other franchises. Two of the other most valuable products at that time were the 7400LS and 7400 lines. These lines were primarily supplied by Signetics, TI and Fairchild. Motorola had not entered the business yet until the 74L came out. They subsequently started making LS. But there was one other line left and that was called IT&T. They were really a secondary line but because they had licenses from Fairchild and other people, they had practically every part number. They weren't very successful in marketing them because the other manufacturers usually packaged them with something else and they didn't understand the business. They only had one other distributor which was Avnet and they weren't intimidated by the other manufacturers of semiconductors so they gave us a franchise. Shortly thereafter they went out of business. In fact, they went out of business just before we got our first shipments, which was a real disastrous blow to us. We had made a big fanfare about signing these people up! Then they also told everyone that if we needed to buy parts, we could buy them from Avnet their other distributor because they weren't going to get around to it. We decided we had to do something fast. I called every manufacturer that all their business was going to be transferred to IT&T until they get rid of their inventory. Well they got furious, called Avnet and said if you buy those parts from IT&T, we're going to pull your franchise. So they really weren't but the threat was enough that Avnet decided not to bother. They called IT&T back and said we don't want your inventory. We stepped in they told us they had about three or four million parts for us and they started shipping the parts kept coming and coming. Before they finished, they shipped us over forty million parts. They had no control over how many parts they had really made. And, of course, every time they'd ship us more parts, we'd tell them the price had to be lower. We were buying gates for about three cents a piece. We were buying MSI parts for a dime. We had a tremendous price advantage over anybody in the business, including the manufacturers. This ran for about a year until we went through our forty million parts. We were able to package that with DRAMs and other things and build quite a successful business. But distribution was very important to us in semiconductors because the average manufacturer primarily bought most of their leading edge parts from distributors, not from manufacturers. As they got into high volume, like in DRAMS, they would switch to OEM agreements because that's the thing that made the most sense. They didn't need the service a distributor provides if you're building a manufacturing line.
RW: Well, where did the reps fit in, the representatives?
BM: Okay, the Sales Reps ..instead of hiring salesmen themselves, they would put on Sales Reps. Most of these were ex-salesmen of various semiconductor companies. That's where the reps came from. They would be able to handle more than just a semiconductor line. They might be able to handle a line making potentiometers, have another line making printed circuit boards. So they could call on the customer and have more to sell than just semiconductors. They would have multiple lines of semiconductors that were not conflicting. They might have semiconductors like analog circuits and another line selling DRAMs and another one selling 74LS. They would be paid a commission and the commission would range from about five percent initially down to a low as one percent over time. In other words, if they got large orders, they got a lower percentage. But they also got an override of what the distributors sold. So the distributors would report every month on what's called a point of sale report. They'd put down the name of the customer and then that was to be sent into the manufacturers. From that there'd be another commission report where they received a percentage of what the distributors' sales were in those accounts where they were franchised. But the thing that made distribution different was the early days distributors used to buy the parts and then they had what's known as return privilege if they had too much inventory. And the manufacturers, if the distributor had the money, of course they didn't want to take the risk, would over ship the distributors. You needed to make the number for the end of the month you needed to ship the distributor another hundred thousand dollars worth of inventory, the distributor would take it. If he didn't sell it he would then turn around and return it. Well that made for very chaotic sales results. All of a sudden when they thought they were going to able to meet a certain sales goal, some distributor would want to return something for credit. Since they had taken it as a sales before, it's a negative sale coming back to the manufacturers. So the manufacturers, in order to get around this, went to using the point of sale reports where they showed what the distributor had sold to this customer and that became the sales report. So if they shipped inventory to a distributor, they didn't recognize it as a sale until it was reported on the POS report. They would be able to check how much inventory a distributor had and maintain the inventory. Then when they had a price change, they would go out to the field, count the parts and give them credit to reflect the price change. If a distributor wanted to get anything returned to the factory, he had to give them an offsetting order of the same size. So it evened out the sales results. Numbers would be up and down by fifty percent some months, so it evened out the flow of sales thirty/thirty/forty on the quarter. Thirty percent the first, thirty percent the second and maybe forty percent the fourth because the fourth
