SEMI Oral History Interview

David Lam

July 24, 2007 , San Jose , California

Interviewed by Craig Addison


David Lam founded Lam Research in 1980. Under his guidance as CEO, the company introduced the industry's first fully automated plasma etching system for semiconductor manufacturing. Lam Research went public in 1984 and has since become a global leader in semiconductor capital equipment. Since his departure from Lam Research, Lam has served as chairman of the David Lam Group, which invests and advises emerging technology companies in semiconductors, semiconductor equipment, computer hardware and software, biotech, networking, and communications.

Lam graduated from the University of Toronto with a bachelor's degree in engineering physics, and holds Master's and doctoral degrees in chemical engineering from MIT. Before founding Lam Research, he worked for Texas Instruments, Xerox, and Hewlett-Packard. Lam was born in China , and grew up in Vietnam and Hong Kong .




CA: David, could you start off by telling me something about your background in the semiconductor industry before you established Lam Research?


DL: I started in 1974 in Texas Instruments and that happened to be my first job right after I got out of MIT finishing my Ph.D. And I found it fascinating walking into a fab which was using primarily 2-inch diameter wafers and the excitement came from the very first 3-inch production line and I was spending half the time in the fab, the other half of the time working on new equipment which was later known to be a plasma etcher. And the work was extraordinarily interesting and I learned a lot. Later on I ended up at Hewlett-Packard in Palo Alto , two years later and, of course, by 1980 I started Lam Research focusing on plasma etching for semiconductor manufacturing.


CA: So what was it about the plasma etcher technology that sparked you to start a company focusing on that?


DL: Well, in 1970, in the 1970s, the industry has already discovered that plasma etching would allow the industry to move way below 5-microns in critical dimensions and the target was 1-micron which was a very big deal. There were a lot of publications, conferences and things like that around plasma etching but industry was reluctant to embrace plasma etching because it has been known to be not quite repeatable and not very consistent and so one of my challenge was to make it so repeatable that the production manager would be willing to take that technology to make chips.


CA: At that time, were there other companies with commercial plasma etching?


DL: Yes, there were a lot. There were many companies, mostly very small, trying to get into the field of plasma etching. Most of them designed equipment for R&D use. I was fortunate to be working on what I would call the “first generation plasma etcher” at TI and then later on at Hewlett-Packard I got the opportunity to work on essentially the second and third generation of plasma etcher. Both of them are for TI and HP's internal use. It was not to sell outside the company. The reason was that the etching technology was an emerging technology that would give the company an advantage against the competition so that's why it was a very important project to keep inside.


Now, in the case of HP…I should go back a little bit. At TI, I was working in the fab half the time and learning [what] the fab manager looks for when it comes to using a new technology or new equipment. At HP I'm learn something different and that is the power of automation in the piece of equipment to make it more consistent and predictable. So those are the two things that I thought of [that] benefit a great deal in terms of concept and in some ways in terms of practice before I started my own company in 1980.


CA: The decision to start your own company, was that something that you mulled over for many years or was it kind of a spark that you woke up one morning and said, “I want to do it?”


DL: Well, things always kind of germinate in your mind. It doesn't really happen overnight but the click happens overnight in the sense that I was one of the many scientists throughout the 1970s to publish papers, give talks on plasma etching and that was fun except that [it] is not being used in production. But towards the end of the 1970s, when I was still working at HP, it became clear that the industry was looking for a solution but there isn't any, perhaps not good enough. So I felt there was an opportunity for me to launch something that I believed the industry would want and that was the basic idea in starting Lam Research, but I was not the first plasma etcher company, not the last. The main focus that I had was to make the equipment so highly automated, in fact it was fully automated…the first time it was computer controlled and, of course, plus the chemistry too has to go with it, and that was the basic premise, to make it production worthy so that the fab managers would be willing to use it to make chips.


CA: Where did you get the money to start the company?


DL: In 1980, when I was starting, I was quite fortunate that two years before that, one or two years before that, the U.S. Congress has passed a law to drastically cut capital gains tax which really spurred the first wave of significant growth in the venture capital industry. So I was fortunate to see the money available to fund emerging technologies such as semiconductor equipment and I raised the first round in 1980.


