Before you went to Princeton University you studied electrical engineering at the NTU. How did you decide to major in electrical engineering?
I decided to study engineering because liberal arts were my weakness, while science and engineering were my strengths. When I was selecting a major for college, I asked my dad, "Which engineering field do you think is better?" The field I understood best was actually civil engineering, because my dad had a bunch of civil engineering books at home. But he convinced me not to pursue it. In his rather old-fashioned thinking, the first few years of civil engineering require going to construction sites, often in remote areas; he believed I couldn't handle going to remote construction sites. He thought electrical engineering would have better prospects.
Another choice would have been medicine, but I didn't want to be a doctor. I had poor health when I was young and often went to see doctors. I could see that being a doctor was really such hard work. This was in 1964, and making a living was quite important. I thought engineers were able to find jobs easily. So when I filled out my major for college, by process of elimination, I put electrical engineering at the top.
Since you didn't know too much about electrical engineering, did you like it?
In the first year of college, all the courses were required. Which subject I was interested in depended mostly on the teachers, not the subject itself. At NTU, the teachers I liked were both EM (electromagnetic) professors. Teachers have a great influence on students, so I studied EM. When I first came to America, I didn't start studying semiconductors right away, and I didn't go to Princeton.
Which university did you go to first?
I went to Syracuse University. When I was applying to schools, both of my sisters were at Syracuse already. My brother-in-law was in the chemical engineering department, my oldest sister was working, and my other sister was studying animal science.
Except for a few very large schools, not every electrical engineering department has all the sub-fields. For example, Princeton didn't have EM. I knew roughly which schools would give scholarships to us. My grades at NTU were not the very best. If you didn't count the overseas Chinese students, I was probably in the top 20%. This wasn't the official ranking, but I had a sense of my own abilities compared to others.
At that time NTU asked each student to only apply to three graduate schools. They didn't want the top students to apply to a bunch of schools and squeeze out their own people. Under these circumstances, for someone at my level, my chances of getting into the first-tier schools such as MIT, Stanford and Berkeley were not great.
I looked at the rankings of the various electrical engineering departments' graduate programs. Syracuse was surprisingly ranked 15th in electrical engineering. I knew it had EM and was quite strong. Since I only applied to three schools, in the end, Syracuse gave me a scholarship, while the other two schools didn't give me money. So the choice was simple.
How long did you study at Syracuse?
I studied for one year. After the first semester, I discovered that because I wasn't an American citizen, it would be very difficult to find work in America after graduation, because most companies required security clearance. And Taiwan didn't have that kind of work either. So I decided to give up EM after the first year and transfer to another school quickly.
I applied to the State University of New York at Stony Brook in late January, and they gave me a fellowship from the National Science Foundation. I believe someone else turned down the offer and I was the replacement. I stayed at Syracuse for the summer to finish the work my advisor assigned to me. Normally I would get $240 a month, but in summer it was $480. Over three months, it was quite a boost, especially since my living expenses at that time were about under $100 a month.
Syracuse was one of the few schools where master's students also had to write a thesis. Generally, a master's degree is two years, but I finished in one year.
Before we talk about Stony Brook, when did you decide to study semiconductors, instead of other electrical engineering subjects?
Electrical engineering was divided into two sides. One side could be called applied math, like computer science and control, which needed a very good math background. The other was applied physics, which included EM and solid-state physics; semiconductors are one area of solid-state, and the most promising one. So when I stopped doing EM, I couldn't just jump to the other side. Plus, I had learned a bit about semiconductors when I was in college.
How long did you stay at Stony Brook?
I arrived at Stony Brook in September, one day before the start of classes. I met my advisor to discuss my future plans. He said I needed to take these courses. Afterwards, for the rest of the semester, he never came to find me again, and I couldn't find him either. He didn't seem to have much interest in me as a student. In the second semester, I did very well in my coursework—all A's. For the summer, my advisor gave me a topic that required a very strong background in chemistry. I said to him, "Why didn't you tell me to study chemistry in the first semester? My chemistry background is not sufficient to do the work." He said, "You can learn." He then went on a year-long sabbatical and handed me over to another professor; at that time, Stony Brook only had two professors for solid-state.
My advisor had some equipment that his previous student had bought to do this topic, but that student was nowhere to be found. I didn't know how to use the new equipment, and there weren't any documents either.
At that time, I thought I had only two choices. One was to go find a bunch of chemistry books and start studying. I had to learn organic chemistry to be at the level of a college graduate. But I felt doing those chemistry things was going down the wrong path—not something someone with my background should study, at least not something I wanted to learn. The other option was to transfer again. I figured as long as there was a school that wanted me, I didn't mind being a year late. However, if I wasn't successful with the transfer, I would have to look for a job. At that time, practical training was two years. You entered a company and hoped the company could help you apply for permanent residence.
