IEEE
You are not logged in, please sign in to edit > Log in / create account  

Oral-History:Luigi Dadda

From GHN

Revision as of 16:41, 11 September 2009 by Nbrewer (Talk | contribs)
Jump to: navigation, search

Contents

About Luigi Dadda

Luigi Dadda
Luigi Dadda
Luigi Dadda, born Lodi in 1923 and among the first researchers on computers in Italy, was the rector of the Politecnico di Milano technical university, where he had graduated in 1947 with a degree in electrical engineering. He research turned toward models and analog computers, and in 1953 he received an NSF grant to study at Cal Tech. In the interview, he describes his participation in the design of a computer at the Computer Research Corporation in San Diego; he would then travel with the computer, packed in hay, to Milan and, with some trouble, get it through customs. He then describes the engineering applications of this early digital computer, as well as the resistance of some engineers to its use. After a break in the interview, he discusses computer education and research in Italy. In 2000, Dadda became an IEEE Life Fellow for contributions in the field of arithmetic architectures for computers and DSP systems.


About the Interview

Luigi Dadda: An Interview Conducted by David Morton, IEEE History Center, 13 August 1996

Interview #282 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 the IEEE History Center Oral History Program,  39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

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

Luigi Dadda, an oral history conducted in 1996 by David Morton, IEEE History Center, New Brunswick, NJ, USA.


Interview

INTERVIEW: Luigi Dadda (#282)
INTERVIEWER: David Morton
PLACE: Milan, Italy
DATE: AUGUST 30, 1996

Education and Wartime

Dadda:

[Inaudible passage] and well, I spent all my student life in the university [Politecnico di Milano] in this school, and then I went Harvard. So I’m all [inaudible passage] I can with the school at the time [inaudible passage] and then I got it in ‘47, in spite of the war, or because of the war, I don’t know anything about war.

Morton:

Were you able to avoid being drafted?

Dadda:

I was asked to postpone my military service because they needed engineers. So provided I was a good student, I was literally asked to stay here and study. I wanted to be an engineer. And of course, during the subsequent time and during the summer we got military instruction. But the main instruction was to become an engineer, and that was the reason why I succeeded in not going to the front.

Morton:

That’s good.

Dadda:

To the front, but no. In any case, it was not an easy life. There was some bombing, you know, at this point. Well, I was going to go in Electrical Engineering, but really, my thesis was on the design of a radio link between Torino and Venice based on microwaves — at the time it was one meter long, you know? Because that was the only technology that was available. And why I mentioned this is that the technology was not really digital but pulse. It was pulse position modulation, and the multiplexing was in time. That was my first work. So when I came in contact with the pulse, or, digital technology, I had a background based on electronic tubes, of course. Remember the first computers were vacuum tubes.

Electro-Synchroton Machine

Dadda:

Well, as soon as I took my degree in ’47 I was asked to stay here in the school as an assistant starting to teach in Electrical Engineering, and doing research in a specific laboratory, which was called Center of Electrical Modeling, chained to the Studio di Model di Electrici. Modeling is a physical phenomena used to measure solutions, determining experimentally the solution of problems. My specific field being the problem of solving Laplace equations [for] different parts of physics and engineering using electric models. One of the things I would like to mention is the work we have done here, taking part in the design of the first electro-synchrotron undertaken by our physicists and being asked to study the magnetic field. It was a very important to determining the stability of electron trajectories.

Morton:

When was this?

Dadda:

And this machine has been built.

Morton:

But when?

Dadda:

In 1957 or so. And well, this was quite an interesting story. The initiative was taken by the University of Pisa, which got funds from a number of local institutions, banks and local government, and this machine had to be established there. But then the central government said, “Well, this is in the national interest so this will be paid by the central government, but the machine has to be in Rome,” of course. At that time the University of Pisa had a lot money to spend, and they asked Enrico Fermi about what to do with that money. And they wrote a letter to the Pisa director. He said, “Well, I suggest that you take an initiative in the computing field,” and that is the origin of the Pisa Group, who designed and built the machine itself.

