Oral-History:B. Jayant Baliga
About B. Jayant Baliga
B. Jayant Baliga was born in Madras, India, in 1948. His father's position as a director for India's largest electronics firm fed his interest in the technology and led to an electrical engineering degree from Indian Institute of Technology, Madras in 1969. Baliga continued his studies at Rensselaer Polytechnic Institute, receiving the Ph.D. in electrical engineering in 1974.
Baliga joined General Electric Company where he invented, developed, and commercialized the insulated gate bipolar transistor (IGBT) and spearheaded commercialization of wideband semiconductor integrated circuits. IGBTs are integrated widely in consumer, industrial, lighting, transportation, medical, renewable energy, and other technologies. Its efficiency has enabled enormous reductions of energy consumption, helping limit carbon dioxide emissions around the world. After fifteen years, he left GE for a faculty position at North Carolina State University, where he led the development of a successful start-up culture based on some of his 120 U.S. patents.
Professor Baliga is a Fellow of the IEEE and Member of the National Academy of Engineering. He has written or edited 19 books and over 500 scientific articles. U.S. President Barack Obama conferred on him the National Medal of Technology and Innovation and the IEEE has recognized him numerous times, culminating with its highest award, the Medal of Honor.
In this interview, Baliga discusses his upbringing, education, and employment in industry and the academy. He recalls his experiences as a high school, undergraduate, and graduate student on two continents. Next, he reviews his fifteen years at General Electric, sharing his experiences with upper management and the corporate innovation process. He speaks at length about his work at NCSU on creating new companies based on his patented inventions, including his leading role in their licensing and commercialization. He also discusses his approaches to solving problems in applied science and engineering, and his surprise at receiving so many professional honors early in his career.
About the Interview
B. JAYANT BALIGA: An Interview conducted by ALEXANDER MAGOUN for the IEEE History Center, 4 February 2022
Interview #873 for the IEEE History Center, The Institute of Electrical and Electronics 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 ieee-history@ieee.org or Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. It should identify the specific passages to be quoted, the anticipated use of the passages, and the user.
It is recommended that this oral history be cited as follows: B. Jayant Baliga, an oral history conducted in 2022 by Alexander B. Magoun, IEEE History Center, Piscataway, NJ USA.
Interview
INTERVIEWEE: Jayant Baliga INTERVIEWER: Alexander Magoun DATE: 4 February 2022 PLACE: Virtual
Family and Electronics in India, 1948-1963
Magoun:
This is Alexander Magoun of the IEEE History Center speaking from his virtual office in Plainsboro, New Jersey with Professor B. Jayant Baliga at his home, I take it, in Raleigh, North Carolina, his virtual office for North Carolina State University [NCSU]. Jay, you’ve obviously been interviewed frequently, and I’m hoping to ask, perhaps, either detailed questions or things you haven’t thought about. If you don’t mind, I want to go back in time to 1948 when you were born in what was then Madras, I take it.
Baliga:
That is correct, yes.
Magoun:
Yes, can you tell me about your parents at that point? It’s a year after Indian independence. Your father is how far along in his career? Brothers, sisters, any other family details you’d like to share?
Baliga:
Sure. At the time of independence, my father [Bantval Vittal Baliga] was already a prominent electronics engineer in India and had risen to the position of Deputy Chief Engineer for All India Radio, which is the radio network for the whole of the Indian subcontinent, and later became also the unit that broadcast the first television in India. At that time, after independence, he was made the Chief Engineer, and so his job was to set up the radio network in India. He traveled extensively to the West to acquire the technology, and of course he was very well known within India itself. Something that should interest you and our listeners was that he was a Founding President of the Institute of Radio Engineers, IRE, [India Section] which later merged with the AIEE to become the IEEE.
Magoun:
Yes.
Baliga:
So, he had a role even in the IEEE in that sense. He brought magazines like the Proceedings of the IEEE, to my house, a rather unusual resource for me as a young man growing up. I read the Proceedings while in high school and was inspired by some of the people, like Lee DeForest and [Edwin Howard] Armstrong, who received the IEEE Medal of Honor.
Magoun:
Ah ha.
Baliga:
It was a very inspiring place to grow up.
Magoun:
So, you were born in Madras, and then he received this promotion to Delhi, the capital, I assume, to. . .
Baliga:
No, actually he was already in Delhi, but my mother wanted to deliver in Madras. That’s why I was born there. They were already living in Delhi, and then in 1956, he was promoted to head up the largest electronics firm, which is owned by the Government of India, called Bharat Electronics [Limited, BEL].
Magoun:
Yes.
Baliga:
And the chairman and managing director of that, so we moved to Bangalore when I was about ten years old.
Magoun:
So north, then south again.
Baliga:
Right, correct.
Magoun:
I’m curious, as a historian, about the sense of national pride and the struggle to develop an electronics industry, educational infrastructure, that your father, obviously, was deeply enmeshed in with his work with IRE. Did he discuss any of that with you, either growing up or later looking back himself?
Baliga:
Not really. He was very busy, and I didn’t see that much of him, to be honest. The only time I remember seeing him is, during my exam times, when he would ask how I was doing and I found it annoying because I wanted to study. I didn’t want to spend the time answering him about how I was doing. But he figured out that I did pretty well.
Magoun:
Did you have any brothers, sisters? Were you firstborn or . . .?
Baliga:
No, I have an older sister who lives in Bangalore right now, and a younger brother who also lives there. I’m the middle child.
Magoun:
I know the feeling. You mentioned at one point that at BEL you used their clean room to build a mesa transistor. I assume this was when you were in high school in the mid-60s, perhaps?
Baliga:
That’s correct. I graduated from high school in 1963 and joined the university in 1964. Before going to the university, I told my father that you have excellent training programs inside Bharat Electronics for young engineers and technicians who are joining, and I would like to join them in getting trained on some of the equipment. That was extraordinarily unusual for a student growing up in high school in India.
Magoun:
Yes.
Baliga:
I actually managed to rotate through various parts of Bharat Electronics, learning how to use a milling machine and then lathe, and then progressing to how to make a mesa transistor, which was the kind of device they were able to make. He actually brought the first semiconductor technology to India by licensing it and that’s why they had that capability.
Magoun:
Was it transferred via Britain or directly from the United States?
Baliga:
It was mostly through Britain, I believe.
Magoun:
Okay. And there’s a clean room.
Baliga:
Yes.
Magoun:
Can you describe a clean room in southern India at that point? Is there anything unusual about it, at the time?
Baliga:
No, it wasn’t really that kind of a clean room that we now associate with what we do now, even at our university. It was very rudimentary. Curiously enough, the clean room that I started working in as a graduate student at RPI was not that different. It was also fairly rudimentary.
Magoun:
When you say rudimentary—
Baliga:
In terms of the air handling and the particulates, those kinds of things.
Magoun:
Did you have to wear a bunny suit or something equivalent?
Baliga:
Yes, we did wear—not a bunny suit, but a lab coat.
Magoun:
Yes.
Baliga:
Bunny suits are not common in those days.
Magoun:
Right. What inspired the wish to build such a transistor?
Baliga:
Well, this was a chance to train in their training program, so I wasn’t particularly sure what I wanted to do at that point, you know? Building transistors wasn’t really something I had aspired to at that point.
Magoun:
Right.
Baliga:
That happened somewhat later when I was at the university.
Magoun:
Okay. Did you have colleagues at either in high school or at university of similar interests, skill levels, intelligence that you spent time with, collaborated with?
High School Competition and Qualifying Exams for University
Baliga:
In high school, it was always competitive and my parents always had high expectations. If I didn’t get the very top rank, they would always say, “Oh, you’re not there yet. You’ve got to work harder. You’ve got to be number one.” That was always something that was taught to me, in a sense, but those high expectations were there, so that, when I finished high school, I did very well. I got the top rank at my high school, but interestingly enough, I got a certificate from the government of India that I was among the top ten students in that exam, which was actually set by Cambridge University. It’s called a Senior Cambridge exam.
Magoun:
Oh.
Baliga:
And I got the certificate saying that “You are eligible for some financial assistance, but your father is too well paid.”
Magoun:
Right.
Baliga:
It was an interesting thing to receive. But at that point, I had the choice of joining a number of good universities in Bangalore, for instance, or aspiring to join the Indian Institute of Technology system. Today, they have become very well-known and very highly regarded. You may remember that even 60 Minutes had a story where, I think it was, I can’t remember her name now [Leslie Stahl]. But she said that it’s harder to get into IIT than MIT and Stanford combined.
Magoun:
I’m sure.
Baliga:
It was a very hard, competitive exam. When I took it, about 250,000 students, high school students, who thought they had a chance took the exam, and the number of people they accepted was 1,000 students among five universities. It was not easy to get in. But I managed to get a rank of about 50, I believe. That gave me easy entrance, and I could choose which one to go to, and I chose to go Madras because it’s closest to Bangalore.
Magoun:
Was there any particular preparations for the exam? Did your high school teachers teach to test? Did you have cram courses or. . .?
Baliga:
No, at the time, there was no such system. My high school and my teachers weren’t even aware that I was going to do this exam. It was something that I chose to do, and it was done after I graduated at the end of 1963 and before the university opened, later in the year in ‘64. At that point, we just did self-study and used the knowledge we had gained in our high school studies to take the exam. As the university’s prestige grew and the desire to join these universities became more and more among the population, they started having all these training centers. So today, there’s a huge enterprise in India where you go for training to prepare.
Magoun:
Right.
IIT Madras and Electrical Engineering for Undergraduates in the 1960s
Baliga:
Yes. It’s a very different picture now than when I was doing it. I should also tell you that, at that point, IIT Madras was graduating its fifth batch of students, so it was very young and did not have a reputation, particularly in the United States.
Magoun:
Oh. Right, right.
Baliga:
[The] IIT that had a big reputation in the United States was IIT Kanpur because it had an affiliation with the United States. IIT Madras was affiliated with the West German government.
Magoun:
Do you know why that connection occurred?
Baliga:
Each of these five universities was set up by support from five different nations. That’s how it was done by the government of India. In our case, in Madras, it was West German.
Magoun:
I see.
Baliga:
As a consequence, I had to learn German in my IIT Madras.
Magoun:
Oh. Did you have German professors at IIT?
Baliga:
There were German professors resident, but they were considered very senior people, and I don’t know how good their English was, so they never taught us, actually, any lectures. But they interacted with a lot of the faculty and young researchers at the university, so it was very good for the university itself to have them there. They also provided a lot of equipment.
Magoun:
Okay. There could have been a possible move to Siemens, perhaps.
Baliga:
Yes, I didn’t explore that option, but I know that some of the lecturers and others did go to Germany for training. In my case, I received the highest grade-point average when I graduated. We call it the valedictorian here, so I got what is called the Highest Rank Award in my engineering batch, but also the highest award in electrical engineering, which was called the Philips India Award. I could have gone to Philips in India, I guess, as a—
Magoun:
Ah. You mentioned that your father instructed you to go into mechanical engineering. Do you know what his rationale was for this? Did he not see a future in electrical or electronic engineering in India or . . .?
