SEMI Oral History Interview

Morris Chang

August 24, 2007 , Taipei , Taiwan

 

Morris Chang has been the chairman of TSMC since the company was established in 1987 and served as CEO from its founding until June 2005. From 1985 to 1994, he was president and then chairman of ITRI, the Industrial Technology Research Institute in Taiwan . Earlier in his career, Chang served as president and chief operating officer of General Instrument, and prior to that he was corporate group and senior vice president for Texas Instruments. Chang holds a bachelor's degree and a master's degree in mechanical engineering from the Massachusetts Institute of Technology and a Ph.D. in electrical engineering from Stanford University , and has been active in the semiconductor industry for 51 years.

 

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>  

 

Morris Chang: I was born in 1931 into a middle-class family in China . And I lived in China until I was 18 years old, when I moved to the United States . So the early part of my life between birth and 18, the background was war, poverty, injustice. So those three things -- war, poverty and injustice, they dominated the backdrop, the environment in which I grew up. War, first the Sino-Japanese War, then of course the Second World War and then the [Chinese] Civil War. And I lived through all of them. China was so poor, most of the people were so poor, even middle class for me that I was born in, we didn't starve, we were not hungry ever but our life style was very, very modest compared to a middleclass family lifestyle now. Injustice. Even back in those chaotic days of my youth I could see and also read about people that were very rich and very powerful and indeed, I even saw, I even met some of them as a young person. But those were just in the very small minority and most of the people that I saw were also middleclass. They also lived very, very modestly and then there was a large number of people that simply didn't know where their next meal would come from. And people that didn't have adequate housing, I saw a lot of those people too. And so how did that background influence me? Well, first of all I knew that war was just a terrible thing. I knew that abject poverty was also a terrible thing and injustice, social injustice. I saw something that lots of people ground their teeth at…myself included. So much resentment, so much hatred in a socially unjust environment. Now the personal lesson to me was, well, I needed to study hard and later on, work hard. And hope that there will be peace. That was the personal lesson to me.

 

MC: Well the United States was paradise when I first arrived at the age of 18. Politically, economically and socially. And yes, certainly I shook off all those shadows in my mind. I quickly shook them off anyway after I arrived in the United States . But you talk about war, you talk about going, having to move from one place to another, yes…our family had to flee from the air raids in Guangzhou . We were living in Guangzhou when the Sino-Japanese War started and immediately there started to be air raids. So my mother and I moved to Hong Kong and really [had] a few years of peace in Hong Kong . Hong Kong was beautiful and certainly it was not invaded or anything until the Pearl Harbor day. Everybody knows Pearl Harbor because of what happened at Pearl Harbor but Hong Kong was attacked the same day by the Japanese and we lived through gunfire for two weeks and then we lived through Japanese occupation of Hong Kong for another year. Then we went to Chungking and we again had very serious air raids in Chungking . And then finally “VJ Day” came and we moved back to Shanghai hoping for a period of peace and development. But then civil war which of course was happening all the time anyway, except that Shanghai was not touched by the civil war until about 1948 when the Communist armies actually came pretty close and we had to flee from that again. So we did move from place to place every few years.