So it made a connection to the state. And I was involved in these kinds of things. At the same time we got the first information about what had been developed in the United States about electronic computers, digital computers, because before that time we didn’t know anything. And our director took the initiative of putting a request to the European Economic Program, called the Marshall Plan. It was a big program started by the United States to help the European industry to get out of the trouble in the fifties. But when this was being done, there was not much hope that it would be accepted, because it was a university which was asking and not an industry. But we presented a problem in which we said that we wanted to develop this technology and also teach industry about the use of this new instrument. And that was the reason why it was unexpectedly accepted.

At the same time, in the ‘52-’53, I wanted to go to the United States, after starting to work in the analog computing field. And, well, I got this call from the National Science Foundation to stay for two years, and they paid also for my family to go over there. The choices being the California Institute of Technology, Cal Tech, to work with a professor, who was working in the analog field. And I was ready to go in the spring of ‘54, but at a certain moment my Rector called me and said, “You know, they said yes to our request. So you go to the States, but for us.” And I was very much [unconcerned with] what group.

Selecting and Building the Machine

Dadda:

The machine that was being selected, and I will tell you how it was being selected, was built by a company in Los Angeles, in San Diego. So in the same place as Cal Tech — Cal Tech is in Pasadena. It’s not San Diego, but it’s still Los Angeles. So I decided to go take care of both things, you know? That’s been a good choice for me. The choice of machine is being a problem, because we didn’t know anything about those machines. But we got some information. In particular we got in connection with a gentleman from the National Bureau of Standards. His name was Dr. Alexander. He was in charge of a program in the National Bureau of Standards concerning these technologies. They had two main projects, one in Washington, DC, and the other was in Los Angeles, UCLA. In Washington they developed a serial machine based on the mercury delay line. In Los Angeles it was not in the university there, it was within the university campus, but it was a different, say, independent laboratory established in a military barrack. I remember, because I went there a number of times. They developed a parallel machine based on Williams tubes memory, and so I had the — since I was familiar with both these two groups I had a chance of getting at the front here, if you want, of this technology, while taking care of my own smaller machine.

Well, through the advice of these people, we decided, we means not me myself, but the group of people taking care of all of these things in Washington, DC, and chaired by a man whom I’d like to mention, Mr. Ortona, who became after that the Italian ambassador to the United States. So, you know, this other connection with different people takes part of this strange activity. Well, the machine was selected because of its cost. It was not that the cost was very high. The cost was a hundred and twenty thousand dollars, and at that time, and that was a lot of money, but it was not enough to buy say, a 701 or 702 or a much bigger machine. But in any case, the machine was part of a group of machines of the same kind which started to be built by different industries, smaller than the biggest of the three machines, but, in any case, quite important. By the way, the magnetic drum was not the memory in that case. Remember the first magnetic drum memory machine built by IBM came later, in ‘55 or ’57, it was the famous 650. This was big event for IBM, who had been on the market for a number of years.

Well, our machine was built by a small company called the Computer Research Corporation of California. It was an outcome of people coming from the Northrop Aircraft Company, and also people coming from National [inaudible], and also from the MIT. The designs and machines was strictly based on the micro-programming architecture. At that time it was quite new. I told you the machine still exists here, not operating, of course. It was quite interesting since it was very easy to program with 42 bits, three address instructions. So it was easy to program, you see. Program in binary or octal, because the addresses were given in octal, but we were able to work with that base.

Transporting the Machine

Dadda:

And well, I went there to do work, to take part in the construction, because the contract was such that when the machine was constructed by the company in the factory, tested, and put into a ship out in the Long Beach Harbor, from that point in time it was my own responsibility. I was too young to understand what this could mean. Nobody had explained. Also about transportation of machines, originally it was believed that we could transport this thing by air. But at the last moment they decided that they realized that there was not possible So I had to rely on a ship. Everybody was worried about the problem of having a machine like this for a month on an old Liberty, it was a Liberty ship. We were worried about vibration. You know Liberty ship was a fleet working, doing all the deep water transporting of things across the Atlantic.

Morton:

So it was a military ship?

Dadda:

Yes, it used to be.

Morton:

Oh, but it was being shipped by some company?

Dadda:

Yes, it was a pilot ship belonging to some kind of company.

Morton:

But some former Navy ship?