the last month of every quarter everybody pushed to maximize sales because that's the quarter they had to report. But that's how distribution works. We got into selling more things than just semiconductors. We started selling equipment. We found the equipment sales were less fluctuating and there was less competition. We became the largest distributor of IBM midrange computers. Not PCs, the midrange stuff. And the reason for that they had only about three or four distributors in the whole country, not eight or ten. So it wasn't a wild fluctuation in pricing that you'd get and they controlled the pricing at IBM. But the distributor business started to die out as more and more of the manufacturing moved offshore. The necessity of distributors was only for prototypes then. They weren't getting any large multi item orders as much because what they would do is just get the new parts. The prototypes would be designed here at the same old companies that designed them before, Cicso or HP or somebody like that, but the production was done offshore. Like most computers would be designed in Palo Alto or somewhere if they were at HP or Compaq down in Texas. But the actual manufacturing was done in Taiwan. The boards would be assembled in Taiwan and then they'd be shipped to Mexico and then put into a cabinet. A microprocessor was inserted in the PC board and the disc drive was attached and then shipped to a customer or to a warehouse. So that changed the industry, it changed the growth of distribution. In fact, it leveled off very quickly and several of the distributors went out of business. It became a very shrinking market and that's what happened to distribution.
RW: Well, what did you do?
BM: Well, we sold our component business and changed our name to Savoir Technology. We then specialized in selling mid range computers and work stations. IBM was our main line and we grew to $800 million in sales. We became the largest distributor of IBM equipment and we sold our business to Avnet, who was the second largest. The combination of our sales and Avnet sales made them by far the largest, so they became the largest Avnet's always interested in being the largest in the area that they're in.
So I went out and started another company developing there were two developing Bluetooth. We made the first laptops that could talk to each other at the '99 Comdex Show in Las Vegas. We were going to sell the company, we had no manufacturing whatsoever, but we could build prototypes. We would build the prototypes we had a very small team. We decided to get outside funding to increase the team so we could set up a manufacturing operation and start building these circuits or do we sell it? We thought about it for a couple days and we decided to try to sell it. We estimated that we could sell
the business for maybe ten million but perhaps as much as thirty million dollars. There's a company in New Jersey that were building DSL equipment they had seen us at Comdex and were impressed. They came in and made us an offer on the spot, all stock, for two hundred million. We almost fell off the chair. We agreed and were ready to sign. But after we signed, we found out he was not going to give us registration rights on the stock. The stock started going down in price. The two hundred million became one hundred fifty million, then down to one hundred million still a great price. But we pulled out because we couldn't get the registration rights. Then we decided we'd only sell for cash and the cash had to be paid in full. The two companies who were most interested were National Semiconductor and Cypress Semiconductor. They got into a bidding war and National finally bid one hundred thirty-five million cash for the business and we sold to them. They hired the engineers, transferred Bluetooth to their operation, very expensive operation for Bluetooth and then transferred the engineers into their cell phone group. At that time, National was a manufacturer of basebands but not radios. These fellows were mostly RF people so their ideal, they were going into the radio end of the business. Shortly thereafter they found that TI and people like that were far ahead of them and they really weren't competitive in basebands so they shut the whole thing down, laid everybody off, including the engineers they just bought and paid all this money to. So after I left, I joined OPTI. OPTI, founded in 1989, was a manufacturer of chipsets. They built the north and south bridge for the chipsets. They became the original chipset manufacturer for the Pentium computer. Intel, while they had the capability in-house to probably build chipsets themselves the Pentium was such a radical change for them they were afraid that if the Pentium didn't work or the chipset didn't work, they couldn't sell the part. So they decided to put all their eggs on the Pentium and let us do the chipset, which we did. The launch was quite successful. Then Intel came out with a second generation of Pentium and we were supplying the chipsets for that. But they started to come out with chipsets themselves because it was a natural evolution for Intel to go into the chipset business because the factories that built Pentiums when they were obsolete would be ideal factories to build