Well, I went to the U.S., I went to Harvard, entered the freshman class of Harvard in 1949 and after a year I knew that in order to get a good job after I graduate I should go to a school with more specialized training, so I transferred to MIT (Massachusetts Institute of Technology). After I got my Bachelors and Masters degree at MIT I wanted to go on for my Ph.D. But I failed the qualifying exam so I sought a job in industry. I had four job offers and just by a very lucky break I chose without really knowing the implications, the ramifications, Sylvania Semiconductor. And I worked for Sylvania Semiconductor for three years and I decided, and I was right as later events proved, that Sylvania wasn't going anywhere in semiconductors. And I knew already that Texas Instruments was rising very fast in Texas . And so I joined Texas Instruments after three years at Sylvania . I was first just an engineering supervisor at Texas Instruments. By that, I mean, I had two or three engineers reporting to me. And I was in charge of a production line. I was on the fast track at TI. I rose very fast. By 1961, which was three years after I joined Texas Instruments, I had already become a manager of a sizable engineering section. It was at that time that my boss told me that TI wanted to develop me by offering me an opportunity to go for my Ph.D. The opportunity was golden because TI would pay full salary while I was a student and they would pay tuition [and] other school expenses. And I could even go to two or three technical conferences in other cities in the United States while I was a Ph.D. student. And that was an offer that is hard to refuse, so I took up the offer and went to Stanford for two-and-a-half years and got my Ph.D. This time I passed my qualifying exam with flying colors so I had learned to study hard and study efficiently too. And so after two-and-a-half years, that took us to 1964, Spring of 1964, I finished my Ph.D. thesis and I went back to Texas Instruments to work fully expecting that I would still stay in the engineering area.

But a few months after I got back to TI, the vice president of the semiconductor business asked me to head up a very important product department in Texas Instruments and of course that was what I had always wanted to do, general management. And I then became the manager of more than 2,000 people. Remember back then production was all onshore, was all in the United States , in Texas . So the 2,000 people of course included a large number of production people, but still there were maybe about 200 professional people. So it was a big promotion for me from just an engineering manager to a general manager. And after that I was still on the fast track and I [was] promoted from less important departments to more important departments and in 1967, early ‘67, I became the general manager of the most important semiconductor department in TI which was integrated circuits. And then a few years later I again was promoted to be the group vice president of all TI's semiconductor business and I stayed in that job for six years until TI's consumer business…it was new to TI and it was not in good shape, and they transferred me over to head up the consumer group and I was in that job for another two-and-a-half years. And the consumer business did not improve. I think that was due to many factors…then I was transferred to a staff job. I was still a senior vice president but effectively I felt that I had been put out to the pasture. Still, I agonized for at least two years as to whether to leave or to stay at Texas Instruments. But finally after more than two years of agonizing, I decided to leave.

MC: Many highlights during my 25 years, a few low lights at the tail end also but many highlights at least in the first 20 years. How did I get on the fast track and get promoted successfully pretty fast? Well I think that I did something in each of my jobs. I accomplished something significant in each of my jobs. In fact, my first big accomplishment happened three months after I joined TI and I wrote about it in my book…my autobiography. In the production line that I was in charge of, I increased the yield. The yield was practically zero in the first three months I was there. And then one day…I kept changing the recipe. Back then the process you know, essentially was a recipe, I kept changing the recipe but I changed the recipe very rationally because at Sylvania I had by self study and by experience became pretty good in transistor technology. And after various numerous changes of the process recipe, suddenly one day three months after I joined TI, the yield was up to 25-30% from zero. Suddenly we were in business, whereas before, we were just cranking out rejects. And it was a very important production line for Texas Instruments at that time. And I remember that night, I mean I couldn't sleep I was so excited. I knew that I had accomplished something big because I knew the impact of the high yield. And the profit that resulted from that was significant to Texas Instruments.

Back then, this was 1958, the industry was still in its infancy and the equipment that we used was primitive. So it was basically a matter of just having the right temperature, having the right pressure at the various process points and in order to even guess at what the right temperature, what the right pressure and what the right dopant and all those things…in order to even guess at what the right things you should do, you certainly needed a pretty good grounding of device physics. And that was the thing, device physics was what I became pretty good at, at Sylvania .

It was alloyed transistor and diffused transistor at that time. My line was a diffused transistor so you have to calculate how fast something diffuses into… it was germanium also by the way… how fast something diffuses into germanium and what the base width is going to be. The base width cannot be too wide. If it's too wide then you don't get a very strong transistor action. And you don't get very high frequency either. But if the base width is very short or very narrow then you run into shorts. So I knew device physics pretty well by then and so I tried various recipes… and temperature of course was not the only variable. There were other variables as well, such as the impurity of the dopant etc. etc. And the germanium wafer thickness, I mean, heck all these things, they were ancient by today's standards. But back in those days those skills of how to achieve good yields in transistors those skills were very valuable skills.