Baliga:
No, it’s just at that point, mechanical engineering was considered to be a little bit higher in status than electrical engineering.
Magoun:
Oh.
Baliga:
I, of course, wanted to do electrical engineering because I grew up with electronics, as I explained to you, so it was, at that point, almost in my DNA to want to do electrical engineering. So I had to take an entrance oral exam, very intimidating exam because there are ten professors and you go into this room and you sit on a chair and they’re supposed to ask you tough questions, technical or not. But when I went in, I’ve said this in previous interviews, they already knew about my father because of his prominence. One of them turned to me and said, “Oh, I’m sure your father would like you to be an electrical engineer,” and I said “Yes, thank you very much.” Got electrical engineering and came out and told my father, and he was very disappointed, but subsequently, I think I did well enough that he had no regrets.
Magoun:
I should hope so. Can you describe the pedagogy or the curriculum at that time, what it was like, how it might contrast with today?
Baliga:
The IIT system was well known to be extremely rigorous and high level in terms of the expectations from students, and the teaching was also at a very high level. One thing I quickly learned was that, yes, I graduated with a top rank at my high school, but everybody else also graduated with a top rank in their high school. I concluded very quickly that most of the people there were smarter than me, so I had to work extra hard to do well and keep up with them and try to be the best in class, so it was not easy. It was very demanding. Something else that’s unusual was that we had to take a lot of subjects at the same time, a lot of courses, I guess I should say. On the average, it was like ten different courses you were doing simultaneously, these were one year long, and we would have an exam in one of these every Saturday.
Magoun:
Yes.
Baliga:
And it all counted, so your GPA was based on all of these tests, plus the final exam.
Magoun:
Was there any collaboration, teamwork with your fellow students, or were you all competing individually?
Baliga:
We did teamwork and also competed with each other when it came to these exams, but we shared our notes and we went over some of the issues that we didn’t understand with each other to help each other understand the material and do better. So, yes, I had some students, two or three, who I liked to work with, and we worked together and met fairly often. Similar groups were there among other students working together, so this was based on probably where you grew up and other relationships that you may have had before you came to the university.
Magoun:
Were there any women amongst the student body in electrical engineering?
Baliga:
At that time, there were zero women in the bachelor’s program. Now there’s quite a sizable number of female students. In fact, we get applicants at NCSU, female students from the IITs.
Magoun:
I’m assuming there’s literally no time for a social life with ten courses for a year?
Baliga:
Pretty much, yes. The time you blew off your steam, is that the right word?
Magoun:
Yes.
Baliga:
Every weekend, they would show a movie in an open-air theater. That was a big thing for everybody to gather and go—
Magoun:
The early or mid-phase Bollywood.
Baliga:
Yeah, there were true Bollywood movies and English movies.
Magoun:
Was there any interest in computing or activities in computing at that point?
Baliga:
No, we had very little equipment in terms of computers. The personal computer revolution took place much later, so even when I was in graduate school, we didn’t have personal computers.
Magoun:
Right.
Baliga:
But while I was there, I had to use the IBM 360 and produce a card deck that you inputted and got the results overnight. It was very different kind of an environment then. I had no access to computers at that point at IIT Madras.
Magoun:
It was all slide rules . . .
Baliga:
Yes, slide rules, absolutely. I still have my lovely slide rule. It has a wonderful aroma that I treasure. I don’t know why.
Magoun:
Really.
Baliga:
It’s something, yeah.
Magoun:
What does it smell of?
Baliga:
Oh, I don’t know. It’s just a unique smell. I don’t know how to describe it.
Magoun:
Was it an Indian brand, or British?
Baliga:
It was a German Faber Castel slide rule. Yes, but everything in it was written in English. [Actually all instructions were in German!]
Magoun:
Okay, and were there, amongst these ten courses, did you have labs, or was it all lecture halls?
Baliga:
This is an average of ten. It varied from year to year. I remember one year, I think, it was fourteen.
Magoun:
Phew.
Baliga:
Yes, this was a mix of classroom instruction, mostly, and some labs. One of the great things about the German education system is they believe in practical training, hands-on training, so we had to do much more hands-on lab work than most universities in India. That’s where they, of course, were good at providing us with equipment, so I had to learn how to do all the basic things first, like carpentry and smithy, and I remember having blisters on my hand from chiseling, yes. It was quite rigorous in the Madras heat to go and work in that environment.
Magoun:
There’s not much of a winter season.
Baliga:
It gets chilly in the evenings, not anywhere near freezing, but it’s cold in the evenings in the wintertime. The worst part was the summer heat and the humidity is really pretty bad in Madras. But, yes, we would have one day a week, I think, doing practical training and the rest of the time with classroom instruction. And then, of course, after that, we had our usual electrical engineering labs, building circuits and testing RF devices and stuff like that.
Magoun:
Did you build your own televisions or anything that?
Baliga:
No, no. No.
Magoun:
No?
Baliga:
We didn’t have the equipment and resources to do that. One thing I did do is to suggest to the professor I was assigned to in my final year, I suggested maybe I can get a transmitter tube from BEL with my father’s assistance. He was very excited about that because they can’t get access to such an expensive, valuable transmitter tube. My father was able to donate such a tube and I and three other students, I think, worked on a project to try to make that work.
Magoun:
So this is a three-year or four-year program?
Baliga:
It’s a five-year program at that time.
Magoun:
Five years?
Baliga:
Five years, yes.
Magoun:
Wow.
Baliga:
It is a long program. The final grade-point average is determined by your performance in every year. That’s input from each of the five years. Some of the early years were problematic because we had to take things like English and social studies and stuff like that where they wouldn’t give us very good grades, and it hurt my grade point average. I recall getting only 55 points in my English exam, which is terrible, because that was considered very low for our curriculum.
Magoun:
Oh, my.
Baliga:
You know, at the end of five years, I got what’s called an S average, which is over 80 points on the average, so if you get 55, imagine all the other scores I had to get to pull it up to 80. But it turned out 55 was the highest number of points that the professor gave to the whole class, so I didn’t have to feel so bad.
Magoun:
You ruined the curve.
Baliga:
I ruined the curve, yes. Most people got less than 50, which is a passing grade, so it was really terrible. But you couldn’t dissuade that professor from doing it.
Magoun:
You concluded, I take it, in the spring of 1969?
Baliga:
Yes.
Social Unrest in Madras, 1964-69
Magoun:
Given the ‘60s in the West, I’m curious. Was there any sort of social upheaval going on in India that affected life on campus at that time?
Baliga:
Yes. During my time there, there were a lot of protests going on in the south about the use of the national language, which is Hindi, across the country because it’s a very diverse country with multiple states that were pulled together to form the country. In the south, most people did not speak the national language, Hindi, and in the north, a certain fraction did. If you take the average, I think maybe about 40 percent of the people maybe had some familiarity with Hindi, so they declared that as the national language. What that means is, all the government jobs and all the government programs had to be done in Hindi.
And that was very uncomfortable in the south, so there were big protests. Where it stood out for me was in terms of my going out into the town of Madras from the campus. We had a beautiful, isolated campus on the outskirts of the city but if you wanted to go into the city for any reason, we had to use the local bus transportation. When I first joined, all the signs were in the local language and in English and in Hindi, and they removed all the English and Hindi because of these protests, leaving us in a very sad place in terms of figuring out what was going on. That lasted only for a few months, I think.
Richard Feynman, Semiconductors, and Postgraduate Education Choices
Magoun:
Okay. So, at graduation, what were you thinking about your career, staying in India versus taking your chances abroad?
Baliga:
The reason that I decided to come abroad, to the United States specifically, is because in the final year, I came across the famous Feynman books in physics.
Magoun:
Oh.
Baliga:
You know, the story behind Feynman’s lectures—
Magoun:
Richard Feynman’s lectures?
Baliga:
Yes, Feynman’s lectures [The Feynman Lectures on Physics, 3 vols. (1964)]. These were not prescribed for us, but I found them in a bookstore. I decided I should purchase them and read them, so I would sit and read these books while the other students went to watch the movie every Sunday night, actually. I found it very inspiring to read those books, and I said, “I should do physics,” but with my background, there was no chance I could get a position in physics in the United States, so what’s the next closest thing? It was semiconductors, which involves a lot of physics. That’s of course, an electrical engineering topic, so this is what moved me to work on semiconductors. You asked about why I went to transistors. That was the spark that led to me going to transistors. I credit Feynman for being one of the people who influenced me in my life to choose my career path.
Magoun:
Goodness. What course work had you in semiconductor physics at IIT?
Baliga:
We had one course in the final year on semiconductor physics, and basic transistor physics. But there was no lab to do anything, no semiconductor area at the university to build anything.
Magoun:
This is, in some ways, you would say, it was a theoretical course?
Baliga:
Yes, purely theoretical, yes.
Magoun:
And you’re blending that with Feynman, and now you’re saying you want to go to America to pursue this in practical fashion?
Baliga:
That’s correct. I decided, if I’m going to do this, I should go to the finest universities in America. So I applied to MIT, Stanford, and a few others, the top universities. I thought, I am the valedictorian, and this is one of the best universities in India, I must have a reasonable chance of getting admitted with financial aid. What happened was not one of them would give me financial aid. That’s because IIT Madras was not well known at that time, as I explained before. It was the fifth batch of graduates, and it was actually little more than the fifth, I guess, by that time, but it had not made a mark, you know?
Magoun:
Yes.
Baliga:
Then I decided, okay, if I’m not getting into those universities, I’ll just go to the best university in India, which is in Bangalore, Indian Institute of Science [IISc]. That’s a very venerable university going back more than 100 years, so they had a very good program. I thought maybe I’ll study, do a master’s, and then try to come to the United States. But, before I could join IISc, Professor [Sorab K.] Ghandi at RPI, Rensselaer Polytechnic [Institute], suddenly made me an offer rather late in the year. At that time, there was no internet, so we got a telegram saying, if you’re interested, I can support you, and this was in August of 1969. The university opened in September, so that was an extremely short time to make all the arrangements.
Magoun:
Now had you applied to RPI, or what was the network that made this connection?
Baliga:
What happened there is that my father knew a professor at IISc, and he suggested I talk to him about my interest in going to the United States. I interviewed with him, and he was impressed. Of course, being a valedictorian at IIT Madras is a pretty big deal, at least in India itself, so they were very happy to admit me to IISc. He also was very impressed, so he said, “I can ask Professor Ghandi at RPI if he would like to take you and you should submit your application.” I did, and then he sent me this offer. The story behind that, I learned later, is that I was not his first choice. In fact at that time, the department believed that students from India could not be successful, we would not be able to handle the curriculum, and so they were not admitting any students from India. Unfortunately, one of his Chinese students could not come, one of those he offered the research assistantships to, so at the last moment, he decided to take a chance on me. I guess the rest is history, as they say.