CA: From venture capital companies?


DL: Yes, from two small venture capital firms and then a year later I raised additional ones from larger firms like Hambrecht and Quist and Mayfield.


CA: And didn't your mother put some money in, the very first money?


DL: Yes indeed, and it was kind of interesting. I was thinking of starting the company and I was visiting my mother who was living in Hong Kong at that time and I showed her what I was trying to do and she didn't understand anything except that she knew that I was trying to start some business and my father was an entrepreneur and he became a very successful businessman. So it didn't take too long for my mother to realize what I was up to, so she was very, very supportive and invested the first dollar, it's a small amount, from her own savings into the company and I came back to finish the business plan, refine it and before I hit the road and talk to the VCs. But that gave me a boost of confidence, as well as feeling very fortunate to have my mother's blessing on this.


CA: David, who else did you recruit to be the founding employees or any partners?


DL: Well, shortly after I got some commitment from the venture capital firm and the money has not come in yet…I was actively looking for good people and…I became quite aware that it takes a lot of effort, a lot of good people, different skills, different disciplines to accomplish and the first one I hired was someone from Applied, George Koch, and then later on I brought in other people -- Steve Muto from HP and later on Tom Voehl in sales from Tegal and, of course, added finance people and manufacturing and other salespeople as well.


CA: So what about your business plan; I believe you went to Bob Noyce and got his approval and that was a big help. Could you talk about that?


DL: Yes indeed. I was introduced to my first investor, David Arscott , who happened to know Bob Noyce and while he was looking at my plan, he said, “Well, perhaps we should talk to Bob and see if he might think that this plan is a good one.” So we sent in the plan to Bob and then later on we went in together to see Bob and he had already read the plan and his conclusion was that this was a good plan, it would work. This is, of course, from the chip maker's perspective, and the plan is all about making equipment. That was a great boost of confidence for me coming from Bob.


CA: Just going back to approaching VCs, you said that the tax break was a big help but did you get a lot of rejections from potential VCs?


DL: Well, I think anyone raising money should be prepared to get a lot of rejections and that's when I've learned to be thick skinned. Indeed, even though it was easier because of the timing, it was still very difficult mainly because, first of all, the industry was very small. The venture capital industry was small and the semiconductor industry, while exciting, the equipment industry was exciting and emerging. There was also a lot of risks so it's really always a balance there and it was never easy but looking back, I think I had good timing getting started in terms of funding.


CA: What about market acceptance for the product? How did that go?


DL: We introduced our first product a year after I started which, looking back, was pretty fast but it would still take quite a while before I get the first purchase order and luckily for me most of the fabs are in the Silicon Valley at that time. This is back in the early 1980s and they were relatively close but the geographic closeness is only one advantage. The product still has to work and work well and we were able to get into several fabs, only one at a time, or the major fabs in the Silicon Valley within the first year of introduction.


CA: You said the first product came out one year after starting the company. During that year were there a lot of struggles or did the development of the product go fairly smoothly?


DL: Well, there was quite a lot of struggle. The product development was relatively smooth but the decision to pick what to focus on was much more difficult. Plasma etching is a very broad field, even in those days. There are three different materials that we faced -- polysilicon, silicon oxide and aluminum. And all of them were important from the customer's standpoint. But what I recognized was that the equipment was so new it was going to be a big challenge to get it right and working well.


The process development is another challenge on top of the equipment so I had to look at the bigger plasma etch segment and look at the smaller pieces that I can handle in the beginning which was polysilicon. So there was a strategic choice that I made. Looking back, I think that was a good choice because that allowed me to launch a product with a working process which even though polysilicon was the simplest of the three, and perhaps not exactly the Holy Grail, but I have something working in the hands of the customer. So I was able to build some credibility in the industry and gain some trust from the customers who said, “This is working but only for polysilicon. If you are going to work on the oxide, I'll wait for you. I would buy from you if it is working too.” So that is one thing that I have learned in that process and how to zero down and focus on a segment that you can do very quickly as the starting point.