And then I developed… when I became a bigger manager, still engineering manager… I developed a number of new germanium transistors. In fact one of them, another one now, turned out to be extremely profitable to Texas Instruments again and that was already after I came back from Stanford, from my Ph.D. And so I was promoted to the general manager job that I spoke of earlier, and in that general manager job I finished the development of this second important germanium transistor that I had started to develop. I finished the development and put it into production. I also had to negotiate a deal with then the most important customer of Texas Instruments, IBM, and then we put this newly developed transistor into production just to sell to IBM and that turned out to be very profitable for Texas Instruments again. So I quickly got promoted to become the general manager of a more important department, which was silicon transistors. Now in my year as the general manager of silicon transistors, I think I did maybe one important thing. I set up a factory in Canada because there was a Canadian market. The new factory in Canada would allow us to exploit it. That was probably the first time that TI set up something like that. And then I went to integrated circuits, became the general manager of integrated circuits and there we started a learning curve, a methodical use of the learning curve… as a control, as an influence on pricing and as it influences our cost reduction efforts. And this was a pioneer effort in the semiconductor industry. We did it together with the Boston Consulting Group, in fact, they talk ed about it quite a bit. This was in the late 60s. And Boston Consulting Group was a very small outfit when we did this, and I was the key contact guy that worked with them…they were of course our consultant. And we used what we worked out with them…loads of data, a lot of theory and a lot of effort put into the practice of the theory and use of the data we had already taken. The result was so called learning curve pricing. And we would automatically reduce the price every quarter. Even when the market did not demand lower price. And this was a very successful effort even though it was somewhat controversial. A lot of people thought that we were being foolish. Why do you reduce price when you didn't have to? But we did that because I believed in it and indeed our market share just kept expanding. That wasn't our only strategy… price reduction, learning curve pricing was not our only strategy. We had other strategies too which were just as important, such as developing new products, developing medium scale integration at that time. And any way, all those strategies combined to make our integrated circuits business, the TI integrated circuits business, the biggest integrated circuit business in the world and the most profitable also.

And that was when I became promoted to the semiconductor business job. Just during that time when I went from integrated circuits to the semiconductor job, the IC world was changing. It was changing from bipolar to MOS. And TI was nowhere in the MOS game. I, as the integrated circuits general manager was not responsible for MOS development at all. I was just in charge of the business, of the integrated circuits business, which was almost 100% bipolar. The development of MOS was in another group, but when I took over the TI semiconductor business entirely of course, MOS became increasingly more important. It pretty quickly took over the bipolar integrated circuit world. So I would say the highlight in the time, at least the early highlight during the days of my semiconductor general manager job was to catch up and in fact even surpass our competitors in MOS technology and MOS sales. And we did that mainly through calculator chips and through the DRAMs, the memories. And not just DRAMs, but also PROMs and EPROMs. And those kept me busy for quite a few years. The highlight in my consumer days I have to say was probably one product, the educational product, Speak & Spell . That was developed and marketed during the time I was in charge of consumer. It happened to be I think the best product that the TI consumer division ever did.