Magoun:
I’m curious. I would have thought that Professor Ghandi would have been aware of the filtering process to get into your university and then to finish at the top of it.
Baliga:
You’re right. That’s probably true, but he had to also convince a lot of colleagues in the department.
Magoun:
Yes.
Baliga:
They were not so enthusiastic, and fortunately, after I came to RPI, I actually found the classes very easy, compared to IIT Madras. I got a straight 4.0 grade-point average. After that, all those doubts were dismissed, and they started admitting a lot of students from India.
Coming to the United States and Rensselaer Polytechnic Institute
Magoun:
So, you flew from India to the United States, take a train up to Troy or . . .?
Baliga:
That was a bit of an intimidating experience. At that time, India was very tight on foreign exchange, so all I was able to get was around, I think eight dollars was what they gave me, and so maybe I had a little more than that, maybe 20 or 30 dollars. Of course, you buy the ticket and everything is covered between India and arrival, so I had to get all the way to Troy, New York, and I had other friends of mine who graduated with me going to SUNY [State University of New York] in Long Island. We traveled together up to JFK [John F. Kennedy International Airport]. Then I took a flight to Troy and I reached Troy at about seven or eight in the evening, and I had only a few bucks in my pocket and nowhere to go. It was very. . . what should I say? Comforting to find a family waiting for me that RPI arranged to meet me and take me to their home and host me for a few days before bringing me to the university.
Magoun:
Oh, my.
Baliga:
Very, very kind and gracious, very appreciative, very much, to this day.
Magoun:
May I ask whether they were Indian American themselves or . . .
Baliga:
No, the host family program was trying to get the foreign students accustomed to the American family, so they would then invite me for Thanksgiving and Christmas and on occasion, and I would go and enjoy those experiences.
Magoun:
I think I may have mentioned in the email, I was curious about the development of an Indian American network of mutual support and I tend to think food is rather important. I’m just sort of curious how one continues to eat dal and enjoy Indian cooking, or not?
Baliga:
Well, at that time, Indian food was not easy to obtain and certainly not at the university, and so I decided to stay in a dorm and all I could get, of course, was the cafeteria food. But it wasn’t easy for me to adjust to be honest, so after about a year, I moved into an apartment and started cooking. I had no knowledge of cooking Indian food, so I tried a little bit.
Magoun:
Your mother didn’t send you recipes?
Baliga:
She’s tried to teach me something, I’m sure, but you couldn’t get most of the ingredients to do it anyway.
Magoun:
Right, right. Did you go back to India in summer?
Baliga:
What is interesting is, even after my wife and I got married and she joined me in 1976, it was hard to get many of these ingredients. [I went back to India after three years when I could afford to buy an airline ticket.]
Magoun:
Now RPI obviously has a relationship with General Electric [GE], which seems to be drawing on the resources of power engineers, not just from India, but coming through RPI. Did you encounter Bimal K. Bose?
General Electric Company and Rensselaer Polytechnic Institute
Baliga:
Let me just first mention the relationship between GE and RPI. Employees of GE would come to RPI and work towards advanced degrees, so they had this very nice arrangement, and I benefitted from that because my Professor Ghandi asked me to work under the tutelage of a Ph.D. student who was an employee of GE. His name was Armand Ferro, and he asked me to set up some equipment for him and do some work to provide the capabilities that he needed to do his Ph.D. work. I did a good job on it, and he was very impressed, which helped me later.
He insisted with GE to give me an interview and hire me when I graduated. So, when I graduated from RPI in ‘74—I did not realize this at that time—it was not a good year to graduate. There was a downturn. Employment was not easy to get. In fact, I recall that no other companies would offer me an interview. Only GE gave me an interview.
I thought I had done a very new, exciting technology called MOCVD [metal-organic chemical vapor deposition], pioneered that, and I wanted to continue to work on it. But at that time, it was so new, people thought it was they didn’t know what to do with it, so they didn’t want to hire me to work on it. Now it’s become the technology to make everything from LEDs to lasers and so it was difficult to give that up. But GE gave me the only job offer, and I had to take it, and it was to work on power devices which were already 25, 30 years old at that time, so most people felt it was a very mature area, not exciting to go and join. But, of course, once I joined GE, I revolutionized all of the things we were doing in power devices.
Magoun:
Some of the context for that would be those oil shocks of ‘73 and ‘79 that presumably put a premium on finding more efficient ways of consuming electricity.
Baliga:
That is correct, but the reason that GE created the group I joined was because of competition from Europe for power distribution systems.
Magoun:
Ah.
Baliga:
They were building power distribution devices, and they felt they were being severely threatened by Asea, Brown Boveri, Siemens, and other companies.
Magoun:
Oh, okay.
Baliga:
But Bimal Bose was a professor at RPI. I was a student at that time, so he was a senior person. But after I joined GE, he joined GE later because the power electronics activity at RPI was not doing very well. He left and came to GE, where we had a very thriving power electronics activity, so we got to become colleagues, and he appreciated all the work I did, and of course I got a lot of prominence at GE because of my invention of the IGBT and becoming Coolidge Fellow, which is a very high rank.
Magoun:
Yes.
Baliga:
The highest ranked position for engineers there and one of the interesting things that happened is we were at dinner together once and he suddenly mentioned that, “Did you know that you were known as the golden boy from India while you were at RPI?” I said, “Nobody told me that. If I had known, I would have asked for more money as an RA [research assistant].” But he said, “You did really well and impressed everybody while you were there.”
Developing M.S. and Ph.D. Thesis Topics in Wideband Semiconductors at RPI
Magoun:
Did GE underwrite your Ph.D., then?
Baliga:
Not at all. That was funded by Professor Ghandi from other sources. [I did get an IBM Fellowship for one year.]
Magoun:
Ah ha.
Baliga:
GE had no interest in this process that I developed at that time.
Magoun:
Oh. You somehow managed to do something your advisor didn’t approve of?
Baliga:
No, no, let me give you a little background there. Most of the time, a professor comes up with an idea and assigns it to a student to do a Ph.D. That’s what I do all the time in my program at NCSU, but at RPI, when I did my master’s work, I actually did something very novel. I came up with the first way to make a planar junction in gallium arsenide. That is pretty revolutionary, because planar junctions are the cornerstone of all integrated circuits and no one had figured out how to do it in gallium arsenide. I managed to do that, so that was a very good master’s thesis.
Then he asked me, “Do you have any ideas to do a Ph.D.?” I had heard a lecture on what’s called band-gap engineering, where you combine semiconductors of different band gaps to get different properties. I thought about how to use that, and I came up with the idea of making gallium-indium-arsenide based, very high-speed transistors, which had not been done before. So I went to him and said, “I’ve done these calculations and analyses, and I think this would be a very promising thing to do.” He said, “Fine, but you have to make the layers to build the transistor. How would you do that?” At that time, there was something called chloride transport to grow the layers. I said we would have to set up this equipment and then do it. And he said, “I don’t have the money to build that equipment, but there’s another student who wants to try something new called MOCVD. Only one person in the world has grown gallium arsenide with this method at Rockwell [International]. We want to copy that, and he’s building a machine, so if you want to make gallium indium arsenide, maybe you can add indium to that machine and you can figure out how to do it. Nobody’s done it before, but maybe you can work it out.” Okay, it was an exciting option. I spent about a year, maybe year and a half with that student–Paul Smith was his name–to build this machine, by hand.
Magoun:
Yes.
Baliga:
It was quite challenging, of course, to put it together, but when I went to the library to learn about how this process works, I learned that you have to mix two ingredients. The first one detonates on exposure to air, and the second one is arsine, which is extremely toxic. I went back to Professor Ghandi, and I said, “Do you know this, we could kill everybody in this building, not only the students, but everyone else?” And he said, “No, just do enough leak check testing, and make sure the equipment is not going to do that, and if you like, I’ll buy you a canary.” Every time we met afterwards, for many, many years, I’d tell him, “You owe me a canary. You haven’t gotten me my canary yet.”
Fortunately, we didn’t blow up the machine, and we survived, but it was not easy to grow these layers. For a year, I was not having any success growing the indium arsenide nor the gallium indium arsenide. No one else also had done it, so there was no literature to lean on to figure out how to do it. I had to investigate all kinds of temperatures and process conditions. Then, finally, I figured it out. Once I got it going, then I got really fantastic layers, and I managed to make every good measurements on them. Then Professor Ghandi said, “You’ve done enough for several Ph.D.’s. You don’t need to make transistors. You need to graduate.” What’s interesting is, about three years after that, he called me at GE and said, “Has anybody asked you about the details about how you were successful?” I said, “No”. He said, “Well, if anybody does, do me a favor and don’t tell them, because there are people competing with me for money. They are telling the funding agencies it’s impossible to do it, and you did it three years ago.” That’s how difficult it was, actually, to grow it.
Magoun:
Were you drawing at all on any of the work at RCA on gallium nitride, gallium arsenide under [Herbert P. Maruska or [Jacques] Pankove or [James J.] Tietjen?
Baliga:
Oh, we read all those papers of course, but the processes they used were very different, as I said.
Magoun:
Okay.
Baliga:
They used chloride transport and also liquid phase epitaxy. We were pioneering something new; and later, after about fifteen years, it was recognized to be a very good process. Big manufacturers like Aixtron [AG], so on, started building equipment for the industry; now it’s very commonplace to have commercial machines for growing these layers.
Innovating the Insulated Gate Bipolar Transistor
Magoun:
If we turn, I don’t know how briefly, to your landmark in the IGBT, apparently it took you two months to make it happen.
Baliga:
Well, it depends on what you mean by make it happen.
Magoun:
Reduce to practice.
Baliga:
No, no. That’s not accurate. The story there is I had been working for about five years on power devices at GE and I was trying various ideas. There was a new idea that was proposed in Japan by Professor [Jun-ichi] Nishizawa, and there is a medal named in his honor by IEEE, so he’s very well-known and recognized. He came up with this idea called the SIT [static induction thyristor] and GE decided that it’s very important to investigate it to make sure that we don’t get blindsided by this idea. So I looked at it and worked on it for about three or four years, one of the very few people, apart from Nishizawa, who actually worked on this.
I was successful in making devices, but the applications people would not use them because they weren’t what’s called normally “off devices.” They needed gate bias to shut them off and that doesn’t work well in a power electronics situation because, when you turn on the equipment, you cannot ensure that the device is off when you get a lot of current flow into the system. It’ll blow up the power electronics, so I was struggling with that. Then I got this idea for the IGBT, but I had not made it. At that time, a vice president of a new division wanted to make adjustable speed drives, which are commonplace now, but were not at that time. He wanted to sell them to Carrier [Corporation] and [The] Trane [Company], [to] make heat pumps for homes and businesses. He was using the existing bipolar transistors, and they were failing and very difficult to control. They needed what are called snubbers to prevent failures, and the whole thing was very bulky, expensive to build, not very reliable.