Now, there are other very strong companies. Some of them chose to focus on silicon oxide as a starting point which was what the industry was clamoring for. Aluminum was even more difficult and that was the real Holy Grail in those days, and most of them were taking so long, perhaps too long, to get the product out in the market and ended up not being able to survive.


CA: Without getting into too much technical detail, could you talk about the differences between polysilicon and aluminum and silicon oxide? What benefits the more complicated technology provided to the customer?


DL: Polysilicon was routinely etched in some of the less critical processes with bench-top systems. So for that reason it was not considered the most critical. Oxide had a most difficult challenge of so-called selectivity, meaning that when you finish etching oxide, if there is not enough selectivity, it would keep etching the underlying substrate. That is undesirable. Aluminum is even more complex because aluminum, when you take the wafer out of the chamber after etching, the water vapor in the environment, in the ambient, would tend to continue to attack the aluminum even without the plasma therefore making it extremely difficult to define the critical dimension. So those are the challenges.


When I chose polysilicon as the first process to start, people questioned my decision because it can be done, even though it's not great, but it can be done by other people and why do I chose that? The question was not asked, if it was so good, how come people are not using it widely in production? And they were not and for the same reason I said earlier, it was not repeatable, not reproducible, and fab managers could not count on it to produce chips out the door. So because of that, there is still a need for a production worthy polysilicon etcher and that was why I chose it. Now, again, it is only a matter of steps, a matter of priority and a matter of focus. All three of them were eventually launched by Lam Research and other companies


CA: David, what did you call your first product and can you talk about the pricing strategy?


DL: The first product was the AutoEtch 480. I chose AutoEtch because automation was the most important thing that the industry was looking for because it implies consistency, repeatability and, indeed, our first product was fully automated with a microprocessor which happened to be the Z80, an 8-bit processor. The 480 was a number I picked because it is for the 1980s and 1980 was just starting and is an exciting decade for our industry and so that's how the name came about -- AutoEtch 480.


CA: And the pricing?


DL: Well, pricing was very difficult and I had no experience in pricing commercial products and, in fact, it is an art. But one thing I have learned at that time that the pricing should never be based on the cost of making equipment alone. And value pricing was a new concept but I recognized I need to follow that thinking. I was looking around at all our competitor's product. Some of them are bench-top which was sold for about $35,000. Some of them standalone that were as high as $50,000 or $55,000. And I priced it at $160,000 base price, plus options which I know the customer would want to have. That adds up to roughly $200,000, and that was all based on the value that I feel that the product would give to the customer. And it was a little bit of a scary thing. If your price is too high, the customer may feel you are gouging them and resent. If you price it too low, you are leaving too much on the table. There is never really right or wrong but a lot of gut feel of what you think you can get away with.


CA: Can you talk about the competition at the time and in terms of competitors who kept you up at night and just what the competitive landscape was like?


DL: There were a lot of plasma etcher companies already when I started. In fact, the first question that I faced from the venture investors was that there were so many plasma etcher companies out there, why do you want to start another one? That was the first question, actually. And I had to explain the difference between just having one in R&D lab and one in production. There were very big differences and, again, that was what I learned when I worked for TI and later on at HP. But the interesting thing was that I was able to differentiate myself from the existing players, the existing suppliers. The biggest challengers were Applied Materials and Perkin-Elmer. Now Applied Materials uses a very different approach. They use a batch reactor that they license from Bell Labs, a very good system, very large and a lot of manual work to load and unload the wafers which, again, if someone committed to batch systems, they probably would pick Applied Materials. If someone believes single wafer would be better, I may have a better chance.


The challenge I had was that Perkin-Elmer also picked single wafer approach and they were the largest. They were certainly among the largest, if not the largest, semiconductor equipment company at that time. The lithography systems [unit] was the leader and they were now getting into etching and with some very good people that they hired to start the new division. Now, I went after automation, full automation. So did they. I used Z80. They used a Perkin-Elmer computer which was worth many, many more times a Z80 chip. I used a simple 9-inch display. They used a very sophisticated display to show the motion of the wafer movement and the motion of the robot that sends the wafer in and out. I used very simple ones with line drawings and numbers. So they were having a very sophisticated display along with single wafer etching and knowing how big the company is and how resourceful it is, I was very concerned about the competition coming from Perkin-Elmer.


