Oral-History:Sung Mo (Steve) Kang

About Sung Mo (Steve) Kang

Sung Mo (Steve) Kang was born in Yang Pyong, Korea, a suburb of Seoul, later moving to Seoul in seventh grade. After high school, Kang entered the Korean Air Force where he learned about electronics through a US Air Force correspondence course, adding to his interest since early childhood in math and science. Kang left the military in 1965 and began attending Yonsei University in electrical engineering, coming to the United States in his senior year to attend Fairleigh Dickinson University, a sister school of Yonsei. He received his masters at SUNY Buffalo, and attended Berkeley for his doctorate working on memristive systems, finishing in 1975. After graduating, Kang taught for two years at Rutgers University before going to work at Bell Labs in 1977. Involved in microprocessor development, Kang stayed at Bell Labs until 1985, when he moved back into academia at the University of Illinois, Champaign-Urbana, eventually becoming department chair. In 2001, Kang went to UC Santa Cruz as Dean of Engineering, and began serving as Chancellor at UC Merced in 2007. Kang also joined the IEEE in 1972, and has been editor-in-chief of a journal and society president.

In this interview, Kang discusses his career in both industry and academia. He talks about his experiences in Korean and American universities, and at AT&T how he came into management. His shift from AT&T to Illinois – and industry to academia – is also covered, with the difficulties involved both personally and professionally. Kang also discusses engineering education, both in building an engineering department at UC Santa Cruz and the general state of education and its challenges. He emphasizes the importance of links between industry and academia, while he also talks about technology relationships between the US and Korea, with wider implications to globalization and education.

See also: Sung Mo (Steve) Kang’s First-Hand:The AT&T BELLMAC-32 Microprocessor Development

Note: The interview was conducted in  two parts, the first on 17 December 2009 and the second 4 March 2010. There will be a note in the text where the second part of the interview begins, with a repetition of the date.

About the Interview

SUNG MO (STEVE) Kang: An Interview Conducted by John Vardalas, IEEE History Center, 17 December 2009 and 4 March 2010

Interview #525 and 525A for the IEEE History Center, The Institute of Electrical and Electronic Engineers Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center at Stevens Institute of Technology, Castle Point on Hudson, Hoboken, NJ 07030 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Sung Mo (Steve) Kang, an oral history conducted in 2009 and 2010 by John Vardalas, IEEE History Center, Hoboken, NJ, USA.

Interview

Interview: Dr. Sung Mo (Steve) Kang

Interviewer: John Vardalas

Date: 17 December 2009

Location: Merced, California

Background, Education and Korean Air Force

Vardalas:

This is December 17th. I'm interviewing Dr. Kang in his office at UC Merced. Thank you so much for agreeing to be part of this.

Kang:

Thank you.

Vardalas:

I want to break it up into several sections: your pre-university years, we want to find out all about the person, who the engineer is; your undergraduate and graduate student years, your life at Bell Labs. We're very much interested in what you did at Bell Labs; your life as an academic, as an engineer academic; and then finally, I'd like to examine your experience as an IEEE member, the society you're involved in, etcetera. So let's start with the beginning, then. Where were you born and where did you grow up?

Kang:

So I was born in Korea, in a suburb of Seoul. Seoul is now a mega city. There is a town called Yang Pyong in Kyunggi province. It's a suburb of Seoul, about 35 miles from Seoul. That's where I was born and raised. I went to school there until seventh grade. Then the family moved to Seoul. So the rest of my life in Korea was in Seoul. I graduated from high school and went to college in Seoul. In fact, I first went to the Korean Air Force before college as a private soldier. That's where I learned about electronics.

Vardalas:

How young were you when you first became interested in science and mathematics?

Kang:

I really liked mathematics and science when I was even in elementary school and junior and high school. And I also grew up with my grandfather. He always used to say that Sung Mo - that's my first name - “you will study electrical engineering, go to US, get a doctorate and become a professor.” Those were his instructions. I think that was his dream. But he didn't realize his dream, and he wanted to maybe impose his dream on me and he saw potential and he encouraged me. In the early days, he was fighting for Korean independence. He was not able to realize his dream, but he saw a grandson who had a potential, so he used to say that.

Vardalas:

He was encouraging. Were your parents professional?

Kang:

My father was not in engineering. He worked in a bank, not in engineering at all. My mother was a bright lady. She was a full-time homemaker, so I did it pretty much on my own with encouragement from my grandfather. So I served three years in Korean Air Force right after high school. My dream was come to the U.S. for college study. When I was in Korean Air Force, I was in the communications field. My boss was a Sergeant, Sergeant Kwak. Aspiring to be promoted, he took a US correspondence course from the U.S. Air Force, all in English. But he couldn't understand English. So he asked me to do it for him. I did all the homework. It was a very extensive course on electronics and I learned electronics well at that time.

Vardalas:

Do you remember the correspondence course?

Kang:

It was an Air Force Correspondence course.

Vardalas:

Well, U.S. Air Force Correspondence?

Kang:

Yes.

Vardalas:

Really?

Kang:

Yes, a very, very good course for practice. So I think he got almost a perfect grade.

Vardalas:

Thanks to you. And where did you learn your English? In school?

Kang:

In school we had English education when I was in high school and a course in the university. But, you know, the university I went was founded by American missionary, Dr. Horace G. Underwood. It was a good university. That's where I learned English. By the end of 1965, when I finished my military duty I took the Yonsei University entrance exam. It was very competitive. I was fortunate enough to be able to make it into the electrical engineering program. I attended Yonsei University until the seventh semester. My university had established a sister relationship with Fairleigh Dickinson University in Teaneck, New Jersey. So that's where they sent me as a senior student. I came in August and started my senior year at Fairleigh Dickinson. Studied one year and I did complete my bachelor's degree at Fairleigh Dickinson.

Vardalas:

That, that was my next question, how you came from Korea to Fairleigh Dickinson. Now I understand.

Kang:

Fairleigh Dickinson promised to accept one student and I was selected as the student. That is how I came to Fairleigh Dickinson University.

Vardalas:

How did you find your high school education in terms of encouraging science? Did you find teachers encouraged you in high school? Was it a good environment for you?

Kang:

Yes. In fact, I was a good student in high school. Usually you have a teacher in charge of a particular class in high school. He was a math teacher Mr. Cha. Cha was his last name. He was a very good teacher and I think I did well. I always liked physics. More than chemistry, in fact, so I knew electrical engineering would be a right field for me.

Vardalas:

And why didn’t you go into physics?

Kang:

Grandfather said electrical engineering. At that time in Korea after war, engineering was actually more popular than physics, basic science, because you probably earned better salary after graduation.

Vardalas:

After three years at Yonsei University and one year at Fairleigh Dickinson, how would [you] compare the two experiences? What struck you at the time?

Kang:

See, when I went to Yonsei University, I learned a lot. I think we had many great teachers. Definitely I learned a lot. But then society was in political turmoil. So many students demonstrated on the street. In fact, the first semester of my senior year was consumed by demonstrations, there was unrest in the society. In Korea, at that time, people studied very, very hard to get into good college. Once they got into good college, they were not as serious as in high school, in fact. Once they got into good college, they expected something good would happen. Whereas in the U.S. I was impressed when I first came in the senior class, everybody came to class, and everybody was so quiet and attentive to teachers. That was a distinct impression that I had at Fairleigh Dickinson.

Vardalas:

Now, in '65 to '68, you were in Korea studying at the university or -

Kang:

'66. So '66 to '69.

Vardalas:

Did you feel at the time you were getting the current subject matter in electrical engineering?

Kang:

Yeah. Our textbooks were in English. Some were in Korean, but many of our professors, they were U.S. educated or England educated professors. You know, I felt really privileged to have a good education. We used always the English textbook, and then I remember in one course the textbook was in Korean. It was a translation of a book on automatic control theory. I couldn’t figure out one page - it really didn't make sense so I had to go to library, look at the original version in English. Turns out they skipped a few pages in translation.