chipsets that didn't need state of the art processing. But they weren't very successful in ousting us from the sockets. So when they came out with the Pentium 3, Pentium 2, I should say, they decided to license the bus. And they patented the bus, the Pentium bus, and said if we connected to it they would sue us. You have to get a license. So they gave licenses to a company called VIA and a company called SIS in Taiwan whom they thought were no threat and eliminated us from the desktop business. We were still in the notebook business but out of the desktop business. So I started looking at their devices and I said they sure look a lot like our old devices. We had patented a lot of the circuitry that we had designed as time went on and we found out that they were probably violating our patent. Now Intel also had enough patents of their own and Intel was a very successful company. I've competed with them at AMI and they are very impressive. So I didn't want to get into a lawsuit with them. I kept negotiating with them over the phone ..you know, I think you're violating our patents and they told me they weren't, so sue them. Well, I didn't want to sue them because we didn't have the economic muscle at that time to sue them. We finally convinced them to have a meeting and they talked about it. They told me that they didn't think they violated our patent but they would go through an arbitration meeting with us. If we hired an arbitrator it'd be binding and if our patent was valid, theirs was invalid we would resolve it that way and save ourselves a lot of money. So finally they came back with another offer and they said that, look, they really wanted to do this and if we will sign with them to do this arbitration, we'd get to pick the arbitrator, they'd pay for it. We didn't even have to split the costs because they wanted to get it behind them. That didn't seem to entice me, so finally they came to an offer that if our patent was invalid and theirs was valid, they'd still pay me a million dollars just for going through the exercise. So right away, I recognized they were more interested in solving it that I was maybe. So they finally settled and they gave us a license. They took our license and they paid us thirteen and one half million dollars and gave us the choice of any three patents they owned and a cross license across the board on all their patents. So it was a good settlement for us and we recognized that the reason we accepted thirteen and one half million was at that time there was the Hart-Scott-Rodino Antitrust Improvement Act of 1976 (HSR Act) and the two largest manufacturers of any product line had to go to the Justice Department if they license each other. They were afraid that we'd never get a license past the Justice Department because of all the controversy with Intel and all the people who were screaming about their performance. So in lieu of that, that's why they gave us a choice of any three patents they owned. And we never really did that because we didn't want to be a patent troll. We wanted to only go after things that we had invented. So we took patents and we went after companies like National Semiconductor and Nvidia. And currently we're suing both Apple and AMD for violating our patents. They're coming up for trial next year, as well as companies who violated some of our other patents. We're really a patent licensing company now at OPTI, that's our principle business. And we sold off our other businesses and I've been running it now for ten years.
RW: And does this make you really popular in the valley?
BM: No and I don't have any real problem with it. We returned over a hundred million dollars to our stockholders in royalties that have resulted in dividends to them. So they're quite happy with what we do. But I think people recognize the value of the patents. And other people, I notice, are licensing, spending more time on their intellectual properties and licensing their property too and they should. Now, as you can well imagine, the best defense on a patent suit is your own patents. You counter-sue the plaintiff on their infringement of one or more of your patents. And so what you do is the judge will say okay why don't you two guys just cross license each other and disappear. Well, one of the reasons we sold off all our operating businesses they can't counter-sue us. If you don't make a product, you can't be sued. So that's really why we're only in the licensing business. It's hard to be in both and make a career of it. But the semiconductor business has now grown into such a mammoth industry and it affects so many people in so many ways that most people don't recognize how important semiconductors are because people born after 1965 or 66 even as early as '60 don't know the world without semiconductors. And in 1960, I remember going down as a young engineer in Ohio when I first was hired, when I was waiting to find out what that new product I would be training others. Young engineers who were not much older than I would be lined up to see me just to explain to them how semiconductors worked because the concept of current amplification when you were trained on voltage amplification sounds like well why weren't they able to really understand it. Any they didn't understand how it worked. And they would just ask you to spend fifteen or twenty minutes just explaining how these things work because they were trained on vacuum tubes and they were quickly realizing they're becoming obsolete, especially in the aerospace industries.