Now what I learned later on was that they did not pay attention to segmenting the market as I did and other people did or the processes at the same…time which will make their product very difficult to have the first success story. The second thing was that they did not perhaps pay enough attention to cost and everything they do was what the market wanted but by the time you add them all up, it was hugely expensive so those two things combined make them less competitive and as a result, as you know very well, they were unable to survive in the plasma etch market.


CA: Did you have a defined strategy to compete with Perkin-Elmer or you just kind of toughed it out and survived?


DL: Well, initially I was very concerned and then I went to a tradeshow, SEMICON West, and I walked around and stopped by their booth. I know some of the people there. They were showing me how great their product was and they are proud to point out that they use a Perkin-Elmer computer to run the whole thing. Now the Perkin-Elmer [computer] at that time runs somewhere between $35,000 to $37,000 which is very inexpensive by today's standard but very high in those days. That was the price of a whole etcher if it was a bench-top model. I use a chip which costs only $500 doing the same thing so immediately I know that I would have a significant advantage over their approach in cost so I was able to price at the point where I can be very profitable but still way below what they would have to ask for.


Now in addition to that, I also looked at the complexity of their system. The way they designed it really would make it very difficult to have high reliability and high up time and there was sort of a superficial review by me but I have used a lot of equipment. Now I was building one. So as a result, I see that the true several weaknesses in their approach [that] would make it possible for me to compete against a big company.


CA: Did you come up with any customer concerns that they were a big company and you were a small company and maybe we [the customer] shouldn't risk buying from a small company?


DL: Yes. There is always that. All companies would be concerned about buying from a small company that may or may not be around two years from now, that kind of thing. So I do have a lot of discussions with early customers when they bought the first one and when they bought the first one, they actually had in mind that if this works, we'll be buying a lot, so let's figure out if the company would be around. If we buy from them, can they really continue to support us?


Now, I have to point out that in the 1980s the requirement for support is much more relaxed than it is today. It doesn't give 24 hour response. It doesn't need 24/7 and that kind of thing, but still, it is a very important issue. So what I did was, when I shipped the first systems, I would tell the customer, “Here is my home phone. If you have any problem, you call me directly. You can call me at home and I would come to take care of it.” Now, in those days there's no cell phone. You have to give them the home phone in order to provide the access beyond office hours. But that also makes a statement of a commitment to support them. Now they don't always call me, but [sometimes] they did. So when they do call me, the next morning I get to the office, I round up a team and I get on a plane and that was what happened to Intel when they bought our first system in Oregon and there was some problem and we stayed there for three days and I, along with two other engineers, we fixed problem before we left, and as a result, we got a lot of good feedback from the customer and they became very loyal to Lam Research for years.


CA: So who was your very first customer?


DL: Actually, my very first customer was Trilogy. That was a company started around the same time as I did by Gene Amdahl of the Amdahl Computer fame. They were able to raise a huge amount of money trying to build a computer on the whole silicon wafer. It turns out that it was just too challenging so the company was unable to get the product done. There are other customers immediately following that including Intel, AMD and National.


CA: David, could you talk about subsequent product developments? I mean, did you eventually develop an oxide etcher, an aluminum one? What came after the polyetcher?


DL: After the poly etcher we targeted oxide and aluminum wasn't too far behind, but I think that it appears oxide was in greater need at that time. I think it was not as easy as most people thought. You can etch even with good selectivity but the rate tend to be very, very slow and when it is getting too slow and you do one wafer at a time, it is just not good enough for production and, in fact, doing one wafer at a time was a concept that was quite new in the early 1980s. As you recall, the whole industry was using only 3-inch wafers and 4-inch wafers and doing multiple wafers in each load was the norm…was the thing that our customers were used to. So the reason why I proposed doing one wafer at a time was because plasma chemistry is such that it is extremely sensitive to minute variations in the ambient around the wafer when you do etching. So any little change, any little leak, any little leftover water vapor on the wafer and things like that would change the result so that is one of the reasons why it is so difficult to make it consistent and if it was easy, everybody would be using it at that time.