So I was back to applying to schools again. I applied to three schools: Stanford, UC Santa Barbara, and Princeton. At that time, my dad said if I couldn't get a scholarship, he could support me for one to two years. But I knew that was half of his savings. UC Santa Barbara was my safety school, and they gave me a scholarship. I was accepted at Stanford but without funding. Princeton put me on the waiting list. And when I went back to Taiwan for the summer, I received a phone call from Princeton, telling me I had been accepted off the waitlist. That's how I got to Princeton in 1972.
What was your impression of Princeton at that time? After being at Princeton, was it different from what you imagined?
I had only seen Princeton from photographs. I didn't have any big thoughts when going to Princeton. I only knew that this time I couldn't mess up. I couldn't transfer again, and I had to succeed.
I tried to ask around, but contacting my former classmates already at Princeton wasn't very easy. We didn't have mobile phones like we do now. I didn't even know where they lived. And they themselves weren't clear either, because living in dorms meant you moved to a different place every year. There were no phones in the dorm rooms, only in the hallways. So finding people wasn't that easy unless you wrote letters.
My sister drove over from Morristown and dropped me off at the dormitory. I was very impressed when I arrived at Princeton. I felt the buildings were very similar to those English colleges you saw in movies. I was a bit disappointed that I wasn't living in the Old Graduate College. I lived in the New Graduate College next to it. The new and old parts were connected. Later I learned the new part had some advantages—the radiators were better and not as noisy.
Did you have a roommate? What were the students like?
Single rooms were rare. Most were doubles, which consisted of two rooms: a living room with two desks and a bedroom with two beds. It was quite spacious. My roommate turned out to be from Hong Kong who went to college in the U.S. He spent most of his time with his Hong Kong classmates and usually only came back to the room to sleep. He was in computer science. When I entered, he was in electrical engineering, but by the time he graduated, the two departments had already separated. I tried to learn Cantonese from him, but my language ability was too poor.
I quite liked the environment at Princeton, which I could feel within the first two weeks. First, everyone living there was very serious about studying. The atmosphere was very good, though of course there were one or two exceptions.
At that time, there was a guy in the physics department who people said was the best student in physics. Every day after dinner, he would just wander around looking for people to chat with. No one had time to talk with him, and everyone tried to avoid him. He was just so smart that he didn't need to work very hard. He graduated in three years, and afterwards joined the physics department.
Some people said Princeton people were very boring, that everyone was too serious. Maybe that was me as well. But I felt very comfortable. Of the four universities I attended, Princeton was my favorite. I felt at home. The graduate students at Princeton were all very gentle. I feel those few years were a highlight of my life. It was a very good four years.
There were some relaxing situations. We had a bar in the basement that opened at nine o'clock, where you could go drink beer and eat peanuts. But it only stayed open for a little over an hour. Of course, you could continue to stay there, but no more beer was served. I only went once to experience it. After dinner, I usually went to the large lounge to read the New York Times for half an hour. Afterwards, I would go back to my room to study.
I know in the old days the graduate dining hall required the students to wear black robes. Did you have to wear a black robe?
At that time, there were two dining halls. Both sides had the same food. One was formal where people served you. The servers would pour water for you, and when you finished drinking, you had to wait for them to come to refill it. I felt very awkward and didn't like it. I only went to the formal dining room a few times. I didn't buy a black robe and only borrowed it from someone else. There were people who liked the formal setup. The doors opened and everyone would be seated about ten minutes later. We would pray together then eat. It was very formal, just like in the movies.
The students at the two dining halls were different. The reason they opened an informal one was the number of students had increased too much; the formal one couldn't fit everyone. Later on, the informal one did better business than the formal one.
I often went to the dining hall alone instead of as part of a group. Sometimes I would meet people from different countries. The dining experience really depended on whether the person next to you had some interesting topics to talk about.
There was one table where people all spoke Chinese. They were all Chinese students with only one foreigner, who was a student of Wen Fang 方聞 studying Ming dynasty paintings. I think he took eight years to finish his PhD. He spent two years in Taiwan. His Chinese was very good. He said when he first came to Taiwan, no one could understand him because he spoke with a Beijing accent.
It sounds like your life was very regular at that time. You would go back to the dining hall for dinner at the same time every day.
At that time, if you didn't go back to the dining hall, you had nothing to eat.
So during the day you were just at E-Quad. Did you walk or ride a bicycle? The E-Quad is very far from the Graduate College.