Vardalas:

Did you protest with the students at all? Were you a serious student? Did you go out at all?

Kang:

I was not on the street, actually. Engineering students mostly stayed on campus.

SUNY Buffalo and Berkeley

Vardalas:

They were mostly - ah, okay. All right. Because, because you came - 'cause the next thing is you went to Berkeley at a time when things were going on, too.

Kang:

Right, so I was an observer on the Berkeley Street.

Vardalas:

So tell me how, how did you - what was the process that got you from Fairleigh Dickinson to Berkeley?

Kang:

So, you know, at Fairleigh Dickinson I finished my bachelor's degree and then I went to SUNY Buffalo, in upstate New York, for my master's degree. At that time, Nelson Rockefeller wanted to make SUNY Buffalo a Berkeley on the east coast. He brought in many Nobel prize winners and they tried to invest into SUNY system, so SUNY was a good system at that time especially the Buffalo campus. I went to there. It was so cold, you know. I ended up finishing master's degree in one year at Buffalo, and then I got married in 1972. My wife is also from Korea. I met her at the Teaneck campus of Fairleigh Dickinson, and we got married in 1972 and I was debating whether to go Berkeley or Columbia. Everybody said go to Columbia because it's nearby. But Berkeley was always in my dream. It's a wonderful campus, so our honeymoon was crossing the country, driving. I came to Berkeley in the summer of 1972.

Vardalas:

Well, that's an adventurous honeymoon.

Kang:

It was. It was fun.

Vardalas:

Yes, driving across the country. So what particular area of electrical- engineering did you zero in on for your master's program?

Kang:

At that time, I was very much interested more in applied mathematics, automatic control was an area of my interest. Control and system theory. At Buffalo I met a very young and dynamic professor. Although his Ph.D. from Cornell was in control theory, he was getting into bioengineering. So my master's thesis was related to bioengineering. It was on the stability properties of a Purkinje fiber model - the Hodgkin Huxley equation. So that was my master thesis topic.

Vardalas:

Ah, and then so then in Berkeley, did you continue this direction?

Kang:

No, actually. When I came to Berkeley, I was still interested in similar fields so I was going to study in the system field or control theory. I met my advisor, Professor Leon Chua who was in nonlinear circuit theory. He was my advisor, and I wanted to work with him on nonlinear circuits, especially modeling and analysis of nonlinear circuits. So that was my PhD thesis topic. I remember he once asked whether I could look into “memristive” systems, which was a very new field at that time. He had defined a new element called memristor.

I worked on the extension of memristor. Prof. Chua presented that topic to me as an extension of memristor. I remember I asked as a naïve graduate student, “professor, if I study this field, which is so new and not many people know about it, will I be able to find a job?” He said, “Steve, don't worry, no matter what you do, if you do well, there will be always job for you.” But anyway, I did it and it turned out it was linked to my master thesis because Hodgkin Huxley model was a memristive system, which I later found out. We didn't know it in the beginning. Through my investigation I made that link and that became very valuable information because it's related to neurmorphic computing.

Vardalas:

So the thesis topic was essentially handed to you by your supervisor? How did you decide on the thesis topic?

Kang:

Well, he asked me what I did in my master thesis. I showed my master thesis, in particular a paper that we published in a journal, I think it was Computers in Medicine journal. Prof. Scott and I had published one paper based on my master thesis. Prof. Chua looked at the paper and he may have guessed that there may be a potential link. He suggested “what about working on this one.” So it was on mutual agreement. I said “let's work on it”. That's how I got in. But my thesis was not totally on memristive systems. I did something else also.

Vardalas:

Thinking back on the whole Ph.D. process, can you recall the highs and the lows? Did it go smoothly? Were there days you were saying “oh my god, I'm not going to solve it?” How was the process for you?

Kang:

No one has done anything in memristive systems. It was so new. I had to dig deep into it. There was no one to talk to. It was a difficult time. For instance, I used to go to the mathematics library to see whether I could find any literature relevant to it. It was under the general topic of modeling of nonlinear systems. Can you create nonlinear differential equations that will describe behavior of a devices or systems? I was in the Berkeley mathematics library and I found one book. The title was Synthesis of Dynamical Systems. I said “This is it. Let me study this”. The opening remark, when you opened the book cover, was “Dedicated to my wife a dynamical system that cannot be synthesized.” [Laughter] So I remember that.

Vardalas:

You remember those words.

Kang:

Those were good memories. It was challenging. But you know, once you find something new it’s very exciting.

Vardalas:

Yeah, okay, okay. So you finish your thesis.

Kang:

I finished the thesis in 1975. In fact you know, I finished by the summer of 1975 but I remember in February (my birthday's February 25. I was born in 1945, so that was February 25, exactly my 30th birthday) I received a phone call from my advisor. He said, “I have a job for you. Would you take it?” And that happened to be a job at Rutgers University. They were searching for new assistant professor and so that's how I got my job. You know, it was good to come back to New Jersey.

Vardalas:

I was going to ask you how you - I mean, why back to New Jersey, but now I understand.

Kang:

At that time, the job market was very tight, and I was most fortunate to have that opportunity, even before I graduated. I interviewed with the new dean at that time and took the job in summer of 1975 and came to teach at Rutgers.

Vardalas:

Because I teach at Rutgers.

Kang:

Oh, do you?

Vardalas:

The History Center is at Rutgers University.

Kang:

Right, right. Good, good.

Bell Labs

Vardalas:

You're two years there, and then very quickly, you wind up in Bell Labs.

Kang:

Bell Labs.

Vardalas:

Why did you make that transition? How did it happen?

Kang:

Although I did study and I enjoyed the teaching, I also thought “gee, industry experience would be also very good.” Also at the time, the assistant professor salary was very, very low. So my friend who moved to Bell Labs from Berkeley Lab recruited me. I interviewed at Bell Labs and I took the job with a much higher salary. Bell Labs at the time was a very respectable place. You could get a really good research environment and a good experience. So that’s how I took the job in 1977.

Vardalas:

You had heard a lot about Bell Labs, obviously.

Kang:

Absolutely.

Vardalas:

What were your expectations when you walked in and what was the reality? Did you have any preconceived notions of what it would be like to work at Bell Labs and what happened when you got there?

Kang:

Right, so the first place I went to was Holmdel.

Vardalas:

You were in Holmdel.

Kang:

I spent the first day in Holmdel, because that's relevant to my own background in systems theory, systems study, nonlinear systems, so my first assignment was doing satellite based private network design.

Vardalas:

Okay.

Kang:

I didn't study that subject matter at Berkeley or in my other graduate studies, but as I delve into it and found a great link, actually. The way they have done network design was based on experience and at the time all network design was based on three basic nodes based design. So I studied on my own - the theory based on statistics, stochastic models and then I formulated this network design problem and with a great struggle after in depth study, I applied what I learned at Berkeley. It was an optimization theory and I remember the Lagrangian multiplier technique led me to decompose this complex network into sort of a linked three node networks. So you know, I finished that design and then what people said was gee, this proves what Bell Labs people have done in the past. This is really correct because it showed that mathematically the conventional design was an optimal design.

Vardalas:

Oh, interesting.

Kang:

Because before, it was based on experience, but, you know, through very elegant network modeling and then formulation of nonlinear equations, applying very vigorous known optimization theory led me to decompose this complex network into coupled networks of three nodes.

When the solution was given, it was very impressive to a lot of people in Bell Labs. So for that, I even got a best paper prize when I presented it at a conference.

Vardalas:

What kind of environment was it for a young engineer? Was there any mentoring? Was it a place where they helped you learn, or was it a place where you were just thrown in and told you're on your own?