MC: I left TI without any job offer, and I left because I felt that essentially I had been put out in the pasture at TI. I felt sure that I would still continue to have a job, in fact even a job with a good title, but my hope of further advancement at TI was gone. I decided that there was no chance that I would advance further. Actually there was only one step further, it was the CEO, you know, and I decided that there was no hope of that, and besides I wasn't doing anything that interested me very much any more. So that was why I left TI. But immediately after I left, and my resignation was published in the Dallas papers, in the electronics trade papers, and in the Wall Street Journal also. Immediately after I resigned I started to get a lot of phone calls, and I pursued three or four of those opportunities, and one of them happened to be General Instrument. It was to be the president and the COO of General Instrument in New York . And at that time General Instrument was a sizeable company, not as big as TI, but it was a sizeable company. It was very definitely in the Fortune 500. I decided that this would be an opportunity for me to manage a sizeable company and accepted that offer. Unfortunately after a year or so at General Instrument I found out it was so different a company from Texas Instruments. Texas Instruments focused on organic growth. Even the branching out to consumer business and to computer business was organic. They just took good people from other businesses, principally semiconductor business, to staff and manage the new business, so even that was organic. Whereas, General Instrument was mainly about acquisitions, mergers, and then after a business became successful, sale of that business, so that General Instrument could acquire new businesses, and remake them, or improve them and sell them again, so.

But when I was being interviewed I did not know that it had become that kind of a company. I mean there's nothing wrong with it being that kind of company, but it was just very different from what I was used at Texas Instruments. So after a year of General Instrument, a little more than a year, I decided to leave again. Again, without another job prospect, but then Taiwan beckoned, and the offer was to be the president of the Industrial Technology Research Institute, ITRI, in Taiwan . And at that time ITRI was a very important institution in Taiwan . It was the largest and the only significant industrial technology research in Taiwan . The Premier [of Taiwan] who appoints the president of ITRI told me that he particularly wanted to use my ability to transfer technology from just research results to economic benefits for Taiwan industry.

And it was that part of the Premier's appeal that he wanted me to use my abilities to transfer research results into economic gains for Taiwan industry that really appealed to me because I had not been a purely research or technical person for a long time when I took the job in ITRI. But I had always, at Texas Instruments and later for a short while in General Instrument, I always thought of my job as one of transferring R&D results to economic gains for the company. So I felt as if I could do this ITRI job also.

Rather than being a corporate executive, the ITRI job, the presidency of ITRI would mean serving the whole industry, serving the whole Taiwan industry. Using the research results of the institute of ITRI to serve the whole Taiwan industry, and try to make money for the whole Taiwan industry with the results of the research in ITRI. This was a completely new kind of a job for me, and also the environment would be completely different. A Taiwanese environment as opposed to the U.S. environment. The kind of people that I would work with would also be very different. So it was the newness that appealed to me, the difference from what I had been doing for several decades first in Texas Instruments then in General Instrument. And, I mean, everything seemed to me so different. And by then I was financially pretty secure. I was not rich, but you also have to realize that the standards of wealth were much lower back in 1985. But still in absolute standards I was financially secure which meant that I could live according to the way I desire, which was actually pretty modest… for the rest of my life without having to earn a living or a salary. So I came to Taiwan at a much reduced level of compensation, but because it was so new, and appeared to be so challenging… that I came. And this was why I said that you need to follow your interests, not where you think the big money is. Because, obviously, back in 1985 the big money was thought to be in financial venture capital, or maybe even continuing to manage a company in the U.S. , but I felt that the Taiwan opportunity appealed to me from the interest point of view.

But as it turned out, when you don't chase money, money comes to you. Now, not in the ITRI job, the ITRI job itself is just a salaried job. It was at that time, it is now. But I got a chance of starting up TSMC after I came to Taiwan , and TSMC, obviously, was an opportunity to make some money personally.