Do you know why the Graduate College was placed so far away? The first Dean of the Graduate College really disliked the President. He deliberately put it in a very remote place, even separated by a golf course. I bought a second-hand bicycle, which was only useful for going from the Graduate College to the edge of the main campus. During class time, there were many people on the main thoroughfare, and you couldn't really ride fast. Sometimes I had to get off the bike and just push it. Also, parts of the campus have a steep slope. So I ended up selling the bicycle.
.jpg)
Where did you eat lunch?
Usually lunch was a sandwich that I made myself. At Princeton, the cost of the dorm plus dinner was very expensive—$200 was gone. At that time, prices were rising very quickly. I remember in the first year we had London broil once a week, and they even let us take seconds. In the second semester, no more seconds. After another year, sorry, no more of that stuff, only meatloaf. The inflation at that time was incredible, over ten percent a year. In the first semester, I thought the food was very expensive but the quality was very good. After the second semester, I felt things started going downhill.
Did the students from Taiwan often get together?
Basically everyone was busy with their own things. People were in different departments, so it was very hard to discuss with each other. So it was just during dinners everyone would gather for a bit.
Several of your classmates from NTU's electrical engineering department were also at Princeton.
Three people from my class went to Princeton: Chin-Tay Shih 史欽泰 *75, Shang-Yi Chiang 蔣尚義, and Zhuang. Among these three, I knew Shang-Yi the best. But he and I didn't overlap at Princeton. He transferred to Stanford after a year at Princeton. I did overlap with Chin-Tay.
How did you like the coursework at Princeton?
The first semester was mostly required courses. Princeton had a rule, at least in the electrical engineering department, to cut people after the final exams of the second semester. Usually 30% of the students would not continue. Actually, they would call you after the midterm exams. If the results were not good, you would need to be prepared to leave.
So some people would just get a master's degree and leave?
Whether they gave you a master's degree depended on how you chose your courses. With 24 credits, you could get a master quite easily. But if you didn't take enough courses, you would get nothing because Princeton didn't have summer school. You couldn't make up the credits. I didn't know Princeton was like this before I arrived.
In your first year, did you have to take courses from other departments such as in physics?
I mentioned before that electrical engineering had two sides: applied math and applied physics. For those in the applied math section, it was mandatory to take courses in the math department and have math as the first minor. They focused more on communication, so their second minor was usually computer science.
For us on the applied physics side, two of the four courses were in the physics department. Physics was our mandatory first minor. You could choose your second minor. Ching-Chu Chang 章青駒 *76 was the only one who chose math. The math we learned in college was very far from real math; Princeton's math was very strong. So you had to compete with students in the math department, which was extremely tough. After he went and studied math, our advisor (we had the same one) told all of us not to take math anymore.
Besides math, the other choice was chemistry. So I took a course on physical chemistry. It wasn't real chemistry. What they were actually teaching you was the physics needed to do physical chemistry.
With our two physics classes, the system basically forced us to compete with the students in the physics department, which was very strong. The more senior students in our department told us to pay special attention to the two courses in physics. If you didn't do well, you might be cut. The courses in our own department weren't a problem.
I read an interview with Shang-Yi Chiang 蔣尚義 where he said Princeton was too academic and that was why he transferred to Stanford. Did you feel the same way about Princeton?
I think Princeton's teaching method, which focuses more on fundamental things, is correct. The chance that what you learn and what you use later will be the same is very small. And after you start working, what the industry needs, especially in semiconductors, will keep changing constantly. So the most important thing is to get the fundamentals right and get some concepts right.
You might say an undergraduate degree is enough, but in reality undergraduate coursework isn't quite sufficient. Graduate school gives you a very good base. The chance that your graduate school thesis topic is the same as the R&D topics you work on later is almost zero, unless you stay in the university. Like me, it was completely different, because what industry wants and what school teaches are two different things, at least in semiconductors. So what do you learn when you are doing your thesis?
What I learned is when you get a problem, how to look at that problem. This problem must be new, otherwise it won't be your thesis topic. So how do you look at this problem? How do you approach it? Graduate school gives you the time to tackle the problem. Once you enter the industry, they won't give you that much time. At school, you can spend one or two months pondering how you should tackle and study the problem. You go discuss with your advisor. This is a training process.
Shang-Yi later went to work on III-V compounds. One thing Stanford was definitely better than Princeton at was Stanford had more choices. From a ranking perspective, the semiconductors program at Stanford was also better. And if you were interested in entering the semiconductor industry, then you should go find teachers who were already collaborating with the industry. This was another area where Stanford was stronger. Princeton was more remote. But I don't think Princeton's teaching method was wrong. I think it was right.