Kang:

Well, it was a good environment because there are so many talented people. We had many physics majors, very bright people, so it was a team of talented people. Most of the theoretical people didn't realize what network planning was about. It involves a cost model and a sort of systems analysis. Usually when you said costs to scientists, it was not really motivating on the surface. I took a more practical approach. Whatever problem, can you really come up with a very good formulation of the problem and then can [you] come up with a vigorous solution. That's what I did, very, very satisfying. I did that for one year, and then I wanted to look forward. What was really the upcoming technology? Actually, if I stayed with networks, that could have been also a very flourishing new field. If you look at today's world, it's all about networks. But I wanted to get into a new field – integrated circuits field. So I asked to transfer to the integrated circuits division. And they said “oh, you didn't do integrated circuits in your Ph.D.” But my boss in Murray Hill, Bernard Murphy who came from England, he just grabbed me. He saw the potential. He himself was a biophysicist I think, and got into this field and became very, very successful. That was really the beauty of Bell Labs. They saw the potential of a person and then trust that person to get into new field, explore, and do new things. That was a tremendous experience. After my research experience even in university has been always in that mode. I was not afraid to get into new field and to do something new.

Vardalas:

So Bell Labs encouraged you to experiment with new ideas. The R&D group you were [with] was integrated circuits [in] Murray Hill?

Kang:

Murray Hill.

Vardalas:

It wasn't part of the basic research part.

Kang:

No. It was area 20. It was an electronics division. I was in the electronics field. The physics and the material scientist or mathematicians, communication, information theory people, they were in area ten.

Vardalas:

Ten.

Kang:

Ten, area ten. Electronics it was in area 20.

Vardalas:

Twenty, I see.

Kang:

The people who recently got the Nobel Prize for CCD. They were in area 20.

Vardalas:

Was there much communication between these groups? Did you find that in Bell Labs, all these various groups communicated? Or were they just silos?

Kang:

No, we had a lot of good communication.

Microprocessor Development

Actually, a particular project that I was in was on microprocessor development.

Vardalas:

That's what I wanted to ask you.

Kang:

It was a fully CMOS, 32-bit microprocessor chip, which at that time, no one had done. In microprocessor business, Motorola and Intel were ahead, but they hadn't done that in full CMOS.

Vardalas:

So, so that's the significance of what you were doing was that it was done in full CMOS?

Kang:

A 32-bit microprocessor chip in full CMOS technology. That was a very exciting project.

Vardalas:

So what were the challenges? Obviously, it wasn't the simple thing to do or otherwise everybody would have done it. Do you recall challenges?

Kang:

First of all, experienced companies didn't believe in CMOS. They thought NMOS was the way to go. That's what Motorola and the Intel were doing. In Bell Labs the people figured out that, in terms of power dissipation, CMOS was the way. Although it may take more chip area, CMOS would consume less power. And because of that, you could make even more efficient circuits. When they used NMOS, they found that power consumption was too big. Then they tried to add more circuits to save power using power management, but that made the chip bigger. So CMOS was the right choice.

Vardalas:

Then correct me if I'm wrong, but why did the center of CMOS development and industry move to California? Why didn't it stay at Bell Labs?

Kang:

Okay. At that time AT&T was being split into two pieces, right. Judge Green wanted to split AT&T into long distance business and the local telephone operating business. AT&T really didn't want to do it. But when pressure was so much, they thought “okay, the only condition that can help is having computer business.” They wanted to compete with IBM. So in order to do the computer business this 32-bit microprocessor development was critically important. So it was a top project within Bell Labs to develop a 32-bit microprocessor chip, tremendous pressure. I was a project manager at that time. Later I was a technical supervisor and our team really worked day and night.

Vardalas:

How many were in your team, do you recall?

Kang:

In Murray Hill alone there were 30 people and another 30 people in the Holmdel doing logic design and then the Indian Hills systems people. So there were more than 100 people involved.

Vardalas:

And what group were you supervising, the whole project or the -

Kang:

Yeah, the - mainly supervising the circuit and physical design people in Murray Hill plus some interface people in Holmdel. It was a big project.

Moving into Managament

Vardalas:

<flashmp3>525 - kang - clip 1.mp3</flashmp3>

I want to pursue this project some more, but before I do that, how did you find the transition from being a bench engineer to being in management? Now you have to supervise people. How did you cope with that? Was it easy or hard for you?

Kang:

It was a big challenge. I was picked because they thought I may be technically capable. Then the challenge was to harmonize the people who were all under tremendous pressure, right (everybody was under tremendous pressure), and yet achieve something. In fact, before I got involved to manage the project, there was a big failure. Once we failed.

Vardalas:

Really?

Kang:

In the beginning, people underestimated the complexity of the job. At the time Lee Thomas, he's no longer alive, was in Holmdel. He was in charge of the total project. The people doing the chip architecture design there thought, “oh, don’t worry. Our architecture is so good. The circuits will be very, very simple.” So people operated under that assumption. So the assumption was if you have PLA arrays already and a data path, logic would be so simple that you can just simply connect them and you will have a microprocessor. It turned out that this was not the case. Logic grew and then interconnection lengths became very long. We were targeting in 3.5 micron technology, 4 megahertz clock frequency. When chip was made it was barely running at 2 megahertz. Total failure. So [the] person who was in charge almost got fired. So he asked for a second chance. He then decided to put me as one of the project managers in charge of Murray Hill and interface with the Holmdel team.

So it was a really critical, important project. I made a commitment to make it successful, and at that time, I was reading some books like Soul of a New Machine, a famous book. There was another software book which had a chapter on “How does a Big Project Get Delayed a Year or Two?”. The answer was one day at a time. So I got the point. You have to schedule, to make things happen in a given day, not to pile up to where it becomes a significant delay because AT&T was waiting for this one.

I used to go to labs early in the morning, as early as 3:00 in the morning to plan for the day's job before the team members came. I asked them, “okay, this is what we needed to do, individually each day.” And that's how we made things happen.

Vardalas:

So the pressure on you must have been enormous.

Kang:

It was a good pressure. After the success of the project, they gave me a special recognition. On top of that, we gave what is called the LDI, a document, to Western Electric for manufacturing. We handed it over to Western Electric in Allentown, Pennsylvania. Then we found out that they couldn't do it. They said “this is too complex for us to manufacture.” There was a big crisis again. They said, you know, “Steve, this is most important, can you form a special taskforce and go to Western Electric to make it happen.” So I hand-picked a few people to commute to Allentown from Murray Hill every single day for the next several month and worked with the manufacturing people.

Vardalas:

Now, so what year would this be?

Kang:

I think that was 1980, 1981, that timeframe.

Vardalas:

How long did that take - the actual design and the testing, the prototyping and all that?

Kang:

Design including failure took about three years.

Vardalas:

And your role as supervisor was two years?

Kang:

Supervisor, I did formally supervision '82-'85.

Vardalas:

Oh, so even after the project ended, you were still supervising other things there.

Kang:

Yes. We were doing further development on microprocessors and the microprocessor family. We had the Bellmac 32, which was the Western Electric WE32,000 and the next generation was WE32,100 and the WE32,200.

AT&T and CMOS

Vardalas:

Why didn't AT&T - why didn't it succeed in pushing this into the computer? What happened to this technology?

Kang:

AT&T was not really experienced in the microprocessor business. The customer interface. Would anyone buy new microprocessor if it is from a new company - compared to Intel or to Motorola? So that was a huge disconnect. After we had demonstrated the microprocessor chip working at speed, and even beyond speed, what we heard was, “well, the customer wants it this way.” They would say that customer wanted reverse byte ordering. So we went through that change to make reverse byte ordering and making the mask changes. But that was not really what was critical for the customers. It was more trust. As a new company, they wanted to know “can you support our product?”

Vardalas:

Even though it has the AT&T name behind it, it's still a new -

Kang:

We're a newcomer. So AT&T brought in the Italian computer company to try to do it.

Vardalas:

Olivetti?

Kang:

Yes, Olivetti. That attempt was not successful.

Vardalas:

So your group produced the first 32-bit CMOS microprocessor?

Kang:

First CMOS.

Vardalas:

Do you know which company next produced CMOS microprocessors?