RW: Boy, I guess. Well, you've been very successful in Silicon Valley. Now why does Silicon Valley work? I mean, how come around here we're such innovators?
BM: Fairchild really started it all. Dr. Shockley and Arnold Beckman had assembled a highly intelligent, well-paid group of talent at Shockley Semiconductor. This group was the initial starting team at Fairchild. This group attracted additional talent from all over the country. Fairchild created an atmosphere where people didn't ask you how many years experience you had or how old you were. You had the opportunity to excel. And it wasn't based on what you knew and who you knew; it's what you could do. And there were a lot of risk takers at Fairchild. People would take risks and there were some extremely bright people. I mean, I remember Vic Grinich and Vic had a great personality besides being a very bright guy. But Jean Hoerni was there and Jean was brilliant. Now he had a caustic personality. And there were some people who had both great personalities and great intelligence. Two of the people obviously were the two founders of Intel, Noyce and Moore. I mean, you liked working with Gordon. You liked working with Bob. They had true leadership qualities and they could attract talent.
While at AMI we were interested in hiring T. J. Rodgers out of grad school at Stanford. We had a tip that he was probably one of the brightest guys to come out of Stanford in a decade with a PhD. So Dr. Don Trotter, who was the head of R&D for American Microsystems was working for me, so I assigned Don the job of making sure T. J. comes to work for us. Well he'd come back week after week highly discouraged. I said, what's the problem Don? Oh, he says, we'll never get him. I said, shy? He said, well Bob Noyce is taking him up to the wine country, Napa, with him, wine tasting and Gordon takes him out fishing in his boat for salmon. He says, there's just no way, what do we do? I explained to him - we'll pay him more money. They're going to hire a lot more PhD's than we are this year and they're going to run into a salary scale problem. We'll pay whatever it takes to get him. So we hired him and Intel didn't get him.
RW: Well, it certainly is an environment where it is okay to fail, assuming you you're not stealing or doing something of that nature. But if you go out and you want to go out on your own and you try and it doesn't work
BM: You can try again.
RW: they'll hire you back.
BM: Some of them will hire you back and you get another opportunity plus there's something about failing you learn from. You know, people always think of failure, it's just a downer. Yes, it's a downer at the time but you did learn something, you learned what not to do and you don't do it a second time the same way. So, some of the more successful people in Silicon Valley weren't always a success the first time. I mean, it's the Silicon Valley start-ups made some mistakes. There were several of them that got into trouble primarily due to consumer electronics which they didn't understand very well. I remember the experience I had at AMI making watches. We were the firt ones to make the LCD watch. And the watch, which I still have, it's about one half inch thick. We sold it under the brand name of Gruen we sold it to a company called Gruen, whether they were really the original Gruen, I don't know. But the watch sold for some huge premium and quickly came down in price. But the semiconductor people really had no idea of the other important things, merchandising or case design. And those two things alone prevented them from really dominating their business. Turns out making the integrated circuit wasn't that important a factor in who was successful in the watch business.
RW: Even Intel failed at that.