Did you have to be a Teaching Assistant?
Our department didn't have many Teaching Assistants (TAs). But Princeton's system required all Ph.D. students to have one semester of TA experience. People usually did it after passing the preliminary exam. I was an assistant for the undergraduate lab. Being a TA wasn't a burden, although I knew I wasn't very good at explaining things. I did the preparation work well and made sure every instrument was working before the class started. There were 12 people, all quite good and very serious. Towards the end of the semester, I asked everyone what they planned to do after graduation. To my surprise, not a single one was going to continue in EE. Several wanted to go to medical school, and a couple wanted to get an MBA. I wondered if I had a particularly strange class. My advisor said this problem had been bothering them for two or three years already. We didn't know the reason. Maybe the students just decided that EE didn't have much to offer.
Can you talk about your advisor?
I didn't get to choose an advisor. Our applications went to the EE department, which did the first screening. The remaining applicants would be discussed by the professors. The professor with the funding would decide which students he wanted. However, when you got the scholarship, the school didn't tell you who you would be working with. This was because they were afraid of some changes later. You only found out when you got to school.
My advisor Walter Johnson came to Princeton in 1937. I was born in 1947. He was well established. He didn't have a Ph.D., only a B.S. At that time, Ph.D.s were not that common. Nevertheless, it was still uncommon for someone with just a BS to become a professor at Princeton, which showed he was very special. His technical knowledge was very good. He worked at Bell Labs and was later recruited to Princeton. He wrote a book on electromagnetics. But I never read that book because it was already outdated by then. When he arrived at Princeton, the EE department was slowly starting to change—moving from industrial electronic power and slowly changing to semiconductors and communications. He was involved in a lot of curriculum planning—how should it change and how should an electrical engineer be trained?
My other advisor was Murray Lampert, and he didn't have a Ph.D. either, just a master's. He was very, very sharp. Johnson thought more deeply but not very fast. Lampert was fast and fierce. We were all afraid of him, especially during preliminary oral exams. He was very tricky. At my exam, he kept pressing me with one question about silicon: if an electron hits a silicon atom, how much energy is needed to knock it out? I said, "I don't know, I'd have to look up tables to calculate." I thought this question was unnecessary. He forced me to answer. I told him three times, "I don't know," before he gave up. He then asked me whether hitting a silicon or germanium atom requires higher energy. I said germanium. "Why?" "Because it's heavier." He said, "Correct," then told me the answer. This was a trap. If I had answered randomly by guessing, I'd be in trouble. Basically, he was saying when you do academics, you shouldn't guess. If you don't know, say you don't know. But he still wanted to test whether you understood the theory. What does it take to knock something out? Someone studying chemistry might think it's related to chemical bonds. Actually, it has nothing to do with that in this context—it's momentum and energy. Heavy things need more energy. He was testing you on this.
At that time, we had a weekly seminar where our professors would invite outside people to come speak. Everyone had to attend as it was mandatory. Missing once or twice was okay. Everyone knew the two old professors would look around the room to see who wasn't there. These talks let us get in touch with the outside world. The people who came were usually quite famous. I remember one time they invited a rising star from Bell Labs. He was in his thirties. After he spoke for a short time, Professor Lampert started questioning him. He pinned the speaker down and didn't really let him continue. He thought the speaker's assumptions were wrong. He paid a great deal of attention to assumptions. Assumptions must be correct; otherwise, everything that follows is garbage. In the end, Lampert still allowed him to present his results even though the assumptions were wrong.
This was also an education—some things you take for granted, but after you put them together and integrate them, you need to check again. Otherwise, you have contradictions in the middle, and what you produce is garbage. I heard Tsung-Dao Lee 李政道 was even fiercer than him—in such situations he wouldn't let people continue speaking. Why waste everyone's time? Lampert was still more polite.
Our teachers knew people from Bell Labs and industry, which made job hunting more convenient. But when you came out in the wrong year, it still didn't work. Semiconductors are very big now, but in the seventies, semiconductors were very small compared to other things. It was highly dependent on government funding. When the government caught a small cold, we got pneumonia. So I was very nervous before job hunting, because two classmates ahead of me went out and couldn't find jobs. Ching-Chu Chang 章青駒 couldn't find work. If such a smart guy couldn't find work, wasn't I doomed? He returned to Taiwan. Han-Sheng Lee *76 first went to Southern Methodist University as a postdoc for nine months or so before going to GM's Research Lab.
When did you start looking for work?