Dadda:

No, of course, but it originally was used for transporting military things around the Atlantic. And I was very fortunate because the captain was from Trieste. Everybody was very curious about this strange kind of things we were transporting. One day I went to visit him in Long Beach. I found that they had just come from a port in the Mexican Gulf. They had collected a number of cotton bales. I say yes. I have idea. I say, “Look, take some of these bales out from the center, put the machine in, and then put them on top.” And it was a perfect situation for damping vibrations, and after a month of navigation the machine was taken out in Genova [Genoa] and we transported by car, by truck in Milano, and it worked almost immediately. You know, it seems very simple, but it was not so simple at that time.

By the way, in Genoa I had a problem with Customs, because though this machine was a gift, but we were given it to take care of this. The first problem was what is it? Based on the names written in the book, you know, of goods, there’s no computer; it doesn’t exist. But the problem was that in this country there was a law compelling you to pay a tax and to put a label on each vacuum tube. It was a radio tax. But I have six hundred of those tubes, and moreover, I have six thousand diodes, Germanium diodes. I tried to convince the custom people that those Germanium diodes, after all, they were resisters. But they say, “No, these are diodes. They are tubes.” It was a very small tube, and the problem was that the stamp was much bigger than the tube itself. It was impossible. And so the solution was found the following way. I got a pile of those sheets of stamps and I signed a declaration. I say that I will myself apply those stamps on the tubes. And they knew very well it was impossible. But I finally succeeded in getting through Customs to bring it down, you know.

Uses for the Machine

Dadda:

Well, when this happened I came to the States in early Spring, passed all the Spring and Summer, and then came back by plane, of course. And before the machine arrived they already started a short course for professors and assistants of the school. And what I can tell you is their reaction. The reaction was of two main kinds, you know. People were very enthusiastic and willing, really, to think about what new could be done. And other people, particularly my colleagues, you know, you know at the time an engineer typically had a slide rule in his pocket, you know. They said, what are we going to do with a machine that is so fast that it computes seventy operations per second? I can’t do this with my slide rule. And they just have to convince those kinds of people that they had to change their way of thinking. But I think it’s typical of everywhere.

Anyway, we started just to use the machine, and we had immediately a number of industries coming in. Two of the most important then were the Pirelli Company where they had started a problem to compute by numerical calculus the electric field, like I told you before, in high voltage apparatus. They have done this by hand. They established a group of thirty people using iteration procedures. The problem was that those people had to be sent every month for a week in the mountains, because they had become crazy working that way. So they came immediately to us and they obliged us to immediately apply our machine to that problem. And this was a big success, a very big success. We went for years doing computation for Pirelli, and they used the computer for the longest time taking, say, up to eight hours.

And the machine itself was not so good — you know, no vacuum tube machine is very good. The mean time between failures could be, say, three or four hours at most, and so it was done during the night. And I was sleeping at night in the center, you know. And from time to time somebody would call me and say, “Come, the machine is stopped,” to resume — to save the computation, and resume the computation after having found the fault. Well, this is being, you could say, a tremendous life, but we were very enthusiastic with this. And the second application was being done by a company called Edison. Edison, by the way, this is not just a name given to it, but Edison was involved in the establishment of this company in the eighties. I will tell you their story. They had the section, a division, in charge of designing dams, you know, in the mountains. Water dams.

Morton:

Dams, yes.

Dadda:

It’s a big construction, of course, very delicate. They, well, the design was based on a solution of a number of linear equations. And all I remember is that why the people from Pirelli started with, say, two hundred equations between the points, and they went up to thousands. They were very modest. They gave us first twelve equations. Then six equations being done in ten minutes, they went up to twenty-eight, but said "no more."

In any case, you know, it was what they wanted. They had a different attitude, you know, the first people. It was after all electronics, and so they did not know how much they could ask. The other people were mechanical engineers so they were more modest in the beginning.

Edison Company in Milan

Dadda:

Well, I told you about Edison. This has nothing to do with the situation, but I have to tell you that this city, Milano, was the first city in Europe having had the first power station, the second after New York. Edison took part in this. The things happened the following way. In the 1882 or so our elector, he was from Israel. So he was interested, knew Edison, and asked him to do something in Milano, and he accepted. They founded a company called the Edison Company. This company still exists. And the place where this power station was established is right in the center of Milano, not far from the Duomo.