Kang:

I think Intel and the Motorola, they followed actually CMOS. You know, they sort of have a quasi CMOS version, but later when the power problem became so difficult, they converted it to CMOS. But they were a few years behind AT&T in that.

Vardalas:

Do you remember the fight for high performance between bipolar circuits and CMOS circuits?

Kang:

See, bipolar at that time was already gone.

Vardalas:

Oh, it was already gone by that time.

Kang:

It was NMOS and CMOS.

Vardalas:

Oh, by that time it was gone.

Kang:

Yeah - problem. It's a vertical device structure problem. You know, you cannot really fabricate VLSI chips using bipolar technology.

Vardalas:

Yeah, but that - by that time, you had the integrated -

Kang:

People had figured out that has to go to a planar technology.

Vardalas:

You brought up an interesting idea that AT&T wasn't well placed to market, to push the microprocessor out the door. They didn't know how to deal with the customers. In general how was AT&T in moving from ideas to commercialization? How did you find it from your experience?

Kang:

At that time AT&T’s philosophy was whatever great scientific thing we can do we will do it and demonstrate it. It was not really linked to customer needs intimately. But when company was broken down, then AT&T faced stiff competition. They did have to survive, right? Now, we are competing in computer markets. So what we did was used the chip in switching machines. So it was not total failure because it still was used in 3B2, 3B5 switching machines in Indian Hill. So they used it for switching applications.

Vardalas:

Oh, switching applications.

Kang:

It was used. Yes, switching was like a computer. They used it. But they didn't really win the market in the desktop computer or that sort of a thing. So it still was a significant project. The other thing they found out was it really needs to be customer driven, not just, “produce some great invention here” or “a great product” you buy. AT&T couldn't survive that way.

University of Illinois, Urbana-Champaign

Vardalas:

Why did you leave Bell Labs?

Kang:

That's another interesting story.

Vardalas:

You seemed to be very successful there.

Kang:

Very successful. The manufacturing success was tremendous. We produced more than they could consume in Indian Hill, in Illinois. Naperville, Illinois we called Indian Hill site of AT&T Bell Labs. So then AT&T got very excited. Western Electric got very excited because with our contributions, they were able to fabricate, produce, very complex chips. Then they said “oh, wouldn’t it be nice for you people to be together with us.” And for the company, it was very important to have a business success, to meet the need of Western Electric as the manufacturing arm. AT&T Bell Labs said now we understand. So your group should be moved to Allentown. The whole department was to move to Allentown, actually. I knew in corporation, you do what company wants you to do, although I had the option to remain in Murray Hill in different department. I knew it was good to move. So I was trying to move to Allentown. Then I knew a lot of friends in academic circles and one of my colleagues, Prof. Tom Trick who was department head at Illinois Urbana-Champaign of electrical and computer engineering department, called me one day suddenly in Allentown. I hadn't yet moved to Allentown, but I was commuting to Allentown. He said “we hear that you are going to move to Allentown, and since you are going to move, would you be willing to move to Illinois?” So that's how they invited me. They said they will give me tenured professorship at the University of Illinois. It's an outstanding department. And they invited me for interview. I went and they offered me a job. After getting the offer I was debating a lot, whether to go to Illinois or remain at AT&T Bell Labs. When I mentioned that I may move to Illinois, AT&T Bell Labs said “you are making the biggest mistake of your life if you leave Bell Labs.” So I switched my decisions many, many times.

Vardalas:

Can you share with me what the plus and the minus side was going on in your mind? What were you thinking about?

Kang:

The biggest hindrance was family. We were living in Bridgewater, New Jersey and my wife’s sister was living just a short distance away. We were living with my father-in-law, and he enjoyed having grandchildren from both families. We got together quite often. That's why they said “please do not move, stay at the Bell Labs. You are doing so well and if you move to Allentown still shorter distance.” And our daughter was a beautiful piano player. She had a piano teacher who moved from Russia and loved our daughter. Our daughter used to get first prizes in many piano competitions. So the piano teacher said that I should send her to Julliard, she was still young, ten years old. But she was so good. So she even said “can you please find another job in New Jersey.” She thought that I was losing my job in New Jersey so that she could continue to teach our daughter. My family was totally against my move to Illinois. So that was a difficult.

Vardalas:

Obviously, you found Illinois intellectually more attractive?

Kang:

Yes. It's academic environment. Now, you know, Bell Lab was now getting into more business oriented activity, more competition, more development, less freedom in that regard. You really have to align what you do to the company, you know, profit business and so on but it was a tremendously good experience for me. But I felt that my chest was congested, you know. I felt that gee, I am not as free as being in academia. That's how I made a decision. In the beginning, everybody was mad at me. Wife, our two children. Father-in-law and my brother-in-law, sister-in-law and their children.

Vardalas:

That must have been difficult for you.

Kang:

Very, very difficult, and then at - this is another interesting story. My father-in-law, he used to be in charge of forestry in Korea. He said “even trees, you do not transplant too many times.” If you think about it, very significant words. Moving means that you're uprooting the family tree and then go to new soil. You will hurt all the roots and it takes time to do that.

Vardalas:

So he didn't help you very much, did he? He wasn't helping you.

Kang:

But fortunately after we moved a few months later, he wanted to come to Illinois.

Vardalas:

The grandchildren. He wanted to see the grandchildren.

Kang:

Grandchildren, because they were young and the other family grandchildren were bigger. They were in high school.

Vardalas:

Illinois is the first of a series of interesting positions you had. How did your work evolve as an engineer in academia? Did your focus and your approach change?

Kang:

Not really, because I gained very valuable experience at AT&T. We were at the forefront of integrated circuit technology.

Vardalas:

Probably more than [the] university had seen.

Kang:

Absolutely. No university would have provided such experience. A tremendous experience. That’s why I think they wanted to recruit me to Illinois as, you know, Illinois is sort of isolated place. It's not in Silicon Valley, like a Stanford or Berkeley.

Vardalas:

Right.

Kang:

But the way they kept abreast of modern technology was to invite industry folks, good people. So it was very exciting. They provided a very good environment for me. I introduced new courses on VLSI design and then at the same time in the university environment, you cannot really do serious practical VLSI development. So I switched my direction into more computer aided design for VLSI.

Vardalas:

Oh, computer.

Kang:

So, you know, it means that computer modeling, design automation, that sort of things.

Vardalas:

'Cause the universities don't have the infrastructure to do actual -

Kang:

Right. So that's how I tuned my research direction, went through the modeling simulation, physical design automation. I introduced some new course, far ahead of many other universities in physical design of a VLSI. That was extremely rewarding.

Vardalas:

Now, you come from industry. Did you try to maintain a strong tie to industry while you were in university?

Kang:

That was enormously helpful. I knew what industry people were talking about. Immediate trust. So I worked closely with Texas Instruments, worked still closely with Bell Labs, IBM, Motorola and Intel Corporation, many, many other companies. They knew I had experience in industry and so they gave a lot of good research contracts and they favored our students. So it was a really good start.

Vardalas:

Do you find that, talking about electrical engineering now, do you find that in universities, it tends to be very theoretical?

Kang:

Not necessarily.

Vardalas:

Is there a bridge between industry and electrical engineering? What is your view about the relationship between the electrical engineering departments and industry in general from what you experienced?

Kang:

Illinois always, as I said, promoted industry interaction because it's sort of isolated. In old days, I think major electronics industry was in Chicago. But in a modern electronics is in integrated circuits, it's all on the West Coast or East Coast - IBM and other companies. So Illinois recognizing the importance of industry link, that's why we took extra effort to link to industry, always bring industry people for seminars and so on. The other important component was the Semiconductor Research Corporation, SRC. It gets support from industry. SRC will solicit proposals and give out contracts to professors.

Vardalas:

So it's like precompetitive sharing, precompetitive ideas.

Kang:

Right. Absolutely.

Vardalas:

Yes, yes.

Kang:

So we had good contracts with SRC which really benefited our industry, and our companies. We used to have review meetings with companies and I used to send the students for internship to the industry and so. It was a good experience.