In 1976. I gave Professor Johnson periodic updates on my progress. He said I had enough data from the lab work and could start to write my dissertation. The fact that he felt I had enough was the key. After all, I could have continued to do more, such as study another material.
Johnson also asked if I had financial problems. By then I was married, and prices kept going up. My monthly stipend only increased to $300, and I couldn't get into the university housing. Princeton graduate housing guaranteed first-years a place to live. Second-years had the highest priority. Third-year students needed to be lucky. Fourth-years had zero chance. I found a big room on the side of a big house. The owner partitioned the room and rented it out for $280. After paying some taxes and bills, I basically had nothing left. We were eating into our savings and had to borrow money from home. Professor Johnson told me to look for a job and to write my dissertation while working. So I quickly went job hunting.
Can you talk about your dissertation?
In the EE department at that time, basically the school expected your advisor to support you plus people in the department who couldn't get money. Princeton had an overhead of 100%. If you got two dollars from the government, one dollar went to the university and the other dollar went to the engineering school. Your one dollar got divided again because some of it had to go to pay for the overhead in the E-Quad. Basically, all the private universities were like this. Frankly, what we chose to research was all decided by DARPA (Defense Advanced Research Projects Agency). How DARPA decided which topic to assign, I don't know. Our group was all looking at the interface between the semiconductor films and insulating films. Several people's theses, such as Ding-Yuan Yang 楊丁元 *75, Ching-Chu Chang 章青駒, and Han-Sheng Lee, all revolved around this. Our work was monitored by the Naval Lab.
At that time, the U.S. Department of Defense systematically wanted to foster electric engineering talent. Their basic research gave money to schools to do some relatively less applied and more fundamental studies. At that time, MOS was the preferred transistor. There were two types of transistors—MOS and bipolar. Everyone was optimistic about MOS being able to do large-scale integration. Around 1975, Intel's 8080 came out, the first one that could be used for PCs. Prior to that, these things existed in the military but weren't mass-produced. The Department of Defense was very interested in studying these things and using them to make military equipment.
So job hunting was equally difficult for you then?
I thought it would be difficult, but it actually wasn't. In just one year, the whole environment had improved, and I quickly found three jobs. The first was at GM Research Lab where Han-Sheng was already working. Apparently he was doing well. After only six months, his boss called Professor Johnson to ask if there were more people who could come work. Two weeks later, I had an interview at Texas Instruments (TI). The third offer was from David Sarnoff Research Center at RCA. TI's pay was the lowest, only $20,000. Sarnoff was $23,000, and GM was the most generous at $25,000. I asked Johnson which one I should accept. I didn't want to go to GM because semiconductors weren't the main part of their business, and Michigan was too cold, colder than Syracuse. I really liked the Princeton area and wanted to stay in New Jersey. But Johnson said RCA was already in the second tier. If I really wanted to do semiconductors, I should go to a mainstream place like TI or Intel. So even though the salary was the lowest, I accepted the offer from TI. Johnson paid a lot of attention to the industry. I really appreciated Johnson for giving me advice from his perspective as an elder to a young person. You couldn't ask for more from an advisor.
In retrospect, going to TI was the right choice?
Yes, otherwise I would have gone in circles and wouldn't know where I'd end up. When I left Princeton, Johnson told me, "When you are at work, you'll find some people don't like you. And if you can't find a reason why, then they just don't like you. Avoid these people if you can. If you can't and if you have to work with them, then change job." He was talking about racial discrimination. He said you can't win on this thing. For him to be willing to say that was really rare. I don't think any graduate school advisor today would tell their students this kind of thing. He also said, "Don't let this affect your thoughts about other people." This was another very important point. He was a remarkable advisor. After I started working, I would go back to visit him once a year.
How long did you work at Texas Instruments?
A year and eight months. I finished my dissertation, but I still had to go back to Princeton for my thesis defense. At Princeton, Professor Johnson asked me how things were at TI. I told him I would arrive at work at 8:30 in the morning, and I would be dealing with things in the processing lab as well as the measurement lab. Except for lunch, I didn't sit down at my desk until 3:30 in the afternoon. Then there were usually a pile of things for me to look at. This was pretty good for the first year because I learned a lot. But I didn't have time to think.
Johnson asked if this was just my boss or the whole company. I said the whole company had this trend, but our group was especially busy. He said, "Do you want to change companies?" I said I wanted to go to a real research lab. He said there were only two places—Bell Labs and IBM. He said he knew people at both places and would make some calls. So I later went to work at IBM. I ended up working at IBM for thirteen years before coming back to Taiwan.
The interview was held in 2025 in Taipei. The interview has been translated from Chinese and edited for clarity.