MC: Well, the political leaders deserved all the credit for getting the RCA license, but that didn't mean that Taiwan entered the semiconductor industry just by licensing the technology from RCA. All it meant, this license of RCA…was that Taiwan got the then RCA current technology which was probably a generation behind the leading technology in the U.S. then. Now, after Taiwan got it…a group of people were sent [to RCA] by ITRI and were there for several months…they brought it back to Taiwan, and they continued to work on it in the laboratory, you know, small scale. It was not business. It was not production. And then they licensed the technology from RCA in 1975, and in 1980, which was five years later, they spun off a company. And so that company, of course, used the technology to some extent. And then seven years later, 1987, TSMC was started. And by 1987 the ITRI technology, which was first obtained from RCA, and then worked upon for 12 years between '75 and '87…in the 12 years they managed to get another generation and a half behind. So by 1987 the ITRI technology was already two and a half generations behind the leading technology. And, I mean, that's understandable because first of all RCA was not a first tier semiconductor company when it licensed its technology to Taiwan , to ITRI. And secondly, without a real commercial base ITRI was limited in improving the technology, in keeping up with the pace of other companies that are commercial companies, like Intel and TI and so on. They set the pace, and ITRI couldn't keep up with the pace. So when they got the technology it was a generation behind, and when TSMC got the technology in '87 it was already two and a half generations behind the leading level. Philips played the role as the only willing investor in TSMC, or the only willing significant investor, I should say. Now, I have to go back to the weeks after I arrived and took the presidency of ITRI now. A few weeks after I arrived in Taiwan , and became the president of ITRI, I was called in by Mr. K.T. Li. K.T. Li was a Minister without portfolio.

And K.T. Li turned out to be a very important man in my career in Taiwan , but he was not the Premier. And he said, "Well, we want to promote a semiconductor industry in Taiwan ." Of course, they wanted to do that in 1975 also when they licensed the technology. And Mr. Li told me, "You have a lot of experience managing large semiconductor businesses, so you seem to be ideal to start a new company in Taiwan in semiconductors. And you come back, tell me and tell the Premier how much money you need” etc, etc. In other words, give us a business plan, you know, in a week or so. So I went back to my office in ITRI and started thinking. And it was during those few days -- actually I didn't have to wait a week…a day or so later he wanted me to go back and make a presentation in four days, I think, not in week. It was during that time when the idea of the pure-play foundry gelled. Now, how did the pure-play foundry come about? I think that was going to be your next question. It was really two or three things combined together. First of all, or course, I had read some of Carver Mead 's writings. Carver Mead is a prominent researcher. He wrote several important papers. In fact, he published a book back in the late '70s on IC design and architecture, and he made a point that the design part could be separated from the technology. But he didn't advocate the advent of pure-play foundries, but he did make the point which would lead to the conclusion that you could start up a pure-play foundry. But the market really didn't appear to be there for a pure-play foundry. So I had read Carver Mead …[his] writings were known to me even when I was at Texas Instruments. But the only trouble is, of course, that there didn't appear to be a stable market for that kind of pure-play foundry. So that was one thread of thought as I paused and thought about the task that Mr. K.T. Li gave to me…he wanted me to present a business plan, he wanted me to start a semiconductor company.

And another thread was what I had already observed, closely observed, for three decades. I had been in the semiconductor business for three decades before I came to Taiwan . And I learned at close quarters how competitive the industry was, and how good some of the players were, companies like Intel, Texas Instruments. Even then the Japanese companies were very fierce also. So I knew how competitive it was, and how difficult it would be to carve out a niche for a new Taiwan company. So that was the second thread of thought. The third thread of thought was I paused to try to examine what we have got in Taiwan . And my conclusion was that [we had] very little. What strengths have we got? The conclusion was very little. We had no strength in research and development, or very little anyway. We had no strength in circuit design, IC product design. We had little strength in sales and marketing, and we had almost no strength in intellectual property. The only possible strength that Taiwan had, and even that was a potential one, not an obvious one, was semiconductor manufacturing, wafer manufacturing. And so what kind of company would you create to fit that strength and avoid all the other weaknesses? The answer was pure-play foundry. So maybe you could call it the least evil choice. The least evil because we've got no strength in all of these and I knew how difficult it was to compete even when you have strengths. It would be impossible to compete when you really had very little strength. But if we chose a pure-play foundry business model then we at least have the potential manufacturing strength that we can lean on. And we also manage to avoid the other weaknesses. We have no strength in design. We have a weakness in design. Well, we don't need design as a pure-play foundry. We've got no strength in sales and marketing. Well, sales and marketing for foundries is relatively simpler than sales and marketing for a conventional IDM company. At that time the IDM model was the conventional model. In fact it was the universal model. We got no strength in IP. Well, it turns out, at that time, that process was the part that was least vulnerable to IP attacks from other companies. Most of the IP disputes were about circuit designs. So in choosing the pure-play foundry model I managed to exploit, perhaps, the only strength that Taiwan had, and managed to avoid a lot of the other weaknesses. Now, however, there was one problem with the pure-play foundry model and it was a fatal problem, it could be a fatal problem which was, "Where's the market."