Department Chair

Vardalas:

You became chair of the department, too, didn't you?

Kang:

Right, so it's a long story. I taught for ten years at Illinois from '85 to '95. The department head who recruited me, Tim Trick said, “I have done enough and I'm stepping down.” They advertised nationally for the next department head. I was nominated internally and did end up getting the department head position. That nomination was based on prior experience. We had an Engineering Research Center from National Science Foundation, Center for Compound Semiconductor Microelectronics. Illinois is very strong in compound semiconductors. That's why NSF gave us that engineering center so that they could develop some technology to benefit the industry, especially with the emphasis on optical interconnect.

Soon after I got there, like, 1987, they had the first review, by the NSF review team. The review said “this is terrible.” We'd like to take engineering center back. It was a big disaster.

Vardalas:

That must have been terrible for you.

Kang:

I was not involved at that time but it was a big crisis for the department. So my department head said “Steve, can you help us with this,” because I had industry experience. So with some systems perspective, how do I help? Well, I suggested we create a test bed because the criticism was that individual researchers did their own things. There was no linkage, no system wide effort. So I proposed that we create a test bed, like an optical interconnected test bed. For that test bed to function properly, you should make components, such as laser diodes, photo detectors and the wave guides, that sort of things. So if you were working on laser diodes, you had to meet a spec and deliver it using your new process technology, new device concept. Likewise, for photodetector or waveguide. NSF loved that idea. So that's how I became associate director of the center in charge of research and industry interaction.

Vardalas:

Ah, but these people working on this project were all university faculty.

Kang:

University - all Illinois faculty members.

Vardalas:

When you're a supervisor of AT&T group, they know it's to their interest to do this, otherwise, the job is on the line, but for academics it is not the same, is it?

Kang:

They wouldn’t get their money. If they don't want to do it, we’ll find somebody else. As long as you want to get the money then you better do it. So that was tremendously helpful. We created a test bed. I worked with a computer science professor. Highly interdisciplinary action, you know, getting a computer science professor with device people to have a common dialogue. Although it was very sluggish in the beginning, it really was a rolw model in the nation, for engineering research centers. So it saved it. For the next ten years, we had that center. Otherwise, it would have been gone away because NSF threatened to take it away. So because of that contribution, they trusted me and they ask me to become the next department head.

Vardalas:

Oh, I see.

Kang:

That's how I became department head.

Vardalas:

Did you find that a gratifying experience, to be department head?

Kang:

It's one of the nation's prominent departments, and outstanding faculty, and good students, good visibility, so I was willing to do it.

Vardalas:

So did you see any - did you see yourself as having any specific challenges as chair? What did you think you had to do as chair? Did you set yourself any goals as chair of that department?

Kang:

You know, how to even enhance the department position in terms of recruiting graduate students, to build up our visibility in terms of our national ranking and hiring good faculty members. Retention of faculty members was also a big challenge. So a lot of good challenges. How do we bring more research centers to the department and that sort of a thing.

Vardalas:

Do you feel looking back that you were successful in that?

Kang:

I think so. At the time, I found out that, while many other universities were having many, many endowed chair professors, we only had three. By the time I left, we had more than ten endowed professorships so I tripled the number of endowed professorships. I more than doubled the scholarships. I thought it was really a good contribution to the department. And it was very gratifying because is one of the most really eminent departments in the nation - in the world, for that matter.

UC Santa Cruz and Engineering Education

Vardalas:

Okay. So let me just follow up, then with some general questions before I go into - let's go to Santa Cruz very quickly 'cause we have about 15 minutes.

Kang:

Right, okay.

Vardalas:

You continued the same line of research when you went to Santa Cruz?

Kang:

When I was in the department head, people said, you know, because with the department almost 100 faculty members. They said now you have to give up your research. Research I couldn’t give up. So I continued my research. I had a really good research group - more than ten graduate students.

Vardalas:

In Santa Cruz.

Kang:

At Illinois, and then I used to get a lot of phone calls to be recruited as a dean at many universities. Before I moved, I had a few different opportunities to become dean at multiple universities. But when I look, Santa Cruz was a new engineering school. They thought Santa Cruz would be more interesting, more exciting place, so that's how I chose to come to Santa Cruz as of January 2001.

Vardalas:

And no discussion in your family about uprooting the trees again?

Kang:

By that time, we were empty nest. Our children were not there. My wife always had thought of California, you know. Illinois was too cold.

Vardalas:

I want to ask some general questions about education. You've obviously had positions of importance in the engineering educational system. Have you given any thought to what you see are the main challenges facing engineering education in the next decade or two? Have you given that much thought and what are your conclusions?

Kang:

I found -

Vardalas:

In the United States, I'm talking about now.

Kang:

Yes, the number of engineering students is decreasing in this country.

Vardalas:

Is that right?

Kang:

Right. Yes. Engineering is not as popular as before. Actually, and this is happening in Asia also. It's much stress they feel that the technology is moving so fast in terms of the career. The career span is rather short. There are a lot of people who get laid off at the age of 50 or 55. Very hard to get back to get a job. So young people choose a different field instead of engineering. Even when I was at AT&T, I interviewed many young people. They said oh, this is too complex. I don't want to do this. It's a very complex job. Yet all the advancement, major advancement that we make for quality of life is through engineering.

Vardalas:

Yes, yes.

Kang:

So there's a tremendous need for engineering. It is a challenge to promote engineering. Even students coming in, how do you retain them because it's a shock to them? That's why I think that hands-on learning experience can be very helpful to retain good engineering students.

Vardalas:

What about the complexity of the curriculum? The amount of specialization. Can any student cope with the amount of material that's being pushed on them all the time? It seems to just - is that a challenge, you think?

Kang:

There is a challenge. I know there are so many course requirements. I think there is a need to really look at the curriculum. See whether we can reduce the number of courses and instead provide more depth in what they learn. Sometimes, when you learn fewer subject areas and build the depth, then you can make link and come up with a new field and that sort of thing. If you tried to teach them too many subjects, too shallowly, I don't think that they really learn well. So there's a sense in reducing number of course requirements and instead, you know, putting more depth.

Vardalas:

Is there any tension between industry needs and preparing engineers and, you know, academic issues? In other words, does industry often want people specialized when they graduate ready to run in an area, keep them narrowly focused at the same time you want to keep them broadly trained. So is there tension there?

Kang:

Yeah, that tension is always there, and then also industry people want young talents out as quickly as possible. They think that if they tap young talent, then company can benefit more, right? So usually executives ask that, “Why don't you graduate your PhDs earlier than holding on? and so on.” But the university also has its goal to delve into more depth and do something very innovative which eventually can benefit.

Vardalas:

There are a lot of other topics I'd like to approach, but given that the time is running out, there's ten minutes left. I'd rather just end it now so we can close it and then maybe another time in the future we can have part two of this interview when you're available. Thank you.

[End of Interview #525 on 17 December 2009. Beginning of Interview #525A on 4 March 2010]

UC Santa Cruz Dean

Vardalas: It [is] 11:00 o'clock, Thursday, the 4th of March. And I'm here with Chancellor Kang, for part two of our interview.

Kang:

Yes. Thank you, welcome.

Vardalas:

And we're sitting in his office at UC Merced. Thank you again for agreeing to do this.

Kang:

Thank you.

Vardalas:

When we last left off, you had started to discuss your experience as UC Santa Cruz's Dean of Engineering. What prompted you to leave the University of Illinois and go to Santa Cruz to become Dean?

Kang:

As the Department Head at Illinois I used to receive many offers, especially from Canada and the United States, inviting me to become a candidate for their Dean positions. And I was interested in moving on to have a bigger influence in engineering education beyond the areas of electrical and computer engineering. Because I always thought that, in nature, there is no boundary in terms of discipline. We can have really good impact through interdisciplinary education. That is why I was interested in moving up to become a Dean - to have a greater influence based on what I know in electrical engineering. While at Illinois I was also on computer science faculty and also I had an interest in power engineering among other areas.