The single wafer concept in part was driven by the need to be consistent so in a very small environment, where you only have one wafer and you isolate the chamber from the ambient using load locks and other mechanism…then you can produce a very consistent, isolated environment for etching and, as a result, you get better repeatability. Now, that is fine except that it is also vulnerable to attack by your competitor who may be doing multiple wafers at a time. So I had to come up with a way to counteract that by pointing out that a single wafer, the area being processed, goes up as a square of the diameter as the wafer gets bigger, so that's a very significant movement from 4-inch to six to eight [inch].


Batch wafers, on the other hand, actually would drop slightly as the wafer gets bigger because more area inside the reactor would be wasted so not to count the overhead time, the handling by people, therefore creating dust and other contaminants. Apart from that, the throughput is actually comparable when the wafer size gets to about 5-inch. So I was able to publish the paper based on that analysis and be able to blow away some of the concerns about single wafer being so slow that it would never make in production. And that was a very critical. I would call that a marketing battle, if you will, to overcome, to win. Because once the industry believes that single wafer would also make it in production, then they would not mind buying it now. And to drive home the point that the wafer size would be in our advantage if you etched one wafer at a time, I made sure all the equipment that we built was 6-inch compatible. Now, when we launched the product, there was no such thing as 6-inch wafers. It was to come a few years later but there was the talk in the industry and that was how the users would envision being able to use the same equipment. When 6-inch comes along, they will have to throw it away. In reality, they never really used the same equipment when the big wafer comes. They always buy new ones but that is the idea behind it. So I had to do a few things -- combat the idea, convince the industry, and design the equipment that would be consistent with that message.


CA: So Applied was the biggest supplier of batch systems?


DL: Yes, Applied was the biggest supplier of the batch system and in the end it became a two horse race for quite a while until TEL severed the relationship with Lam and then became the third one.


CA: So how did that play out, the batch vs. single wafer? Which became dominant?


DL: Well, Applied for years…after TEL became an independent supplier, starting about 1989, then it became a three horse race. Each of these three companies -- Applied, TEL and Lam -- had about one-third of the market share roughly and, of course, the few percent left for everybody else and that has been like that for years. I don't have the recent numbers but that is very incumbent in most sectors. In the end you've got two or three. Often times there were only two.


CA: Did TEL have a single wafer system?


DL: Yes. In fact, TEL was allowed to use our design. When the two companies decided to stop collaborating, they were allowed to use our design to sell into their market.


CA: Perhaps now we can talk about Japan and your partnership with TEL [Tokyo Electron Ltd]. What was the beginning of that? How did that come about?


DL: Well, in 1981, when I was launching my first product, it was clear that in the whole world there were only two geographic markets -- the U.S. and Japan . There was very little in Europe . There was nothing in Taiwan practically and nothing else in the rest of Asia . So immediately I realized that the market was roughly 50/50 between Japan and the U.S. If I were only focused on the U.S. , at best I would get 50 percent of the market. That was not good enough. So I tried to figure out a way to get into Japan . Now, I also have visited Japan, I learned a little bit about the country, and I realized that given their culture, it is very difficult for a small, unknown company from California to try to sell to the big chip companies in Japan so it became very, very clear from the very beginning that I needed to team up with a large company in Japan to market our product. Now, unfortunately TEL has already got a product that they were representing so it took me quite a while to really work through the process to convince them that we have a product that they should consider and they should make a change of the supplier they represent.