Selling the IGBT to General Electric Executives, 1980
So he came to us and said, “You are doing a lot of research at the labs, but you need to do something really useful for a product division and come up with an idea that works. Otherwise, I have the ear of the new chairman, Jack Welch, and I’m going to tell him that we don’t need you guys at the research labs if you don’t come up with a good idea,” and he left. We were kind of stunned. We had never been quite addressed that way by anybody.
Magoun:
What year was this?
Baliga:
This was, I think, 1980, early 1980. At that point, I told my management that I had this idea, I already documented it in a patent disclosure, but I have not made it. I said, “Do you think it’s okay if I pitch this?” They said, “Well, we have nothing else, and we don’t want to all be canned, so let’s go ahead.”
I put together a presentation and realized that it’s not good just for adjustable speed drives, but for home appliances, for lighting, many other businesses at GE. When he came back in a month, I pitched this idea to him, and he seemed satisfied. He didn’t say, “Okay, no need to drive to work anymore, you’re considered fired. You’re okay for now.” But what he did, he went straight to Jack Welch and told him about this idea, which, of course, got Jack Welch very excited because it impacts all his businesses across the company. He called the head of the research labs, Roland Schmidt, who was the senior vice president at that time, and told him, “I heard about this idea and I want to know more about it. So next time I come to Schenectady in a couple weeks, I want to hear all about it.”
Well, as things work at GE, I was first-level manager. In between me and Roland Schmidt, there were at least two or three other management levels, so this had not filtered up to him. It was awkward for him to hear about it from the chairman, not a good thing. So, he called me up and said, “Jay, can you come right now to my office and tell me about this idea. Jack Welch is coming. He wants to know about it.” I had to tell my boss, who had to tell his boss, who had to tell his boss. I had to go explain it to Roland Schmidt, and then we put together a presentation, and I guess—I’m reading between the lines—that people said, “This guy has this idea. He’s not even made it, so who should pitch it? Maybe he should pitch it because if it doesn’t go well, he’s the one who will be thrown out of the building.”
So, I was given the chance to pitch it directly to Jack Welch, and fortunately, I was able to convince him. But I had to. I had this other idea that is very important in the story of the IGBT. The idea itself is important, but execution to building a commercial product is equally important. I tell people that it would not have happened other than a place like GE because we had such a strong pull from the application side for this invention. If you do it in a semiconductor company, then you have to find a use for the widget, so you can go shopping around and try to find people who might need it. That’s not easy to do.
That was one, and the other is that I had a facility to build this device in California, where they were making MOSFETs [metal-oxide semiconductor field-effect transistors], and I realized I could build this device directly in their production line. I did not want to build it in the research lab because, if you have a new idea, normally it takes a new process, a new way of building it, because it’s new, obviously. That takes several years to figure out, and then you know it works well. Then you go to the production division and say, “Let’s consider building this,” and they say, “We don’t have the equipment,” so it takes them a year or two to set all that up and put in lots of money. It’s a long process, four or five years, to get a product out.
I didn’t want to do all that. I wanted to go build it straight away, but if you go to a product division and say you want to build something that’s never been made, a research device, they throw you out of the door because they are manufacturing people. They are not interested in research. It’s not their purview.
I would not have been able to do that without the blessing of Jack Welch. Once he approved this, I could go to them, and if they were reluctant to do it, I would just say, “Do you want to argue with Jack Welch?” And they said, “No, no. No thank you. We’re not interested in arguing with the chairman. We’ll do what you say.” So, I had to, at that point, do everything, like design a mask set to build the device, procure the starting material, which is different than what they were using. Make sure that all the parts of the device chip are properly designed to make it functional, then provide all that to the people in the production line, have them run the wafers. That does take time. It took about ten months between my conception of the idea, convincing the management, doing all this work, and getting the product out, and fortunately, the very first product worked.
Otherwise, I think I would have been in deep trouble. You know, once you have the attention of Jack Welch, you have a big spotlight on you, and if you trip up, it can be very dangerous. Fortunately, I did everything perfectly, and it all worked. But see, this, this is the part that is taken for granted sometimes because people say, “Oh, it’s just an invention, just an idea.” But if you don’t do all of these things, it doesn’t become a true valuable product that’s used commercially. That’s not all. The device that is made in the production line is too slow for most applications, like adjustable speed drives.
Magoun:
Ah.
Baliga:
I had to actually figure out a new process to speed it up, which nobody had done before. Without that, it would not have been successful, as well.
Overcoming Reverse Salients in the IGBT
Magoun:
I’m thinking of a number of responses, one of which is the historians of technology have a sense that there are reverse salients in the innovation of new systems, the bottlenecks, that are the really hard part. Otherwise, somebody else would have done it.
Baliga:
Correct.
Magoun:
There seems to be a first one, which is to combine the MOSFET and the BJT.
Baliga:
That’s correct. That’s the inventor part.
Magoun:
Then you’re saying you need to make it a little faster, or significantly faster than it was when you solve that problem.
Baliga:
Right. One of the big problems with the IGBT is you end up with a parasitic thyristor inside the device that can also fail.
Magoun:
Ah ha.
Baliga:
I had to figure out how to prevent that from failing in the very first iteration. Even that was not enough, is what I was pointing out. You had to also be able to speed it up.
Magoun:
Given your inspiration from Feynman and the theoretical grounding you have in solid state physics, how do you understand the knowledge base of theory that you have and its application to engineering to not just applied physics but actual electrical engineering?
Baliga:
Let me answer it this way. It may not be exactly what you had in mind, but one of the strengths of the understanding of physics was my ability to analytically figure out how the device would behave when made. At that time, we did not have these really good computer simulation tools that we now use.
Magoun:
Ah ha.
Baliga:
Now, it’s very routine for us to put in structures and we get very nice outcomes from the simulation tools. But there were no such tools, so I had to do all this by hand, and predict the behavior, and I had colleagues in my own group who were skeptics, who wouldn’t agree with those analyses and calculations. I had to try to overcome them, and there were some people who felt we were going in the wrong direction at that time, but history proved me out to be right. I had a model for the device, which I call the PiN [p-type, intrinsic undoped semiconductor, n-type] MOSFET model, and other people had a different model, which predicted very poor behavior. Once the device came out, and it behaved exactly like I predicted, they had to accept my model.
Receiving Credit for Inventions
Let me mention one more thing which is important historically. Because I convinced Jack Welch of how good this is, he said at that first meeting, “I want to embargo all publication and dissemination of this device because I need to milk it for the GE businesses, and I don’t want my competitors to have it.” That was not easy for me to accept as a scientist because then I won’t get necessarily the credit for having created it. I had to remain silent about it for quite a few years, and so people don’t often credit me for how early I actually did the work and the invention, saying that you didn’t publish it until some later date. That’s part of it, but then, there’s a story behind why it came out. If you like, I can share that with you.
Magoun:
Sure.
Baliga:
What happened was that I was embargoed from talking about it, so I of course did not write any papers, and they would not have released them, if I did. I wrote only internal reports. But suddenly, I see in a conference announcement that the GE Semiconductor Products Division is announcing a new component called the insulated gate transistor, and the author on it is some guy from the sales group in that division. I got really startled and very upset by it, obviously, because this was one of the great ideas that I had come up with and was getting so much attention at GE. Now, all of a sudden, somebody else is taking credit for it.
So I went to my management and said, “What’s going on? How is this possible?” And they said, “This should not be possible.” Somehow, the message didn’t get to that sales manager and he was anxious to get this new product announced because it was being made by the division, so he put it out. They then mollified me by saying, “We’ll make sure that your name is also on that paper.” So it got published with my name first and the other person’s name after me, and at that point they also announced a data sheet, with a product, which I still have to this day, and that was in June of 1983. That was actually the first IGBT product announced by a company and sold by the company.
Why I’m mentioning this is that it took another two to three years for the Japanese companies to start building product in ‘85, ’86. The reason they started building it is because they would visit the GE labs and they would be given tours of the research lab. In the presentations, they would be told about this device because it was one of the new, hot things that GE had created. They [Japanese companies] told me later that hearing this is what inspired them to start building the product in Japan. What I don’t like is recently they have started writing articles saying the IGBT was born in Japan.
Magoun:
Oh.
Baliga:
That’s very hurtful because it is ignoring that true history, what has happened.
Magoun:
I see the same thing with blue LEDs.
Baliga:
It hurts me deeply to see this happening because it is not necessary. There is enough documentation to show that’s not true. They have been given plenty of credit for creating the products that they have. I do not deny them. They made the whole technology successful by building the products and scaling it up and so on, and I say that all the time. They don’t have to claim that it was born in Japan. I think that’s a fallacy. I fortunately have been recognized for my work, as you well know, including being inducted into the [National] Inventor’s Hall of Fame as the sole inventor of the IGBT. Now I feel vindicated, it doesn’t hurt quite as much.
Magoun:
It’s a continuing struggle to maintain priority.
Baliga:
Yes. There are other stories, too, of this nature, of people claiming credit. I like to quote Arthur C. Clarke. I think you’ve seen that quote which is that, when you come up with something, people first say it’s impossible, that it won’t work, and when you show that it works, they’ll say, oh, we already have similar stuff, and it’s good enough. Then, when it gets really good, they say, oh, yeah, it was my idea after all. I’ve seen all of that.
Magoun:
To return to your problem-solving abilities, your unique problem-solving abilities, is there something in terms of resolving these bottlenecks of invention where other people didn’t? Do you see something in your own background that accounts for a particular creativity or a different approach or perspective to problems than most of the people around you?
Baliga:
Well, I think what you say is true because of what I have seen, not necessarily to be boastful about my special abilities, I don’t know what makes it unique in my case. But let me put it this way: a lot of people come up with great ideas. There’s no dearth of good ideas. I think my critical special ability is to be able to make the device very simple in construction so that it can become a product. You can come up with a very complex idea and it’ll be good, but it’ll be so hard to build that it never sees the light of commercialization. I’ve done this now consistently with many, many products. It’s not just a one-time thing.
Magoun:
Right.
Baliga:
For example, I invented a diode, called the JBS [junction barrier Schottky] diode in GE for silicon, and it was kind of okay. It worked well, it didn’t get tremendous commercialization interest, but it has become the essential structure needed to make all the silicon carbide diodes that we make and sell today.
Magoun:
Okay.
Baliga:
There is a market now, about $200 million of those diodes being sold, JBS diodes. Again, that’s a very simple structure, very easy to make, and it changes the physics tremendously. The same is true for another device I created called the MPS [merged pin Schottky] diode. A lot of times, when I propose these ideas, people knock them down, saying it’s terrible, it doesn’t work, this and that. Eventually they figure out that, yes, it is actually a good idea. This has happened repeatedly, and the fact that it’s simple to manufacture is a key part that allows it to go from just being an interesting thing you write a paper on and becoming actually valuable to the electronics community.
Briefing Jack Welch, General Electric CEO
Magoun:
When you briefed Welch, could you describe the room, the overhead projectors, your impressions of the chemical engineer turned most powerful man in America at the time?