So when I was invited to apply for a Dean position at Santa Cruz, I was also being considered by other major universities for Dean position. And when I was offered the Dean position at Santa Cruz, I asked people what they would recommend? They consistently said wouldn’t Santa Cruz be more interesting and more challenging. Also with my high tech background in microchips and microelectronics industry being right next in Silicon Valley, Santa Cruz was very attractive. While at Illinois I had worked with Intel, AMD, HP and many other companies. So the location motivated me to seriously consider Santa Cruz.

UC Santa Cruz was very young. They had only started engineering school a few years earlier. So I was to be the second Dean of the engineering at the Santa Cruz. They wanted to bring somebody in to build engineering program. So that was the motivation for me to accept invitation to become the second Dean at UC Santa Cruz. I started on January 2nd, 2001.

Vardalas:

Ah, I'm sure climate played a role?

Kang:

Climate, yes, I mean I did my PhD at Berkeley. I left Berkeley in 1975. So in coming back to California climate was a factor, also but being close to Silicon Valley, where I would have lots of affiliations, was really another good motivation. Also the opportunity to build something new and have an impact was another consideration.

Vardalas:

Is it straightforward to build up an engineering school? What was the process like? What did you have to face to do this?

Kang:

It was rather challenging because the campus was not used to engineering. Although they had a head start in computer science, as one department. The founding Dean had started computer engineering as a separate department. So when they started engineering they combined the two. Those two departments were anchor departments for School of Engineering. By then it was pretty much IT-oriented; computer science, computer engineering. Later they had started electrical engineering also. When I moved to Santa Cruz my vision was that if we were to be more effective in the twenty-first century, we had to move beyond IT. I advocated nanotechnology and the biotechnology. My thrust was to build the strength in IT, Bio and Nanotechnology. I had the theme of BIN, B-I-N, Bio, Info, Nanotechnology.

People then said, “We don't have a strength in bio. We don't have a strength in nano.” That's why I was promoting them more because that is what is really needed in the twenty-first century. To become known as the forefront School of Engineering we really needed to develop those areas. That's where I put emphasis on hiring; more people into biotechnology area as well as the nanotechnology area. And one of the areas that Santa Cruz was already known for was bioinformatics, which had come out of computer science. David Haussler is a nationally known figure in bioinformatics. So that was a good basis to start with as we moved more to bimolecular engineering and beyond the bioinformatics. And then I also recruited two key individuals from Cornell and N. C. State to develop the field of bioelectronics. And we worked with the University of Southern California School of Medicine, School of Engineering and with Cal Tech to attract a new NSF Engineering Research Center in what is called the biomimedic microelectronic systems (BMES) engineering, to develop implantable chips to help overcome blindness and other physical and cognitive handicaps.

We were able to attract the engineering research center for biomimetics. And worked with USC and Cal Tech. And that was made possible by hiring two key individuals; Wen-Tai Liu from N. C. State and Michael Isaacson from Cornell University. It was a good start.

Vardalas:

When building the school, were you told, “This is your budget”?

Kang:

No. In fact, you know, I have some pioneer spirit. I do not always believe what is promised will be delivered, Although it's nice to get the promise. I always take the attitude - I'm a team member. I will help build the program and if we do well then it will bring other additional support. I had to do lots of selling in the beginning because I had to confront many other faculty members in different schools to explain why we needed this.

So to give you one example, due to tight resources we were told that we could only hire one faculty in a certain area. But when we advertised, we had several outstanding candidates. So I said “Let's go hire two people in this area.” And I was asked “how would you do that?” I said,“I will divide this into two 50% slots” I was then told, “you are going to lose both candidates.” So I had to figure out how to come up with the other 50% for two individuals and I managed to convince other campus people that we could do it. I told them, “Just give me a chance. I will be accountable if I fail.” So they trusted me and I ended up hiring both of them with that one-person budget. I did have some help also.

Vardalas:

These two positions refer to the two gentlemen that you were talking about?

Kang:

Yes, that's correct. That's correct —

Vardalas:

Did you, as Dean at Santa Cruz, pursue your belief in creating links to industry?

Kang:

Absolutely. Yes, we made many trips to Silicon Valley to build the relationships. And furthermore NASA Ames Park at that time was interested in UC Santa Cruz creating a University Affiliated Research Center, UARC. That was for over ten years, with a $330 million budget. We already had foci on bio, info and nanotechnology which was also the foci of NASA. So we worked out a plan with NASA to create a new UARC. That was a huge uplift for UC Santa Cruz.

Nanotechnology Blue Ribbon Task Force

Vardalas:

Oh okay. Tell me about nanotechnology for a second, before we move on, you - I think if I'm not correct - you sat on a…

Kang:

Yes. Blue Ribbon Task Force.

Vardalas:

Yeah, Blue Ribbon Task Force on Nanotechnology in 2005. Can you tell me something about the work that that Task Force did and what were its conclusions?

Kang:

Yes. So we worked with many industry leaders from Agilent, Hewlett Packard, Intel Corporation; many of the companies in Silicon Valley area. This Blue Ribbon panel was created by Congressman Michael Honda for both California and the nation as a whole actually, to plan for a future industry in this important area. And what would it take to help attract business into upcoming nanotechnology areas including the energy industry, and also application to bioengineering, biotechnology.

Over a one year period we had intensive meetings. The Task Force was divided into sub-groups, on issues like how to prepare the workforce for the future, and education, intellectual property technology transfer, and how to promote better technology transfer from university research labs to industry. How could we attract and retain the nanotech industry in California? Those sorts of things.

Vardalas:

Do you recall the big conclusions, the big recommendations of the Task —

Kang:

Yes, nanotech was an opportunity and was a big thing. Nanotechnology is an important enabler. Universities and the industry should work closer to promote better technology transfer which is a difficult issue that we still need to work on.

Vardalas:

Was IP an issue in any of this?

Kang:

Yes, absolutely. When industry is sponsoring particular research they think that they own 100% but that's not the case. As you know, when a professor works on a particular topic, there are other supporting topics through which they attract the funding from NSF, DARPA, or even other companies. So it is really difficult to get a promise from a professor that a project is 100% company X’s project, and that this company will have all the intellectual property —

Vardalas:

So the assignment of patent rights is an issue then? When you have this kind of relationship between business and academia?

Kang:

Oh yeah.

Vardalas:

— between academia —

Kang:

That's right. It's still an ongoing difficulty. Even today companies tend to demand what many universities cannot agree. Usually negotiations take a long time.

Leaving Santa Cruz, Education Challenges

Vardalas:

So you left the Dean at 2007.

Kang:

Yes.

Vardalas:

How big did the engineering school become?

Kang:

When I got there, we had the faculty size of about less than 40, I think it was at 38. Previously when I was Department Head in Illinois we had a faculty size close to100, in electrical and computer engineering department alone. But by the time I left Santa Cruz, we more than doubled the faculty size, close to 80 faculty members. When I came to become the second Chancellor at UC Merced, our faculty size was about the same, because UC Merced got started in 2005. And when I came in 2007, it was only 2 years after.

Vardalas:

You mentioned [the] interdisciplinary approach to engineering education. What about the other engineering disciplines like mechanical, did you bring some representation into these? Or did you just want to focus on one side of engineering?

Kang:

At Santa Cruz?

Vardalas:

Yeah at Santa Cruz.

Kang:

At Santa Cruz, School of Engineering, we had bioengineering program by the time I left. We had a computer engineering program, computer science program, electrical engineering program. And we also had applied math and statistics. We also had a technology and information management program. And we also started offering a Master's course in Silicon Valley at NASA Ames Park on network engineering.

Vardalas:

You've been involved in academia for a long time. Head of department, Dean and Chancellor. So you've gotten a chance to develop views on education, engineering education. I want to probe you about the challenges that you think are facing the education of engineers in the United States or globally, in the coming decade. What do you think they are? You mentioned last time that the number of engineering students is on [the] decline. Why do you think engineering students are declining?