You may recall that in those days TEL primarily represented American-based equipment companies. That has been their business model when it comes to equipment. They were working with Varian, they were working with KLA at that time, and other companies as well. So I was one which was trying to get in the door to team up with them so it took me a little while, but one of the things that was in our favor was that after the first few conversations, TEL decided to send a team of engineers to come visit Lam Research in Santa Clara. The first team that came in was not salespeople, was not process engineers. They were equipment maintenance engineers because they were very concerned about reliability in the field. They got stuck in supporting the customer once we have sold the equipment. So that was a very important point on their checklist so we passed with flying colors because that was my driving point in launching the product because all the automation and other things really made it more reliable and repeatable. So we passed that and, of course, the performance was OK as well. So that's how we got to convince them to drop the other vendor to go with us. And it took me a year before Mr. Inoue , Akira Inoue had a handshake with me to represent Lam and this was done in 1982. So by the end of '82 I was showing our product at the TEL booth and from then on we were able to compete against products, etchers which were made in Japan .


CA: Was there any concern of having people from Japan coming and opening up your equipment? Any IP concerns?


DL: In fact there was. In fact, some of our engineers were very reluctant and they counseled me not to do that but I had to overrule that argument mainly because, first of all, in our situation it is not that easy to copy anything just by looking from the outside and see how it is accessible to make some changes and to repair, to maintain the equipment. Even if there was any intent to do so, [but] I didn't believe there was any intent. The second thing is that in business there is always going to be some calculated risks you may have to take from time-to-time. This was one [where] I clearly felt there was risk but the risk is acceptable so I decided to take that calculated risk and it turns out to work very well for us, because without that review by their maintenance engineers, we would not be able to move on to the next step.


CA: Now David, could you talk about how the distribution agreement with TEL became a joint-venture, the events that led to that?


DL: In 1982, as I mentioned earlier, we signed the marketing agreement and TEL represented us. I went to Japan many, many times to go with the salespeople to visit customers and it went quite well. But a couple of years later we realized that perhaps at some point in time we would want to do some more locally in Japan to serve the customers so that was the beginning of the idea of a joint-venture. The product that we shipped to Japan was over time going to be made in Japan with whatever modifications that we see fit to make the Japanese customers happy…and that was TEL-Lam. The joint-venture was formed, I think about '83 or '84.


CA: And that was 50/50?


DL: It was a 50/50 joint-venture and the market, of course, is primarily Japan , not outside of Japan . Any shipment to outside of Japan , which happened later on in the decade, were all shipped from the U.S.


CA: Ultimately, was this strategy successful do you think, the joint-venture and the partnership?


DL: I thought it was. I think that without that we would not have been able to penetrate Japan as quickly as we did. Not only that, penetration is one thing. Being a leader in Japan is another thing. I think TEL has the marketing muscle, has the sales organization and, of course, the support organization too to combat not only the big company's etchers from the U.S., which later on became Applied Materials, but also the indigenous companies. Hitachi was making etchers. I think Toshiba was making some and several companies were making plasma etchers with similar design ideas so by getting in to work with TEL sooner we were able to dominate the market by the mid and late-‘80s.


CA: So the biggest competition for you was Applied in Japan ?


DL: Yes, that's correct.


CA: Could you just talk a little bit about [the late] Akira Inoue. He's a very respected figure. What were your observations about him?


DL: Yes. I got introduced to him by the head person in the United States . They have TEL America's office and [it was] run by Sam Kano. Kano-san was the one that I contacted first to convince him to come by Santa Clara…his office was in Sunnyvale at the time…to convince him to come over and take a look at us, even though they already have a vendor and so his first look at the equipment was a big surprise in terms of the design concept, in terms of compactness, in terms of the thought that went into repair and maintenance, as well as the automation, the level of automation. So all of those combined intrigued Kano-san who later on introduced me to Mr. Inoue.


Now Inoue-san had negotiated with me on and off during that year and again only after we signed the agreement in mid-1982 did I start to work with him more closely. As you recall, he ran the semiconductor equipment, the capital equipment division which was the largest of all four divisions within TEL. TEL sells chips too [and] other things -- computers, certain kinds of computers. So because of that, that division turns out to be the fastest growing, the most important and, of course, later on Inoue-san became the head of TEL. I had tremendous respect for him and great pleasure working with him throughout the years when we launched the product in Japan .


CA: Just going back to the product strategy, did you ever consider or actually diversify beyond etch?