It was very poorly received. Well, people just dismissed it, you know, "What the hell is Taiwan doing? What the hell is Morris Chang doing?" They really didn't think that it was going to go anywhere. There was no market because there was very little fabless industry, almost none. No fabless industry. So who are you going to sell these wafers to? Who are you going to manufacture the wafers for? Of course, the obvious answer was the companies that already existed at that time, the Intels, and TIs, Motorolas, and so on. Now, those companies knew that they would let you manufacture their wafers only when they didn't have the capacity, or when they didn't want to manufacture the stuff themselves anymore. Now, when they didn't have the capacity, and asked you to do the manufacturing, then as soon as they got the capacity they would stop orders to you, so it couldn't be a stable market. And when they didn't want to make the wafers anymore, well, the chance was that it was losing money for them. The product was losing money for them. And so what do you want to do? Do you want to take over the loss, you know? And so that wouldn't be a very good market either. So the conclusion at that time, the conventional conclusion was that there was no market. Maybe this idea, this pure-play foundry idea, exploited the only strength you have, which is manufacturing, but there's no market for it. That's why it was so poorly thought of. What very few people saw, and I can't tell you that I saw the rise of the fabless industry, I only hoped for it. But I probably had better reasons to hope for it than people at Intel, and TI, and Motorola, etc because I was now standing outside. When I was at TI and General Instrument I saw a lot of IC designers wanting to leave and set up their own business, but the only thing, or the biggest thing that stopped them from leaving those companies was that they couldn't raise enough money to form their own company. Because at that time it was thought that every company needed manufacturing, needed wafer manufacturing, and that was the most capital intensive part of a semiconductor company, of an IC company. And I saw all those people wanting to leave, but being stopped by the lack of ability to raise a lot of money to build a wafer fab. So I thought that maybe TSMC, a pure-play foundry, could remedy that. And as a result of us being able to remedy that then those designers would successfully form their own companies, and they will become our customers, and they will constitute a stable and growing market for us.

MC: Our first customers were still the big companies, the IDMS except their business was not stable, you know, and it was pretty much leftovers from them. So the first few years of our start we didn't grow very fast at all. We started to grow fast when the fabless industry really started, and that was about '91, '92. We actually planned our business model in '85, I made my presentation, we start to raise money and so on, and we raised our money in '86, and we did not physically start operation until '87. Between '87 and '91 we depended mainly on the big companies giving us leftovers, and a few small companies, first in Taiwan , and a little later in the United States . And then from '91 on the number of fabless companies just started to mushroom. And you have the history of the Fabless Industry Association that we can look at. I mean, now there are hundreds of, maybe close to a thousand fabless companies in the world. Most of them, of course, are still in the United States, in Taiwan…well, in China alone there are maybe four or five hundred, so my thousand number doesn't even count all of the four or five hundred. And there are a few in Europe, and there's a couple, maybe, in Japan . So now there are about a thousand. But back before '91 the number was in, at most, two digits, less than a hundred, and most of them were struggling. And we, frankly, TSMC, had to mature a little bit before we could serve those fabless customers. And we had to develop the trust of our customers too, but technology, well, technology and trust. I mean foundry is very important to them. It's the source of their products, and they can't just go on and just trust you immediately. We had to prove it.

Copyright 2007 Computer History Museum, used with permission.

-- end --