Kang:

Many students are keenly interested in their future career. After getting an education, they want to know what they will do. What type of career they can develop, what will be their income level and quality of life. For instance, computer science enrollment went down very rapidly when people thought everybody's software job was moving off shore to India and other countries. In fact the computer science had a tough time attracting more students.

So we started to be more innovative at Santa Cruz. We introduced a computer game design track within computer science, which attracted a lot of students. The computer game industry was very appealing to students. It was fun for students to study. In fact the computer game design track was increasing computer science department’s total enrollment. Yet people asked “What do you teach in computer game design?” We had to teach all the important elements of computer science through computer game design. And people were very excited, and many high school students used to call to find out how they could get into the computer game design track. We attracted top talent faculty members into that program. Now Santa Cruz is very well known in that area, working with the game design industry and getting lots of contributions from them.

So I think in general to be effective in engineering education for the twenty-first century, we really have to educate our engineers to be effective in society, as a member of society. It's not an isolated engineering job. It's not about sitting alone in the corner of a room. It's more about teamwork. They have to be able to work on big teams with members from all over the world. If it is an international corporation, you have a branch office in India, China, [and] Europe as well in the U.S. How can one be more effective in working as a team member with people from diverse backgrounds? Someone may have an arts background. Someone may have a chemistry background, a mathematics background. In that regard, it is important that one must not only know one’s field but also be able to communicate effectively and think reflectively with people from different backgrounds.

Vardalas:

Do you feel the current engineering education in the United States, or in the world, really promotes this kind of looking outside the box. The engineering curriculum is already very dense with a lot of specialization. An interdisciplinary approach requires a little bit of stepping back from specialization and looking cross disciplines, right? How do you think that's going to happen?

Kang:

There is always a tension. People think that we are not teaching enough in specific areas. If we want to have students graduate in a given period, we have to really think hard what are the essential things that we need to teach. In my opinion, teaching never stops at college level or at graduate school level. People will have to always learn. So what I think is important in university education is to teach the basics well. Even in a typical course professors can argue that they have so many topics to cover and we have to teach them all. But that may not be the best approach in my opinion. When you try to teach too much, students really do not absorb it very well. But if you trim down to the basics and teach them well, then I think that they can think more creatively in the future.

I believe it’s the same in any particular discipline. If we trim down and give them a strong education, then they will continually learn and develop new things. So that would have been my approach. For instance if you're an engineer, say civil engineer, and you want to build a bridge or work in the dam or highway, you know, they have to also know what the boundary conditions are in terms of policies and what is practical. I have also advocated always that we want our engineers to be in politics as well. If we have people with more engineering backgrounds working in the legislative arms of the government we can have a better policy as a nation to promote a more solid engineering and science. I think it's important for our students to know economics as well as social policy.

Vardalas:

Yes but there is also that tension where companies would like to have people coming out highly specialized in a particular skill that they need. Right?

Kang:

Right. But then the technology changes so rapidly. I think industry will be short-sighted if they just will tap a person with a particular knowledge at that particular time. If this person cannot create a new technology while being employed at a particular company, that company will become a loser and a follower. So you want to hire a person who is strong in the basics and then be able to learn new things and get into new fields.

For interdisciplinary education, I often used as an analogy - the task of filling a room like where we are now with spheres. If you try to fill this room with spheres, each of which, in terms of physical shape you know, is perfect. And one can argue that to create a perfect education, or a perfect academic unit, you need to create a sphere or typical tall silos in the academic world. But when you fill the space like this room with spheres, I did the calculation, 48% of that space is going to be left empty.

That 48% is interdisciplinary space where future Nobel prize winners will come out. More impact can be found in that space. So we really need to educate our engineers to be ready to get into that space.

Vardalas:

But that requires then a discussion with industry on the nature of education. They have a lot of influence. Don't they? Is American industry properly attuned to this need as opposed to the other countries in the world? In thinking more long term? What are your views on the United States versus let's say Europe, Japan, Korea. Companies in these countries appear to take a longer term view. Is American industry being too short-sighted?

Kang:

I think that America has been very, very strong in future strategic planning within corporations. The other day I talked to a leader in the pharma industry. There's a lot of merger and acquisition activities going on, even in the pharmaceutical industry. And when they merge a small company into a bigger corporation, I heard that the productivity goes below the minimum of the two. Isn't that interesting, rather sad?

Vardalas:

Yes.

Kang:

And that's a huge danger. This happens in microelectronics industry and in any other industry. Furthermore they've been saying that basic research can be done in universities. And industry can just apply. That is also short-sighted because we do research with graduate students. If the corporations cannot absorb these graduate students, they will not come into our graduate study program. So we need to be really careful. They come for PhD training in order to have a better career. And if there is no career in industry, we cannot absorb all of them in academic environment.

Vardalas:

If you were to counsel a young engineer ready to graduate, what would you tell him or her [about] the pluses and minuses of working in academia versus working in industry? Putting money aside?

Kang:

Putting money aside. Yes. What helped me in my career as a professor was real industry experience. That helped me tremendously. When I talked to students I could speak with conviction about what is important when they go out, work in industry. And then when I also did the research, especially with research collaboration with many companies, I knew precisely what they were talking about because of my level of experience in industry. That helped me tremendously throughout my academic career.

I never regret going to industry, getting significant experience, and then coming back to academia where my heart was in teaching and conducting research.

But that is not all. There are a lot of people who may not choose to go into industry - to academia directly rather than to industry. I suggest being open minded. Consider both tracks. But if you want to go into academic life, based on my own experience, industry experience helps. I would recommend gaining some significant industry experience before coming back to academia. But on the other hand, there are people who love in-depth theory. Whether they would need industry experience or not that’s for them to choose. For instance if you're purely in communication theory or information theory it may be okay to immediately start academic career.

Creating Interest in Engineering

Vardalas:

The other thing we discussed [was] about getting people interested in science, in math, in engineering. How would you explain to them the value or the benefit of going into engineering?

Kang:

I would just say that engineering is fun. We have done a lot through engineering. Highway construction, bridges, you know, buildings and all these electronic gadgets and internet technology all came from engineers’ contributions. So engineering career can be very rewarding. And I also say that engineering is much about making things and making it work. You really have a significant impact on society and that you can be very proud of. Engineering is more about the synthesis rather than analysis. Of course, to do the synthesis you have to know the analysis side also.

Vardalas:

It's putting things together.

Kang:

Yes, exactly. Whatever you do, even in society, it's putting a lot of different components together and that is good engineering, social engineering. Person who has a good engineering knowledge can do that better. People ask me, “You are an engineer, how can you be a Chancellor?” I explain that I used to design microchips containing hundreds of millions of transistors and what I learned was transistors don't function by themselves. They have to be connected together. And in university environment it's the same. We have X number of individual students, faculty, staff. But we all needed to be connected to function together. And people get it. And based on my own experience in integrated circuits, I put much emphasis on interconnects. Putting things together, really taught me how to become a better organizer. Making sure that the connections are made correctly and solidly. Even when you make integrated circuits, after the initial processing, the chip may be slow. But when you bake it in the hot oven, the speed can go up, because contacts are made better after baking. The connection is the same in the human society. When you do not have a low impedance in interconnect, communication is slow, as a result society will suffer. But if you have a better contact with the low resistance, communication is much faster. And so, as a whole, society can be function much better.

Vardalas:

I will only add that transistors don't have a mind of their own. Isn’t that the challenge of a university? You have a lot of talented people with minds of their own.

Kang:

[Laughter] That's correct. That's correct. But you know, a Memristor can have a mind of its own.

Vardalas:

What do you think is the role and the responsibility of universities to communicate to the population, to teach and inform, about what engineering and technology is all about and how important it is to them?