DL: Well, as a small company, even after we had gone public in 1984, we were still very small and I had always felt that there was so much more to grow in the plasma etch market. We were barely into polysilicon around the time when we went public. We had not introduced in any significant way the oxide etcher. We were working on that and initially [it] didn't do very well. So as a result, I was under the impression and I also felt that the market for etching was so large that I would want to be number one in that first before we would work on other markets such as CVD, for example, which is somewhat close but not the same.


CA: So you didn't try to introduce other products?


DL: I didn't try that but I think Lam Research later in the decade did try to move into epitaxial reactors by acquiring a company and also towards the end of the decade, or maybe early 1990s, Lam Research also went into CVD by acquiring a small outfit in Southern California . Unfortunately, neither one of those worked out very well.


CA: During your tenure, did you consider any acquisitions?


DL: I was on the board at that time when some of these acquisitions took place but I was not the CEO then.


CA: You just briefly talked about the IPO before. Any highlights about the IPO that you can recall?


DL: The IPO was very interesting in the sense that in 1983 we did what we call the “mezzanine financing.” Mezzanine financing is not as popular today but it was very popular then. It is a financing between a venture investing and public offering and usually you raise money from non-venture sources but still as a private company. So the idea was to raise enough money to give you additional resources to grow big enough to become a public company, so anywhere between 6 to 24 months is good timing to prepare for that and you clean up your balance sheets, you get a little bit more money to grow and things like that.


So we did that in the summer of 1983 and when we came back, we thought that we were going to be sitting tight for a little while before we consider IPO but we have already picked…Goldman Sachs as our banker [for] when we would be ready. So in early 1984, one of the senior partners from Goldman Sachs came to visit us along with one of the economists and essentially the message was that in their analysis…they have economic models, not only the models of the industry but also models of the economy and the high tech market as well. According to their analysis the IPO window for technology companies may be closed or difficult by the end of 1984 and [they had a] whole lot of analysis, a lot of the reasons why they arrived at that conclusion. So they urge us to consider going public sooner rather than later.


Now, we just came back a few months ago. We had no intention, no urge, no compelling reason to go public right away but after listening to them, we decided to take their advice to move, to go public. So by February of '84 we got ready and it takes about three months to prepare and all the filing of papers, the writing of prospectus and preparing for the IPO, so by the first week of May we went on the road show and raised another $20 million. So we had $30 million in the bank when we came back. It turns out they were right. The IPO window was indeed closed towards the end of the year and suddenly the recession hit the United States again starting about 1985 for a few years and we would have had a difficult time going public if we didn't do it right away.


CA: What was that money used for specifically?


DL: Well, at that time we were really experiencing very fast growth particularly…as you recall, 1980, '81 and '82, those were the three years the U.S. experienced a recession. This was a very difficult recession combined with inflation and those who worked in those years would recall that and selling capital equipment in the recession time is extraordinarily difficult. Not only that the general economy was slow but the capital budget tended to be shut out and you don't have a whole lot of budget to buy big pieces of equipment so that was what I faced when I first introduced the etcher and as a result most companies would only buy one in part because it was new equipment. They need to evaluate anyway but also because they don't have budget to buy more.


However, by the end of 1982, the recession was beginning to end and our product, Lam Research etchers, had been tested and evaluated for a period of time, more than a year in some cases and our customers seemed to come to the conclusion that this is a good piece of equipment. If we need it, we want to buy more. And when money is available, the willingness to buy it was there. We started to receive very big orders. I never [before] saw orders coming in for five systems, 10 systems which was something extraordinarily exciting because before that I was only able to sell one or two to each company.


But then there also was a problem and that is working capital. We didn't have enough working capital to really ramp up sales and production so the money we raised in the mezzanine financing as well as the IPO in 1984 was extremely important for us to finance the growth that we experienced.


CA: David, could you just talk about your departure from the company…why you decided to move out of the position of CEO and what your thinking was there?