Baliga:
I think he was not wearing his chemical engineering hat at all. He was there purely as the chairman of this company, and he wanted to create a whole new culture. He had already said that his goal is to make sure every division in GE is number one or two in its area, otherwise he’s going to sell it and acquire other divisions, so there was a threat to all the other managers. He also told the GE research labs that, yes, you’re still going to get the money from the corporate funds, but you’re not going to be given those funds directly. I’m going to allocate these funds to the divisions, and you have to go and sell your worth to each of them and get that money back to fund you.
In our case, fortunately, we had these ideas like the IGBT and other ideas that we were able to do this convincingly. Some other groups struggled to make this happen, but one of the aspects of Jack Welch is he had this reputation of being Neutron Jack, which means that he would go to review a division, and if things didn’t go well, everybody was fired and only the building was left, like a neutron bomb went off. So he was known as Neutron Jack. Everybody was very intimidated and terrified of him; this is why I had to pitch the IGBT to him. The normal practice was for the senior group of people at the labs to do the presentation to him. You know? They didn’t want to have the lower level of inexperienced people having to interact with him, so I had to get in that room, and a lot of people were concerned that I would not be able to do a good job and be successful, but I was successful, and of course, the outcome of that was a lot of pressure to make it work right.
Magoun:
Could I ask how you structured your presentation? Was it, here is a need, here is an approach, or was there something else that you went through.
Baliga:
It was all driven by the impact on the company.
Magoun:
Okay.
Baliga:
We had broken it up into, this is what this device will do for lighting. This is what it’s going to do for your consumer business, for appliances. This is what it’s going to do for adjustable speed drives. This is what is going to do to create what are called numerical controls, and that actually created the Genius I/O product that made GE number one or two in that business from nothing.
Magoun:
Oh.
Baliga:
GE got the order for the Saturn motor car plant actually because of it, so it was a big deal for GE. That was later, of course, after we made it successful, but we told him that these are the things that this device can enable. He was not interested in the physics and all that stuff.
Magoun:
Right.
Baliga:
It was just how it was possible to do it.
Magoun:
Do you still have that presentation?
Baliga:
Yes. It was all done with overheads, just the sheets of the slides.
Licensing the IGBT to Japanese Corporations
Magoun:
You talked about the Japanese visitors. I take it these are the licensing companies from Mitsubishi, et cetera.
Baliga:
Well, I think they would do this to companies whether they licensed or not. They would always bring people around to impress them, that they were doing stuff like CAT scanners or IGBTs or something else interesting, you know? They would go around and see the research center, so I don’t think they were necessarily licensees.
Magoun:
Okay. Did you go to Japan to enable the technology transfer?
Baliga:
No. The way that happened is we did not want other people to build this. But GE semiconductor was not doing a great job in building the device itself. They were doing it well enough, but not pushing it as hard as it were possible. I even volunteered to leave Schenectady and join their division and run it, but they refused to let me do that. They didn’t want me to go.
Magoun:
That would have meant moving to California.
Baliga:
That would have meant moving to not California, but another part of New York State, near Syracuse. Independently, the Japanese realized how good this device was, and they realized it could be made easily, just like I had recognized, so they started vigorously developing it. Once they got products out and those products were more robust and lower in cost because they were able to produce it better in their factory, then GE also started purchasing them from these sites.
Magoun:
I guess I would have thought there was a suitable wall around the patent that had sufficiently broad claims, or is it?
Baliga:
GE did some cross licensing.
Magoun:
Hmm. You started in ‘74. This is going into six to eight years later.
Baliga:
Yes.
Becoming a Manager at General Electric Company
Magoun:
I read at one point you were supervising 40 researchers. When did you end up not reporting to somebody else, but other people reporting to you?
Baliga:
Well, I was always reporting to someone else. There was another manager above me. But I was an individual contributor until about 1978, I think. At that point, they twisted my arm to become manager of this group. I was already helping a lot of the junior engineers and the more recent hires. You know, we were hiring people, at that point, from universities. They had no knowledge of power devices, so I had to train them anyway. They realized I had good skills in managing people, making them productive, so they said, “We have this vacancy, and we want you to head that group up.” I said “Okay,” and it was not my first choice. At that point, I was still deeply involved doing the technical work as well.
Magoun:
So, your mid-1980s GE is thriving. You presumably are thriving as a group manager, a group head, or are you potentially on a management track that you don’t look forward to?
Baliga:
GE had this wonderful belief in training people who are managers, so I was being sent to Crotonville [New York], which is their training center: learning skills on how to manage better, how to manage my time, how to manage people, everything. I was being given the grounding, so that’s why I felt bold enough to say, “I’ll go and run that division” at that time. It was like 1982, ‘83, but it didn’t happen. At that point, I was still deeply interested in doing technical work, so I continued doing that, and of course, they gave me this Coolidge Award, which comes with one year to do whatever you want and they pay all the expenses. I had a tough choice, decision to make, which is to take that fellowship at that time, or wait for maybe fifteen, sixteen years because I had a son who was three years old. Once he got into the school system, it would have been hard to go away for a year. I would have wanted to go to Europe and so on. My wife and I talked it over, and I told her, I had talked with other Coolidge Fellows, but they’re all in their late 50s or 60s. They said they had to wait till their kids graduated from high school, and then they went. And we are talking about our son getting into an elementary school, so what do you want to do? Is it okay if we go now? She said, “It may be hard on our son, but I think it’ll be okay. He’s young. He’s resilient.” My management wasn’t necessarily thrilled, but I told them I can still work on whatever help they needed while I was doing my Coolidge year. So I took my fellowship in 1984. I spent six months in California and six months in Europe.
A Year as Coolidge Fellow, 1984
Magoun:
What did you do in each of those locations?
Baliga:
I wanted to write a book. I decided, that’s the best time, so in twelve months, I actually wrote not only one book, but two books. One was a complete book called Modern Power Devices (1987), which covered these new devices for the first time, and then the other one was on growth of silicon layers with epitaxial growth techniques, a very different type of book [Epitazial Silicon Technology (1986)]. I didn’t tell you this before, but while doing all this in the ‘70s for the power devices, I decided to try to grow silicon layers out of a liquid phase that nobody had done, and I’m one of, I think, the only one, who has done this and written papers on it, so I have like a dozen papers on that technology.
Magoun:
Goodness.
Baliga:
And I wrote this book, I have a chapter in it about liquid phase epitaxy. So, that book I wrote in collaboration with four other people from various other places, Bell Labs and other places, contributing chapters, but the book on power devices was a very intense amount of work while I was on that fellowship.
Magoun:
What made it intense?
Baliga:
Well, it takes a lot of time to pull it together, to write the material, derive equations and all the graphs and so on. It’s a time-consuming affair. I went to California because I wanted to get away from the weather in Schenectady, and we had never lived in California. I set up residence in Redwood City near Palo Alto, and gave, with my fellowship money, some money to Stanford University, to Professor [Jim] Plummer, to support a student, and I also encouraged him to do some work on power devices. So I had a little interaction there, as well.
Magoun:
At this point, you’re beginning to use personal computers to do your writing, composition?
Baliga:
You’re absolutely right. This is when I got an IBM PC, the first one, XT, I think it was called, and I learned how to use that. I learned how to use a word processing software and so on, but my typing was very slow, so I wrote the book by hand and had a lady that put it into the computer and give it back to me. Then I edited it to get it all correct, so that’s how that worked. Then I told my wife after the six months, we’d go to Europe. I wanted to see various parts of Europe, so I arranged to be in France, in Toulouse, for three months, and in Leuven, Belgium, for two or three months.
Magoun:
Who was in Toulouse?
Baliga:
It’s just that I wanted to be in southern France, so that’s my reason. There was a group [LAAS-CNRS] that was just starting to work on power devices. That’s why they were happy to see me there, and I gave some lectures and helped them get started. In Leuven, there is a very big center now called IMEC, very famous. They were not that well known. They were just being founded by Professor [Roger Van] Overstraeten. He was very happy to have me join them and give some lectures and so on. During that part is when I did my second book, the silicon epitaxy book.
Magoun:
What are the restrictions on what you can say based on what’s proprietary at GE versus what you know scientifically or technically?
Baliga:
Oh, absolutely, all the time. After this release of the idea by the Semiconductor Products Division that I told you about, I said “You should allow me to write a technical article about it, right?” And they said, “Okay, yes. Now, it does make sense. The idea is out. You can talk about the physics, if you like, but one of the biggest problems with the device is this latch-up problem. We don’t want to disclose that you have solved it, so you have to cut off the characteristics at this current level. You can’t show that it works beyond that.” This was incrementally published, and it was slow because at that point, it required approval not only by the research lab, management, it went to all these product divisions: the semiconductors, the people making the adjustable speed drives, appliances—everybody had to sign off on anything I wrote. It was a tedious process.
Leaving General Electric
Magoun:
Oh. One would think you were sitting pretty, and what would then make you actually decide to leave for academia?
Baliga:
That was when Jack Welch made a pronouncement. He said [GE] “Semiconductor sucks”. Am I allowed to say that on TV? That’s because he had invested a large amount of money on integrated circuits, VLSI [very large scale integration].
Magoun:
OH.
Baliga:
In fact, there’s a huge semiconductor plant he built here in North Carolina for a billion dollars. They were not productive, and he was very upset, and semiconductors was one big group to him. Power was part of it, and we were very successful. We had done the IGBT. We had good return on investment, good margins. But he said, “No, everything goes.” I had to ask myself, what do I do now? That’s my core expertise, my core research focus. Of course, I was already Coolidge Fellow at that point, and GE said, “We’ll just promote you to the next management level where you can run a lab and you can have, instead of 40 people, 120 or 150 people.”
Creating a Research Center and Patent Culture at North Carolina State University, 1988-1998
And so I had to make a choice. Do I stay technical, or do I move into management, and then somehow it got out that I was interested in leaving, and I started getting offers from all kinds of companies to come and work for them. I interviewed with a few, and then I decided maybe university is a good place because there’s no activity on power devices at any American university. It’s not a good thing for our enterprise.
I happened to know Professor Nino Masnari at NCSU, so I approached him. He was department head, and fortunately for me, he expressed interest. One of the motivations for me was that the state of North Carolina had set up a clean room called MCNC, Microelectronics Center of North Carolina. When I interviewed, I told him, I would do this if they would allow me to access that facility and make my devices. I cannot make them in a university clean room. It’s not sufficiently high in capability because, whatever I do, if it’s not as good as what industry is doing, it’s not going to be good research and people are not going to support it or appreciate the outcome. I actually spent half a day, at the MCNC Center, talking to the president and pitching him my vision of what I wanted to do. They had been set up by the state to work on VLSI, not on power, so it was not in his charter. His response was, “We can give you access, but you will have to pay for everything. I cannot use the state funds that are being allocated for other work.” I said, “Okay. I can bring the funds.” That’s why I set up the center, PSRC [Power Semiconductor Research Center], paying the money to actually start using MCNC.