BM: Well Intel bought a company called Microma and the two guys who started it were two Fairchild people, Don rogers and Bob Robson. And Noyce was very, very concerned as to what he was going to make in this new process, CMOS that Intel had come up with. So they convinced him that this watch was just a first of so many new products they were going to come out with that they would bury him in projects. They didn't have a second project. The watch was the only one and the quality of the watches, the failure mechanism of the watches wasn't the semiconductor, it was the LCD display. No one knew how to make good glass to metal seals so the display just leaked and sooner or later - after a year or so, it failed. And the same thing happened at American Microsystems even though we were slightly ahead of them, but the watch that didn't fail right away was the LED watch. That was pioneered by a lot of people because several people were making those modules, but certainly Intersil made them. The problem is the battery didn't last very long. They used too much power. And there was a fellow in the business at that time that I think was one of the m ore colorful people in the business his name was Gordon Ness. So we were in the business making these digital watches and ours were LCD. Well, the LED watches had one feature that was very, very prominent. You pressed the button, you could read it and it was very bright regardless of environment. The LCD watched worked fine outside in bright light, you could see them quite easily. However at night, if you had them on in bed or they were on your nightstand, they were almost impossible to read. So we were wrestling with what techniques we could use to do that. And we started experimenting with little tubes of tritium. So we were going to use tritium around the light so it's give a nice glow, slightly radioactive but the glass shielded such a small amount of radiation. But there were all kind of talks; your arm is going to fall off if you have a tritium watch, which was not true. We went to one of the big watch shows and Gordon Ness had a company called Ness Time. And Gordon was a promoter and a bit of a character. He shows up at the show with his new Ness light. And the Ness light, I'll tell you, was really bright. I had never seen anything like it. And Gordon would go around showing his watch off and the thing you could see it across the room the light, that is. You could light up the room. It was amazing. And he would not tell anybody how it worked. Well, the way it worked is Gordon strapped a series of batteries on his back and ran cords down his arm to this light which when people found out, that was the end of Ness Time and Ness Lights. But there are people like the characters who'd come into the business and they would get financing. And Gordon, of course, had some very tough financing experiences with the College of Marin where he had convinced the College of Marin to invest in Ness Time before they went belly-up. And the president of the college, which was a relatively poor college at the time, took their entire endowment and gave it to Ness because he thought he was going public in a couple weeks. Of course, they went out and when the college board found out, they fired the president. But they never got their money back.
RW: And they weren't the only people that lost their shirt in the watch business and the calculator business.
BM: Fairchild. National made money in the calculator business and made calculators for quite a while. So did TI and TI still makes calculators to this day. In fact, in the early days when I was at AMI, TI pioneered with a company called Bomar Industries in Fort Wayne, Indiana, the first single chip calculator. They had used a chip originally in Japan because I still have one here but it runs on the Japanese voltage at the time was ninety volts. So you'd have to use batteries to get it to work. It was both AC and battery. But it was the first single chip calculator and we were everybody was shooting to make single chip calculators, such as Fairchild. Fairchild never got, at that time, much beyond three chips. But Fairchild wasn't the pioneer in that National did quite well in the calculator business. And AMI did reasonable well in the calculator business. But the big winner in the calculator business was TI. American Microsystems, when I was there, developed the first HP calculator. We built the HP35 and the HP 45 and the HP80 and they were based on designs done by HP, not by American Micro Systems. All we did was implement the silicon. And what they did is they used the famous RPN, Reverse Polish Notation, which you have to hit enter key on every calculation but that was good for scientific applications. Engineers picked it up quite readily. It eliminated slide rules, I mean, there's nobody who uses a slide rule today. A calculator's used for everything there is. All the scientific calculators are now either built by HP or TI or people of that ilk and that all came out of the true power of a semiconductor. And as time went on, the ASIC business became quite large. One of the pioneers in ASIC business probably was American Microsystems. They were the first true ASIC manufacturer. But then people like Wilf Corrigan and team, including you Rob, came up with the technique of using tools to design the circuits. It was a lot more efficient and certainly a lot faster and lower cost. I mean, Fairchild had pioneered it but never did it on a successful economic basis because I remember using the old 360 computers that they had at Fairchild. We had tried that at AMI using I believe it was the Burroughts 8500 and it didn't work too well. We just didn't have good designs to use tools. So we were trying to do tools and also our accounting on the same computer. It did not work. We ended up using it just for accounting. But that's how the business changed the advent of tools. And then there were specialty tool companies like Cadence and Synopsis and people of that type who really made it an industry of its own. Those tools are now used widely throughout the industry plus all kinds of new ones that have come on since then. That is really why the semiconductor industry did so well here in Silicon Valley. The techniques all came out and they stayed locally and people learned from each other. And then they'd leave that company and start another one doing some specialized version of what they were doing before, because the people who developed the tools might've started at Fairchild but they would soon move to other companies and take that technology and improve it and make it into an industry of its own. And that's really why Silicon Valley became so powerful. But it was also an area that attracted very bright people. If you wanted a job, you didn't have to move to Texas, you're here and you could go to a start=up if you failed at whatever company you were and get another shot. You didn't have to relocate, in fact, you didn't even have to tell your family you were changing jobs. You just went a different direction in the morning.