DL: Well, around '83, Roger Emerick came to join the company as the new CEO and I was very pleased that he was coming in to really work with me in part because I was not very good in sales. I was lucky to sell a number of systems with the help from Tom Voehl. Tom Voehl came to join me as a national sales manager from Tegal and the two of us can sell a few. He was very good but the company was too young, plus I didn't have the experience of building a sales organization and Roger Emerick was the VP of sales from Cobilt and after Cobilt was acquired by Applied and he came to join us shortly after that. So that really made up one of my biggest weakness…with Roger Emerick and our CFO and other engineering teams…and both Roger and I handled marketing. Even though I had a title, I was primarily responsible for marketing and customer service, so together we have the beginning of a complete team and that was what made the IPO very attractive the following year.


Now, towards the end of the decade, I was phasing out of operations but I stayed on the board for several years and later on essentially I was convinced by my family to leave to run a family fund and the David Lam Group is an investment vehicle. But I also use it to help and work with other small companies to advise them, to guide them. All of them are technology-based companies. Most of them are in semiconductor as well.


CA: Looking back on your career and starting the company and so forth, any disappointments or anything you would have done differently?


DL: There were disappointments which very much come as a natural growth of a start-up. Let me give you a couple of examples. They are related. After we got the first few purchase orders from different companies we were, of course, extremely excited because for the first time there are people out there who said that what we have worked hard to build in the last two years were worth something they were willing to buy, pay for it. That was, I considered, a very important milestone for the company. But at the same time, the sales did not pick up as I thought it might. Customers would tell us they loved our equipment, they loved our company and all that, but there's no more purchase orders. So, as I indicated earlier, my sales experience was one factor. The recession and the lack of capital budget among our users was another factor, and there was also a factor which was that all new products would have to go through this phase of evaluation by the customers and that is a very difficult period because you have all the people in the company but the sales were not moving linearly. So that was my first experience of this phenomena among technology start-ups and therefore I was very, very disappointed and I also had to let go a few people. We only still have a small team but we had four people too many. So all of those combined makes a very painful experience but also an experience that I have learned and have taken with me to coach other start-ups because that is a process that all small companies would have to go through.


Now what made us jump through that? Again, the recession ended and Roger Emerick was very, very good in promoting sales, not only in the U.S. but also in other parts of the world and we worked well together in marketing and other things and he brought in another guy from another company to handle manufacturing which made us much stronger in shipping very reliable products and the financing was good because we got the IPO done by 1984. So all of those combined allow us to move up into leadership position within that market segment.


CA: So overall, the wins outweighed the losses.


DL: Absolutely, and I think despite the disappointment at that time, looking back, that is almost a natural process for start-ups. But one of the things that we also did, and I took care of this more than Roger, and that was to work with the press, work with the trade press, work with the business press and make myself available to interview and give talks at SEMICON, not only at SEMICON West, SEMICON Southwest, SEMICON East, SEMICON Japan, and everywhere that I get a chance to promote the message. Not a sales pitch, because SEMICON does not allow that but within the context of allowed presentations I was able to present the new idea that was embodied into our plasma etcher and to promote the advances in our industry. So working with the press and giving talks and interviews allow us to create a little bit of buzz in the marketplace which, combined with a good product, we were able to grow a little bit faster getting into the mainstream market so that was all happening in about three years or so from about '83 to 86.


CA: You said that you've got the David Lam Group. Could you just talk briefly about what you are doing these days?


DL: After I started the David Lam Group, initially [it] was primarily investment which I still do but I found that interestingly money was not the problem for most start-ups. Advice is. And many of them were able to convince investors to put in money but then they made so many mistakes repeatedly, similar to the mistakes I made when I was running Lam Research and some of them just couldn't survive. So when I got into some of the small companies, I used to serve on the board, whether or not I am one of the investors. Then I started to work with the founder, work with the team, of course work closely with the board of directors, and that is where I contribute the most to start-ups that I am involved with in the last 12 years or so. So that has been something that I enjoy doing and I enjoy it so much I'm working on a book to summarize those ideas in written forms that perhaps a larger number of entrepreneurs can benefit from.


CA: Thank you very much, David, for your time.


DL: Thank you.