Magoun:
Were they connected with GE’s VLSI? Was that a state-industry partnership?
Baliga:
Yes, GE was right next door to them, that big plant I just mentioned. In fact, MCNC was set up to attract people like GE to come to the Research Triangle Park, and so were their big success stories.
Magoun:
So, you moved down to North Carolina, and GE presumably sells its site to Intersil or Harris or somebody.
Baliga:
Right, right.
Magoun:
You now arrive in North Carolina, which is looking somewhat entrepreneurially at its university expertise? Are they looking for you to start patenting, getting licenses, making some money for the state--
Baliga:
No, no, no. Actually, patenting was not at all in their horizon, was not a . . . what should I say? A culture. At GE, of course, patenting was a big deal. When I left GE, I think I had 300 patent disclosures, of which they filed a fraction and I got about 60 patents. When I arrived, I had other ideas, so I went to the department head and said, “I have this disclosure. What do I do with it?” He said, “Oh, we don’t know what to do with these things. We just publish the stuff. We don’t like disclosures.” I found out that there was no mechanism, they had no funds, no mechanisms to file for patents. There was no system there, so when I created my own research center, I said “I need to create an ability to start filing patents and cover this area.” The other reason I started the center was because my ideas would be useful to a whole range of companies. I didn’t want to get restricted by funding from one company because then they get ownership of the technology though the funding, so it’s better for me, I felt, if I had funding from a variety of sources.
I went to the university and said, “This is in my vision, this is how I want to do this, but what is your thinking?” Now, at that time, I think NC State was also rather narrow in focus, with just North Carolina as a mission, so when I said I want to do this as a national center, there were some eyebrows raised. They said, “Okay. If you think you can do it, that’s fine.” Then I said, “I want to put together an agreement. Give me a lawyer so I can write in an agreement that I think will work, and I want to put in some intellectual property clauses.” and they said, “Nah, we don’t have resources like that. Just write it, go talk to the other center directors and take whatever they have and put something together.” “Okay, but there were no intellectual property clauses,” so I put all that together, and I sent it to them, and they said, “Yes, there is one problem.” I asked, “What’s wrong?” They replied, “Oh, there’s a spelling mistake over here. Otherwise, it’s fine.”
Okay. Maybe they should give me a law degree, as well! But anyway, I took that agreement, and I went around the nation trying to sell it to all the big companies in the U.S. and I found that my entry point was always at the lowest rung, the first-level manager, and they don’t have much of a budget, so they would keep saying, “Ah, I can give you a small amount of money. I can give you $10K, maybe.” Well, to do semiconductors costs a lot of money. At MCNC, they told me to run one lot, you’ve going to have to pay $65K, so with $10K, what am I going to do? I can’t even get anything done. You have to have serious money.
So in my agreement, I said, I’m going to start with a senior level which pays $150K, next level pays $100K and the lowest tier pays $25K. The university was shocked. You see, most of the companies were paying only like $10K to join centers. They said, “This guy is not going to succeed, but it doesn’t matter. Let him go and try.” So they just let me loose, and they were right. I couldn’t get the U.S. industry to support me. I then went back and said, “It can’t be a national center. I have to go international. I have to go to Japan. I have to go to Europe and see if I can sell it over there.” They said, “Well, nobody has done that, but again, okay, you want to put in the effort, it’s up to you. You can go and try it.”
It was like night and day. In Japan, I was called a sensei. I was a respected person who created the IGBT and had written all these papers and books. They took me to the very top of the company, the president, the board. I talked to them about my vision. They said, yes, this is wonderful. You are inviting us. We’d be very happy to come out and support you. They all joined at that high level. No other way I would have had a chance to make this work. That’s what made all the difference.
Magoun:
It sounds very reminiscent of Nicholas Negroponte and MIT’s Media Lab.
Baliga:
Good.
Magoun:
And close to about the same time of entrepreneurial professors.
Baliga:
Is that right? Okay. Yeah, once I got money from them, then all the U.S. companies came and joined at the lowest levels—and hired all my students, by the way.
From Patents to Start-ups at North Carolina State University, 1998-2008
Magoun:
Ah, excellent. What would you regard as your next high point or challenge at NCSU?
Baliga:
The next challenge was towards the end of the 1990s. There was a recession in Japan, and so they had to pull back on funds. You might know that they had a pretty bad recession there. It’s taken them a long time to recover, as well. When they started pulling funds, I had to rethink what is it that I should be doing. Most of my money was coming from those companies, because of these new intellectual property clauses that I had put in, I had been able to convince all these companies to support many patent applications.
My attorney tells me I’m a very unusual client who writes these patent disclosures. He says I’m his only client—in his company with, I don’t know, maybe 40 patent attorneys—I’m the only client who has a 100 percent batting average with the [U.S.] Patent Office. So all these applications started getting issued as patents. I became one of the most prolific patent getters at NC State. Every year, they would have a ceremony where they would have a lunch and they would hand out plaques to those who got patents, and there was a joke that Baliga has to bring a wheelbarrow every year to take away his plaques. In fact, it got so bad, they even asked me to speak at one of these luncheons, and I gave a speech on the seven habits of highly effective patenting. It’s a play on that book, as you might know.
Magoun:
Sure.
Baliga:
Yes, and that went very well. After that, I heard from my patent attorney that the firm wants a copy of my speech to look at—the other lawyers wanted to hear it.
Magoun:
Absolutely.
Baliga:
The bottom line is, I had all these patents. Then NCSU wanted to take foundation money and invest in creating start-up companies. That was something new for NCSU to get a better return on investment. They hired a bunch of VCs [venture capitalists] to take the foundation money and invest it. They told them, it has to be an idea, created at NCSU by a professor, or student. They looked around, to see who has patents, and I was way up. Most faculty had one or two patents. I had 40, so they said, “We better talk to this guy.”
They came to me and said, “Would you like to start a company?” I had thriving research going on and although PSRC funds were going down, I acquired a lot of money from the government funds: DOD [Department of Defense], DOE [Department of Energy], and so on. I was very busy, so I said, “This is not easy for me to do. I don’t want to leave the university, so I’d like to do a company where we license the idea and I can help the company to design and build the product, and let’s see how it goes.” They said, “If That’s the best we can do at this point. Let’s do it.” That was done through a company called Micro-Ohm. And we licensed the TMBS [Trench MOS Barrier Schottky] rectifier to a company called General Semiconductor. Anyway, it eventually went to Vishay Siliconix, and eventually, they built this product very successfully and announced that it was the most successful new rectifier product in the last 25 years. It was a huge product for them; that’s not something I get much recognition for, but it is one of my really good inventions.
Magoun:
Oh.
Baliga:
Then, the VCs came back and said, “For a venture capital investment, getting licensing revenue is not our best outcome. We want to start a company that we can build and go IPO [initial public offering] and that kind of thing. Do you have another idea?” I looked around, and I said, “Where is there a weakness in semiconductor technology?” I discovered that it’s in the radio frequency space, not in the space that I was actually researching in my center, and this was in cellular phones. When we make our calls, the calls go through a cell phone tower, as we all see on the highways. In the towers, they combine our cell phone calls, and it goes through a much bigger power amplifier, so each of our phones is very lower in power. In the cell phone base tower seen on highways, we’re talking about hundreds or even a thousand watts of RF [radio frequency] power.
I found out that only one company is dominating in this market. They have 80 percent of the market, and their technology is called silicon LDMOS [laterally-diffused metal-oxide semiconductor], and it has a basic physics non-linearity. It’s built into the physics of the device. This is physics we have been teaching for 50 years. It’s nothing surprising, but because of that non-linearity, you get distortion of the signal, and if you have two signals, you also get intermixing of the signals. This produced a sub-band and the sub-bands get spilled over into your competitor’s band. Sprint may have licensed a certain spectrum, AT&T another spectrum, and now Sprint’s signal will go into AT&T or AT&T into Sprint. I don’t want to say that’s happening. I’m just saying, theoretically, it will happen, and that’s a no-no. It’s not allowed by the FCC. They have to build what is called an error correction system; they spent $1,000 for the amplifier and they were spending $2,000 to remove this error signal. It’s a terrible 2G approach and is also low in power efficiency, also it’s terrible.
I said, “There’s a lot of pain in this field and if I can solve it, then I may have a good opportunity.” I had to then invent a new kind of transistor that doesn’t have this kind of non-linearity. That was something amazing because you don’t get to create new physics in a transistor very often, and I said, “Oh yeah, but somebody must have thought of this in the last 50 years.” Couldn’t find it anywhere published. I have now received all the patents on it, so I know nobody seemed to come up with it. I said, “Now, that’s a great idea,” so then I started analyzing how to make this transistor. I found that I can build it in my approach with eight or nine masks, compared to fourteen masks used by the existing company, which means, hey, it could be a heck of a lot cheaper to build.
But I need to come up with a new way of packaging it, as well. I came up with that, and I have patents on that, and then I finally said, “Now it’s time to sell this idea.” And I told the VCs, “This is a really good idea. It solves a pain for a customer. That’s why they’ll be anxious to get this device, there is no device like this in the industry, so we have an opportunity to start a company to do this, but I don’t have a way to do build this, I said no place to build it. So, I need to partner with somebody.” They said, “Okay, maybe we can get a partner to also put some money up front. Where would you like to go to?” I said, “Let’s go to Fairchild Semiconductor. They don’t have such an RF product, and I’m sure they would like to have a market in this area.” So, I gave one pitch for an hour to the senior vice president there who was supporting my research at NC State, and within a month, we had a deal worked out where they gave us $10 million based on my one presentation. That’s not bad.
Of course, then I had to scramble to try to hire people to execute this thing because I had no employees. I had to do all the work myself, which means I couldn’t continue everything I was doing at NC State. I had to go to NC State and tell them I have to take a leave and stop my research center and so on, and so that was also complicated, but I had to do that. Then I told Fairchild, I need a fab. So they gave me access to their Salt Lake City fab and ten people there. I worked with them and we had first-pass success in making the devices with my design. That was really important. We then put them in packages, and we tested them in our lab. We had, by then, hired people in my company, Silicon Wireless Corporation, to do those kinds of measurements, bought the equipment to do it, and so on. Then we started testing our devices at companies like Nokia, AT&T, et cetera, that build base stations, and they all said “Yes. This doesn’t need the error correction network. It’s amazing. And we’d love this kind of RF transistor.” So, again, a huge success in a very short time frame. Mostly, nobody gives me credit for this work.
Magoun:
What years?
Baliga:
This was about 2001 through 2003, 2004. We eventually gave this technology to Fairchild. It actually showed up in their annual reports and stuff. That was the second start-up, which I called Silicon Wireless.
Magoun:
Two good words.