Kang:

Oh yes, absolutely. In these days, if you want to get any government funding, without the education component — especially with the emphasis on K through 12 education - it will be very difficult. So that's what is emphasized by National Science Foundation and the Department of Energy among others. So personally I'm involved in preschool education. In Merced we have a word; it's called the BEAM, that's a Business Education Alliance for Merced County. And the reason I participate in that is I sometimes compare education processes as a river. Like a Merced River which is flowing out of Yosemite National Park, upstream is the preschool education. If there's good water there, downstream, higher education will benefit from it. In fact it has been shown that when children get a good preschool education they tend to do much better in K through 12 education. When they do better in K through 12, they will do better in the higher education.

We are all connected. As the Chancellor of a major public research university, my responsibility includes making sure that we work with other educational entities to help them educate better. I've been doing this now for more than two years. I've worked with the business leadership to provide motivation and to also promote the importance of a preschool and K through 12 education.

Vardalas:

What about the idea of outreach in the sense that creating — even in adults - an awareness of what engineering and science is about?

Kang:

Absolutely. Compared to medicine, the public has difficulty in understanding engineering. When I was a Dean at the Santa Cruz, I used the example of bioelectronics, especially the case of one of our professors who created a microchip and the medical doctor in USC was implanting this microchip into blind person's eye to help recover eyesight. That is a very clear example of wonderful engineering. And when I explained it to people without engineering background, they really said “Ah-ha, you know, that's a great contribution of engineering.” So I think we need to find the good examples where people can understand well the contribution of engineering.

IEEE Participation

Vardalas:

Let’s discuss your participation in IEEE. When did you first become an IEEE member?

Kang:

I first became IEEE member back in 1972. I was a new graduate student at UC Berkeley. My advisor Professor Leon Chua was president of IEEE Circuits and Systems Society, which was holding a conference in San Francisco that year. He suggested that I become a student member and I did. And I never regretted getting involved. And through IEEE I got to know a lot of good people and develop a lot of good friendship, and my involvement in technical activities forever shaped me and helped me.

Vardalas:

You were an Editor-in-Chief of an IEEE journal —

Kang:

Yes.

Vardalas:

What are you recollections of being an Editor-in-Chief [Chuckling]? What did you learn in that process in terms of your own development?

Kang:

Yes. So back then when you tried to start a new thing, people always have some reservation. “Why do you want a new thing?” they ask. “We already have this and have that.” We knew the importance of systems and the need for this new transaction called Transactions on VLSI Systems. We had other journals with related topics but our theme was really different. How do we integrate all different components to build systems?

The other important thing was to shorten the submission to publication cycle time. It took too long to get a paper published because the review cycle was so long. But technology changes very fast, you cannot do that. Your work may become obsolete by the time the paper is published.

Vardalas:

So how did you manage that?

Kang:

I came up with actually a schedule. I promised in the journal that this is what we would do, from initial submission to publication. In the beginning we were able to do it because we didn't really have many accumulated papers. But more importantly, I demanded from the reviewers, by doing diligent follow-up, why they were not giving us feedback on their reviews. We pushed it hard and it was received very well. As a result people also submitted more papers to Transactions on VLSI Systems. It was a good experience. As a founding Editor-in-Chief, I did it for over two years and then handed it over to the next person —

Vardalas:

Can you tell me how straightforward it was to get the Transactions approved?

Kang:

So within the Society —

Vardalas:

Yes.

Kang:

— we started in the Circuits and Systems Society when we tried to create this new journal. The editor of IEEE Transaction on Computer-Aided Design said “We already have a Journal that covers the systems aspect,” although it was mainly on CAD. And the Solid State Circuits Journal would say we have the Integrated Circuits Journal. But usually to get it published in the Journal of Solid State Circuits one had to demonstrate through chip fabrication that it functioned. Our new journal didn't have that as criteria.

Vardalas:

So you had to demonstrate to the society executive -

Kang:

To the Board of Governors in the Society —

Vardalas:

Yes. Kang:

— and then we'll have to also communicate with other societies like Electron Devices Society, and other societies. But they were rather secondary actually, once the mother society supported it then the others came along.

Vardalas:

And did you have to prove that it would be economically viable? That you can generate enough buyers —

Kang:

Well let's just say investment. In the beginning there's always a dip. And then it will recover. But I think the Transactions on VLSI Systems is doing very well.

Vardalas:

Okay. Okay. Did you ever contemplate taking on other roles in IEEE [as a] volunteer? You know, you did the editorial thing. Did you ever —

Kang:

I still do — yes, I did serve as president of a society.

Vardalas:

Oh. Okay.

Kang:

Yes. So I did that. That was very rewarding. At that time you had the title of president-elect position, president and past president. This is sort of a three year deal. I served in many capacities. I also helped organize many international conferences. In fact I'm co-chairing the IEEE International Symposium on Circuits and Systems that will be held in Incheon, Korea in 2012. I'm Co-General Chair for that conference —

Vardalas:

So as president of the society, right? Did you find your understanding of IEEE changed once you stepped into that role and had to now confront other organizational units? Did you discover new weaknesses or strengths in IEEE? As you know, being a member and being president of a society are two different things —

Kang:

Oh absolutely. And as society president you always pay more attention to membership. Is membership decreasing or increasing? If not increasing, why? What should we do to help promote more membership? What are the needs of the, you know, members in different regions. So as a society president I was able to travel different parts of the world also and interact with many chapters. Also promote more active chapter activities. It was a good experience.

Vardalas:

In an interview with Michael Lightner, he told me that when he was president of his society, that all he knew about was his society and he saw IEEE as someone who was taking his money. But as he rose up, his perception changed. Did you have any perceptions about the value of IEEE when you were a society president versus when you were a member of IEEE? Did you ever — see what I'm getting at?

Kang:

I think that your appreciation increases as you see more, because as a president you also have to interact with other societies, such as the Electron Devices Society.

Vardalas:

And you sat on the Board too didn't you?

Kang:

Yes at TAB.

Vardalas:

Were you surprised at how things worked when you got to that level?

Kang:

It wasn't — not really a great surprise. I knew IEEE was a big organization. I used to think that it had 300,000 members but this morning you mentioned 400,000. So IEEE has grown. And I also know that IEEE is a very diverse organization covering biomedical engineering, and a lot of different areas so -

Engineering Technology in Korea

Vardalas:

I'd like to end this interview with your relationship to Korea in terms of engineering technology. Tell me something about that. What have you been doing in Korea?

Kang:

Yeah. Advisory Council for International Organizations. One of them is Korea-U.S. Science Cooperation Center. I advise people on how to promote more intimate relationship between USA and Korea and help build strong proposals for having technical meetings and exchange. I also advise Seoul National University which is one of the flagship universities in Korea. They also want to promote more globalization. In Korea many universities used to hire their own graduates. It was like inbreeding. But with globalization, they're opening their doors to foreigners. Recently they want to hire more professors from abroad, from the USA, China, Japan, Australia, and Europe.

So I think that as we promote more global education and better education, anywhere in the world, globally I think we become better. The quality of life, in my view, will rise higher.

Some people may be worried about the competition. That's true, but I always say that the role model example of a competition is professors. Professors guide their students to do the best that they can. And what do graduates of a PhD program do? Often they will enter academia. They even compete with their own professors in the same field. And professors should be able to compete with his own graduates. So maybe now you have to put in more effort. [Laughing] And, in a sense, that's a global competition.

Korea now is probably arguably number one in memory technology. That benefits inter-corporation, other system builders, because they get the memory chips, maybe if Korea didn't do it some other country would have done it. The US chose not to compete in the memory market. Most display panels are coming from LG, Samsung. US companies also invested in Korea. Globally, Korea also provides job opportunities. And they also have offices in US. Samsung USA in Silicon Valley and other places. So I think it's cooperation. I remember Motorola in old days when I was at Illinois. They say they have a Motorola-China presence. And they went to China saying that we are a China company. Many corporations, not only in IEEE related fields, but in many other fields, are global companies. As we work together, strategically I think we can have mutual benefits for all people —

Vardalas:

Well thank you so much. I think that we’ve had a lovely interview. And I want to thank you—

Kang:

Thank you very much. I hope it will be enough for you.