RW: Well speaking of families, how did you come to meet y our wife, Marian?
BM: I met my wife Marian at her graduation from college. I was working and going to grad school in Cincinnati and my sister was graduating from college. Marian was a math major; my sister was a math major. In fact, both my sisters were math majors. But I was taking a vacation to Cape Cod. I called my mother and told her I was going to take my vacation in June, not coming home until Thanksgiving. And she said you're not going to be at your sister's graduation? She starts crying so I thought better of it. I decided to come for the graduation. I had nothing to do except go to graduation and she introduces me to her roommate we were married a year later. That's how we met.
RW: So how did she take the move to the West Coast here?
BM: Well, first of all, she was a school teacher she was a math teacher. In fact, she taught math one summer at the same school where Jerry Sanders had gone to high school. She was very flexible so we moved to Minnesota in the dead of winter and she had to give up her job in Chicago and move to Minneapolis. Then we moved constantly - we were always moving. I had moved from Cincinnati to Chicago to Minnesota, back to Chicago, back to California, back to Chicago and then finally back to California. So once we were here the second time, we did not move again. I was home; I found where I really wanted to live. We had four children, three boys and one girl. The boys all studied electrical engineering, influenced by their mother being a math major and my being an engineer. But none of them went into the industry. Our oldest son, Tom, was killed in an accident in Florida he was married and working there. His wife and children still live in Florida. Our second son, John and his family live about thirty minutes from here he worked as an analyst in the semiconductor industry for a while. He worked as a design engineer for Bob Schriener at Synertech.
RW: Bob Shriener?
BM: As a summer job. Then he was going to work for AMD but they were not hiring in the end and he ended up with VLSI. He worked for them for a year or two and then joined Alex Brown and Company out of Baltimore. He later joined Morgan Stanley eventually heading up investment banking for them on the West Coast. LSI was one of his big customers. He is now with Texas Pacific, a buyout firm they buy out companies and restructure them. Our third son, Bob, managed the AMD pension plan at a company called Duncan-Hurst. Now he's the head of a firm called Nicholas-Applegate in San Diego. But they are both in the finance business, nothing to do with semiconductors anymore, even though they both started out doing that. Bob worked for AMD in summers while he was in college one summer he worked for me. My daughter, of course was not very mathematical, but she is a special education teacher and lives in Los Altos. That's a thumbnail sketch of the family 4 kids, 10 grandkids, 3 daughters-in-law and 1 son-in-law. We will celebrate our 50 th wedding anniversary in 2010.
RW: Thank you very much, Bernie.
MM: I'm Marian I guess I'm going to say I'm the better half! This is the home we've lived in for thirty-eight years and raised four children. Our living room is where we host cocktails before all dinners for children, grandchildren and our friends. Our dining room table we had made especially for 14 thinking that would be all we would need, but our family now numbers 19, so the youngest children usually are seated at a small table right next to the large table. That way we are all in the same room for our family celebrations. We have a few family photos in the dining room that we would like to share. We have been fortunate to take some family trips - all 19 of us! In 2005 we all met in Lucca, Italy, and stayed in a villa for a week and celebrated our 45 th anniversary. The next big event will be our 50 th in 2010 so we are in the discussion stage as to what we will do to celebrate. Our kitchen is central to all events the family and our friends seem to gather here. But as is in most families and especially in Irish families, which is our heritage, we always seem to be in the kitchen. And this is our small family room which has served us well for all these years we just wish it were bigger. We recently remodeled this room and have a new plasma screen TV perfect for us! Our outdoor area is our main entertainment area for parties and the like. We have our pool, our good size patio, our bar and these amenities make our summer time gatherings very special. We love this house! It probably is a little big for us now, but when the kids visit, we need all the space we have and then some.
MM: The end and thank you!