Baliga:
Once this got off the floor, we needed something else to keep the company afloat, so I invented a chip set to power microprocessors, a completely new idea and approach. This one, we went on our own, we found a foundry in Minnesota, called PolarFab. I guess it’s cold enough in Minnesota that they called it PolarFab. I had to go there and show them how I architectured this new idea and how to make it. You have to convince them they can make it easily, number one; and number two, you have a business plan to make enough wafers; it is worth their time, right? We had both of them because we are a huge market for microprocessor power delivery for all the PCs. Then the second was, I had to create a chip and process architecture they could build. I gave them one pitch and they said, “This is like falling off a log. We can do it for you.”
And they did it. Within three or four months I had product, which is amazing for a completely new idea. We made VRM [voltage regulator module] boards to power microprocessors and showed the efficiency can be improved by a significant amount compared to the existing transistors made by all the companies. Hitachi, Mitsubishi, Motorola, everybody who was building these transistors.
We said, “Let’s benchmark our technology with Intel.” We went to them and said, “We have this amazing technology.” I had to fly over to the West Coast and pitch them all this stuff. I guess I had a reputation, so they let me in the door to do it, but after that it had to sell on its own merits. They said, “Okay, give us the devices for our evaluation.” What I told my guys is, we have to package it in the same package our competitors are selling in. Then, they can swap them out on the board and show the performance improvement with least effort by customers. We did that. We had to go all the way to Malaysia to package them and bring them back but we did it. Intel said it’ll take about two weeks to test this. I said, fine. We’ve already done these tests. It looked good. We heard nothing back from them, two weeks, three weeks, four weeks. We got really worried. We thought this is a disaster. Eventually, after five weeks, they came back, said “Yes, it really, really works. It was so good, we couldn’t believe it the first time. We had to repeat everything two times, and now we know it really works.” That was really satisfying.
Magoun:
Oh, my goodness.
Baliga:
Then I had a fourth idea that was actually a very important innovation. It’s called a split-gate MOSFET, that I invented calling it the GD [graded doped] MOSFET. The IGBTs are used for all the power space above 300 volts, but below that is silicon MOSFET territory. I invented this device to be used in that space, and today, it’s become the dominant technology for silicon MOSFETs in that space. It’s that good. All the companies are now building this device. So when the Inventors Hall of Fame asked me to list the IGBT patent, they said, “You have so many patents, we can allow you to list a second one on the website, on the display.” I said this is the second one, so it’s listed there. What’s cool about this one invention is that my inventions are now used for every power electronics system from low power to high power.
Magoun:
Ah, yes, yes. It’s too bad they don’t have your initials on each little transistor or chip.
Baliga:
When I invented the JBS diode I told you about, they thought JB stands for Jayant Baliga, but it was actually “junction barrier.” No, I didn’t put my name on everything.
Running Start-ups as a Professor
Magoun:
A couple of thoughts, one of which is, with each of these start-ups, you’re juggling an awful lot of technology and management, and how are you managing that? Because you’re starting up a company, and it sounds like five years later, less, you’re starting another company with another technology, and presumably, you still have graduate students and maybe an occasional teaching obligation at NCSU?
Baliga:
I gave up all the teaching and graduate students at that point. So no graduate students from about 2000 to about 2006 or so. Then I came back to the university. All the work at the companies was very hard because I had no employees to hand off stuff to. That was the hardest part when you first get started. Once you have employees, then you hopefully get some very good people. It becomes easier because you can rely on them to do a lot of stuff. So, as the company grew and we had more people, and I was able to hire the right people, it got a lot better.
Magoun:
Right.
Baliga:
Oh, but one thing I should mention, is a lot of faculty start companies through their own students. Either the student goes off and starts the company or the faculty member brings the student in. I felt very uncomfortable with that. I did not involve any of my students in my companies.
Magoun:
Why is that?
Baliga:
It can, if the student is still working on the degree, be a very, what do you say? Conflict of interest. It’s not the right thing to do, in my opinion. After they graduate, of course, they are free to start a company, if they want to. But most of my students get all these fat job offers from companies, so they don’t start their own company.
Explaining the Inventive Process
Magoun:
With each of these newer inventions or significant solutions to problems you observe, [what] would you call this? I don’t know how one describes it, whether it’s cutting-edge technology or simply the insight of experience in having 40 years of expertise in semiconductor engineer and physics that allows you to come up with a solution that other people haven’t bothered to look at, or they don’t have the expertise. What accounts for your ability to come up with this?
Baliga:
Well, I guess if I knew I could bottle it up and start selling it. I don’t have the answer to that, but I can tell you that once I tell people the idea and it’s good, people say, “Oh, this is so obvious, so easy,” and then they kind of start putting it down, saying it’s too easy, too obvious. But until then, nobody came up with it.
I have now several inventions of that nature. We have great success with those, but we are now hoping to commercialize. There is a new silicon carbide-based device that we have created which is called a BiDFET, bi-directional field-effect transistor. This is achieving what we call the Holy Grail of power devices; I’m giving a talk next week about this device at a webinar. The story behind it is power electronics has always needed a device that can support voltage in the first and third quadrants in the I-V [current-voltage] characteristics, carry current in both quadrants, and control the current with a gate voltage in both quadrants.
That device was first made, curiously enough, back in the ‘60s. We called it a triad and sold in millions, but it is a thyristor device, and when it turns on, it creates all kinds of noise. On the old CRT [cathode-ray tube] TV sets, you would get all these fuzzy lines, and it was not making consumers happy, even in those days. It doesn’t give you the kind of control the power electronics people want, so it went out of favor and died out in the ‘60s or early ‘70s. Since then, there has been no device that can do this, and we keep seeking ways.
Now I have finally come up with one, about three or four years ago. I’ve finally received a patent on it, and we are building a whole new type of power electronics converter with it. It’s called a matrix or cyclo converter, as opposed to the previous VSI [voltage source inverter] converters. As a consequence, we can get rid of a big, bulky electrolytic capacitor that is still in use, and that is very much limiting the system. Its temperature is limited. It has a big size and weight, and its cost and reliability is bad. I’m hoping that with my invention, we can enable these new converters and we are able to replace existing approaches, so that can be used everywhere: electric vehicles, or in whatever you want to do power electronics with.
My colleague at NCSU who does power electronics is super excited about the device. He can’t rave enough about it, so we have a program with DOE, the solar initiative, SETO [Solar Energy Technologies Office], and they’ve funded us now very close to the middle of the third year of funding.
Magoun:
Right.
Reviving Semiconductor Foundries in the United States
Baliga:
That’s why we built this transistor in a silicon carbide manufacturing foundry and we built the converter. DOE is very happy to see the results. That’s why NC State also wants to market the device now and see what happens, in terms of getting it out commercially.
Magoun:
I take it that NC State and the state of North Carolina are rather happy with the returns on their investment in you.
Baliga:
Let’s hope so. They’ve certainly given me all the awards that the university can give, including the [Alexander Quarles] Holladay Medal, which is the highest award the university can give. But curiously enough, even before they gave me that award, they gave me the highest award given by the whole university system, consisting of sixteen universities, called the O. Max Gardner Award for the person who has made the greatest contribution to the welfare of the human race. Wow, that was quite an honor. So, yes, I guess they have appreciated what I’ve done.
But that’s not all. You know the other things I’ve done for them, my own center, the PSRC, but after that I helped bring in a big, $50 million NSF grant to do the Freedom Center for Microgrids that went from 2008 to ‘18, ten years. and is very successful. After that, I helped bring the Power America Institute together. This is $140 million program put together by the Department of Energy.
Magoun:
Yes.
Baliga:
Under President [Barack] Obama’s Manufacturing Institute Initiative, I helped stand up the foundry in Texas, and I’ve created there what I call the national process to make silicon carbide power devices. We’ve licensed that to some companies. I’ve had a lot of other interesting irons in the fire, but those are big, visible programs that the university likes to see. Those are the things they really appreciate, I think.
Magoun:
Two separate thoughts. One is the observation that you seem to be contributing to. . . one would call it a revival of foundries in the U.S., as opposed to going to Asia?
Baliga:
I think this was an attempt to revive the foundry business in the U.S. What happened here is that we promulgated the idea that you can take an aging silicon fab, which may be antiquated to make silicon ICs because to make the next generation, you have to invest huge amounts of money in the equipment, Instead of that, build the silicon carbide technology in the same line, and we recognize that, about 70-plus percent of the equipment and processes are the same for silicon carbide as for silicon. What you have to invest in is the rest of the 30 percent bringing up that equipment, investing in those processes. That’s what the Power America Institute funded, so the foundry received a lot of support from DOE to buy that equipment, to install it and start running that foundry. They have now over, I think, a dozen companies that are building product, silicon carbide device products. It’s been a very good success story.
Honors and IEEE Experiences
Magoun:
My other thought is, did your parents get a chance to see some of your successes at GE, if not later?
Baliga:
Yes, my parents were alive when I became a Coolidge Fellow, for instance, created IGBT and so on. So yes, they definitely saw that. They did not see some of the later successes, like there’s nothing as wonderful as being told that your son is going to the White House to be honored by the President of the United States.
Magoun:
Yes.
Baliga:
Something I never dreamt of as an immigrant coming here with, as I said, about eight dollars in my pocket and a suitcase. To end up in the White House with the President honoring you is just incredible. Unfortunately they were not alive when that happened. And unfortunately the IEEE Medal of Honor, for example, that’s much later, 2014. They passed away in the 1990s. They saw me getting other awards, like today, I dug up my Lamme Medal. It’s a pretty nice medal that IEEE gave me earlier on. I’m wearing a pin, which is the IEEE Medal of Honor pin. As a historian, you might find it interesting that I’m the first recipient of this pin, even though that medal has been given for 100 years. They created this pin the year I was given this pin.
Magoun:
That does bring me around to the question of the role of IEEE in your career. Did you have a student section in India?
Baliga:
I was not involved with the IEEE when in India. As soon as I came to the U.S. and joined RPI, I found there was a very good student section of the IEEE. I immediately joined as a student member, and became an officer of that section. We actually did something very remarkable. We held huge numbers of seminars and talks by people which was not happening before. It became a very lively and very, what should I say? Effective student chapter, so after that, when I joined GE, I became an officer in the Schenectady section as well. I kept a very active role in IEEE all along. Then afterwards, I would be on numerous committees and so on. That’s been something I’ve done throughout my career.
But one of the wonderful things I was able to do was start a whole new conference called the IEEE ISPSD, International Symposium on Power Semiconductor Devices. We felt that there was a huge conference on semiconductors, IEDM, which you may know about, International Electronic Devices Meeting. That’s the one we used to go to, but power devices were always a very small subset of papers there. It took people like me to be members of the committees to solicit and get those sessions on power devices. I don’t even think there was a session before I joined the committee. So, we were able to do that over the years. There was still some activity, but we felt there was enough work being done around the world, we should have an international conference. I, and another person from Bell Labs, we got this idea and we convinced the IEEE to support it, but of course, we had to be the first officers. I agreed that he would be the chair of the conference, and I would be the first technical program, what’s the right word?
Magoun:
Coordinator?
Baliga:
Something like that [Chair]. I put together the whole system of how papers are screened and the decisions made on how to accept papers. I put together a committee of people from around the world to evaluate papers, so I created the whole procedure based on my knowledge of how it was done at IEDM. I knew that from my previous experience, and then it became a very robust conference where people had confidence that we were doing this in a very professional way. After the first meeting was successful, I chaired the conference for the second meeting, and then at some point, I suggested we make it rotate between different countries and continents. We started having them in Europe, the U.S., and then also in Asia. Now it rotates between three continents. It’s a very successful conference. It’s still running.
Magoun:
When did it start?
Baliga:
Let’s see, I think the first one, I chaired it in 1991, so I think we’re talking about ‘89 or ’90 [1988] when it was started.
Magoun:
Do you have a fairly dynamic section at NCSU for the students?
Baliga:
How do you mean section?
Magoun:
Student section at the university?
Baliga:
There is a very good student section at NCSU, yes. I am not engaged in it, but there’s activity there, yes.
Magoun:
I see. Is there anything else, Jay, that you would like to add before we close? This has been a very productive two hours.
Innovating Wide-band Semiconductor-based Power Devices using Baliga's Figure of Merit
Baliga:
Yes, it’s been a pleasure talking to you. As you can tell, remembering all these interesting things, facets, is always fun. To me, I’m well known for what I’ve done in power devices, but even there, it’s mostly the IGBT. People forget that I’ve done many other things. In fact, we forgot something that I’ve done, which is probably equally important as the IGBT, which is to create wide-band semiconductor-based power devices as technology.
Magoun:
Ah.
Baliga:
When I was inventing the IGBT in 1980 at the same time, I sat down and derived an equation relating the properties of the semi-conductor material to the performance of what we call MOSFETs and Schottky diodes, and no one had somehow bothered to do that. It was interesting that people thought silicon is fine. We don’t need anything else, I guess, but I had an intellectual curiosity: what can I do if I change the material? So I derived this equation, and that equation is now named after me. It’s called Baliga’s Figure of Merit.
Magoun:
Oh.
Baliga:
A famous equation now. Everybody uses it. You can assess how well another semiconductor material will perform compared to silicon using that equation. At that time, I predicted thirteen times improvement with performance by switching to gallium arsenide, because gallium arsenide was well characterized by a material at that time. GE got really excited by that. You know, if you get an improvement by two times, three times, the semiconductor industry gets pretty excited by that. But when you see thirteen times, not thirteen percent, thirteen times—
Magoun:
Order of magnitude.
Baliga:
—order of magnitude, that doesn’t come along very often, and GE had production going on with the gallium arsenide lasers, LEDs. I said, “We can build product there.” So, while I was bringing up the whole IGBT idea, they gave me ten people to create this technology. I was doing that in parallel, and we had success. We had the first functional gallium arsenide devices, which then became products.
After that, I realized that by switching to silicon carbide, I could probably get even better performance, but silicon carbide material was not ready. In fact, at the GE Labs, I went around asking the materials scientists, and they said, “We have these tiny slivers of silicon carbide. We don’t have wafers. We don’t know much about them.” After I went to NCSU, I talked to a materials science professor, Professor [Robert] Davis who was doing materials, and he said, “Yes, now we are starting to figure out how to grow the material, so you can think about using that.” His students spun off a very successful company called Cree in ‘87. I was able to purchase the silicon carbide material from them and start making the first power devices.
At that point, I thought the material might be 100 times better than silicon, not thirteen, but 100, two orders of magnitude. Then my students and I did these careful measurements, and we found that it’s actually 1,000 to 2,000 times better. Incredible. That has proven to be true, so this has been the platform for all the success of silicon carbide now.
It’s taken a long time for the wafers to get to the size we need and the cost to drop to make it commercially feasible. That took all of the ‘90s, and then around 2005, they made the first commercial JBS diodes, the same JBS concept I was telling you about, and made a product. At PSRC, we had shown that product idea before, in the ‘90s, to make not the product, research structures in the ‘90s, and then in the 90s, also shown how to make silicon carbide power MOSFETs. They needed to work out some more bugs, and finally got a MOSFET product in about 2010. Now that’s become a realistic, there are even electric vehicles being driven using silicon carbide by Tesla and Toyota.
It’s another success story completely parallel to my IGBT activity. One of my colleagues once said, you are a very unique individual who has allowed an area to jump twice, quantum leaps in technology, one with the IGBT idea and now with this idea. It’s quite unusual. I guess it doesn’t happen that easily and often, so that’s very satisfying in order to do it.
Magoun:
No, I appreciate it.
Baliga:
That’s what I’ve been doing recently.
Magoun:
Well, thank you very much, Jay. I appreciate your time, and I guess we will be following up, but there, there’s a lot there, and I hope we can add to your body of work with proper links and more documentation.
Baliga’s Figure of Merit and the Uses of Theory at GE Semiconductor
Baliga:
I didn’t bore you with all my theoretical papers and all the fundamental science papers.
Magoun:
Well, I tend to think that’s explained in the papers, and if I was a better historian of science I would be able to interrogate you on the epistemology of it.
Baliga:
I can tell you one amusing thing.
Magoun:
Please.
Baliga:
When I first joined GE, they were doing lifetime control, something I told you about, to speed up the IGBT that I did later. But when I joined GE, I found out it was all empirical. They would just throw stuff at silicon in the furnaces and find out what happened. How is it working today? Why is it doing this? There’s no theoretical basis for it. So I sat and derived a fundamental set of premises by which lifetime control happens and I wrote this very fundamental paper on it and that predicts exactly what happens with different elements that you can put in [Baliga and E. Sun, "Comparison of Gold, Platinum, and Electron Irradiation for Controlling Lifetime in Power Rectifiers," IEEE Transactions on Electron Devices Vol. 24, Issue 6 (June 1977)]. It is just something of beauty. You know, it was fun to do something that explains after 30 years what people were just doing by the seat of their pants, and you can change all that into science. It’s really very satisfying.
Magoun:
I’m astonished because I am assuming that there are plenty of powerful physicists at GE—
Baliga:
Oh, yeah.
Magoun:
—that are supposed to do that sort of thing.
Baliga:
Exactly, exactly.
Magoun:
But they didn’t.
Baliga:
But they didn’t, yes. Baliga’s Figure of Merit, they could have derived that 20 years before me, if they wanted to.
Magoun:
Was it that nobody assigned it to them, or they didn’t. . .?
Baliga:
Nobody thought of doing it. It’s just nobody had the idea to think of investigating that avenue of the research, I guess.
Magoun:
Is this part of the disjunction between a corporate research center and products, perhaps, that, if you had the product people saying you need to be more accountable for the money you’re spending, this would be a worthy equation?
Baliga:
That’s an interesting point because I just finished reading a book about Willis Whitney.
Magoun:
Yes.
Baliga:
Do you know Willis Whitney?
Magoun:
George Wise’s book [Willis R. Whitney, General Electric, and the Origins of U.S. Industrial Research (1985)].
Baliga:
Created the GE Research Center, so I just finished that book about him. It’s fantastic to read. I had it sitting on my shelf for 30 years.
Magoun:
It was one of the first books I read in graduate school.
Baliga:
My goodness, okay. You know that book. He talks about the fact that the research industrial labs are for product and commercialization, and science is to be kept at a minimum. He only let [Irving] Langmuir do a little bit of science, and he ended up with a Nobel Prize, of course. But GE changed, and was doing a lot of science when I went there, under [Reginald] Jones. Of course, Welch came and changed that culture back to more product driven. But for me to do all these theory things [it] was all offline, meaning off-hours kind of activity. My job was to produce something useful.
Magoun:
Oh, goodness.
Responding to IEEE Recognition
Baliga:
I did that also because I wanted to contribute to literature and produce something useful. It was very good for me from a personal standpoint because that gave me recognition, recognition among the professional people. Something you must have noticed is that I became [IEEE] Fellow in 1983 at the age of only 35.
Magoun:
Yes.
Baliga:
That’s rather young for an IEEE Fellow, so when I started, I thought, “Oh, maybe when I’m 55 or 60, I’ll be IEEE Fellow.”
Magoun:
Oh, dear. All right, Jay.
Baliga:
It was an interesting story because I never thought I would be Fellow at that age, but Professor Ghandi called me once and said, “I’m nominating you to be Fellow. I said, “What? I just turned 35.” He said, “you’ve done more than all these other people who are becoming Fellows,” and he puts it in. I become Fellow the first time he puts my application in, so I guess it was well deserved. Same thing is true for National Academy members. I’m one of the youngest people to be made member of the National Academy at the age of 45.
Magoun:
That’s right.
Baliga:
After I became a member, I looked up the roster. It said average age of the Academy is 70.
Magoun:
Oh, my.
Baliga:
Oh, my. That’s because members do age. I shouldn’t say that’s the entry level. So, once you’re a member, you’re a member forever.
Magoun:
Breaking the curve again.
Baliga:
Yes, but it is nice that, if you can achieve certain things and people recognize it, that’s of course the most satisfying.
Magoun:
Yes.
Baliga:
I always say that I’m hoping the recognition is for my work, not for me as such: what the work is done, what IGBT has done. In the last ten years, I’ve spent a lot of time on trying to bring out the impact of the IGBT.
Magoun:
Yes.
Baliga:
I wrote this book called The IGBT Device [2015], out of which 400 pages is on the applications and transportation and lighting and so on, so that people are aware what this one innovation has done. The other part I’ve done is document the energy savings, the gasoline savings, the carbon dioxide reduction, and consumer cost savings; it’s been a very interesting exercise. [At] the IEEE Medal of Honor, they asked me to say a few words. I had to give an Ignite Talk. That’s a really tough talk to give. It’s a five-minute talk where the slides are run automatically. You don’t control them. I had to think about something interesting to speak about, not the technical stuff. My talk was all about how what people should do in their lives is at some point start assessing what is the meaning of my work, how does it affect my immediate surrounding, my own company, then how to affect the next layer of society around you—maybe the world, you know? I peeled this onion at the meeting and showed you can do things. It was just fun to do.
Magoun:
In five minutes.
Baliga:
In five minutes. It’s a really tough talk, because the slides keep changing, whether you have finished your speech or not.
Magoun:
Ah ha.
Baliga:
Another fun thing to do. Anyway, I did not forget to talk about wide bandgap, I mentioned that to you because that’s a very important phase of my life. So I look forward to getting something back from you and we’ll see what happens. Are these histories available on the website so I can look at other people’s?
Magoun:
Yes. Thank you very much, Jay, and I will be in touch. We will be sending this off to the transcriber, and the turnaround is pretty good. Then it takes me some time to go through and then share it with you to clarify or correct or whatever you feel is necessary.
Baliga:
Appreciate that.
Magoun:
Okay, well thank you very much, Jay.
Baliga:
Okay, fun talking with you. Thank you.
[END OF RECORDING]