Morton:

Oh, really?

Dadda:

Yes, yes. And even the machine that was used still exists, and it is the second in the world. But this is ancient history.

Morton:

Is it in the museum?

Dadda:

Yes, it is in the museum. We have also at our school a machine of this kind.

Updating the Technology

Dadda:

Well, coming to modern times — this was ‘54, ’55. In ‘55 the faculty asked me to give the first regular course in electronic computing in this school. It was in the curriculum of ‘55. I think it’s one of the first in the world, because there was not so much people teaching, and I was teaching programming and teaching, also, logic. At that time we wanted to teach everything from the binary system to logical operations and programming was, after all, only in machine language. So I had to teach everything.

Morton:

Not computer design?

Dadda:

Oh, yes, of course. I will come to this point. This event established here the first computing research center, and a number of, say, consequences and foreign activities in different fields. As a computing center we had development based on the new machines, of course. At a certain time we wanted to establish a consortium with the universities in this region, and then we asked our ministry to do the same in other parts of this country. And, well, this consortium still exists. It runs the machine for the service of the university. Well, the universities already had a number of machines themselves, but in the fifties and the sixties it was hard to have a machine. And I would say that for this reason we can say that Italian universities could use machines which were not small compared with the American universities. They had the opportunity to use this new technology. That was a branch of development based on selecting the best computer on the market.

Well, here there is something to say that they would not like to say, “No, no, I’m joking,” because the problem was to resist the pressure of different companies. We bought an IBM machine, an IBM 7040. But then we decided that the best machine was Univac. So we behaved in a quite independent way. We did not want here to have a company telling us what to do. And this, I think, is a good idea to follow in a university. But this means the user machines, of course, all departments came — Well, we had a different attitude, no. First, among those who were not electrical engineers, who came and appreciated these new possibilities were our colleagues working in the construction field, civil engineers. They understood immediately that the computer was an important tool. And chemical engineers came also. Anyway, everybody now is using machines.

Well, what happens in this department, and this was a small department, is the development of an activity in computer science. And a step at a time we added a new course in the curriculum. Well, the first was half dedicated to analog computing, half digital, because both were important. But now the situation is changed. We have curriculum studies for information science containing all the classical subjects, but, more important than that, we established over time a group of researchers in the languages, in operating systems, in software, and so on, so we have here the classical structure of a department, and we can say we among the first to have developed these kinds of things. Among the research we undertook in the very beginning in the fifties, there were two problems that we took as important, because we were electrical engineers, and we wanted to study filters, and study graphics.

We got one of the first machines produced by DEC — the PDP8. Well, the PDP8 was a very nice minicomputer, but it could not multiply. It had to have a substitute to do multiplication. But that wasn’t really important. And so we did studies in developing fast, fast arithmetic circuits. This has been my specific field, developing fast multipliers. And this is been — I continued to work in this field for a number of years. Even now, you know, somebody tried to compel me to, again, go back to talk about multiplication. I’m quite fed up with it. Well, this is a general line.

When you do graphics, you imagine solids, you know? You have the problem of mapping or deleting the hidden lines. And so solving the problem in general was a problem. And to this problem I say that problem can be solved with lots of research for doing fast arithmetic but also a methodology. Now we have a group of people working in different fields, the philosophy has been first to establish a research group, and then teach in the field.

[end of side one]

[much of side two is inaudible]

Electrical Communication in Milan

Dadda:

— Just before the war we built a radio link between Milan and [?]. You need just two intermediate points. And he asked me, why don't you study a link between Milan and Toledo? What I did was develop the logic part. It was a pulse technique. The technology to develop high frequencies came after the war, with the development of triodes with five grids. That meant that the grids had small distances between them, and you could go to higher frequencies. In this case, it was 300 megahertz. I built a cavity magnetron. It was my first experience with [?]. He was the teacher for two courses; radio techniques, and electrical communication.

Morton:

Was that a small program then?

Dadda:

There were two basic courses. The students were electrical engineering students who studied this branch of the field, electrical communication. I was in this branch.

Morton:

In the United States it was common for there to be a rivalry between the power engineers and the electronics engineers. Was that common here?

Dadda:

Oh, yes. We started to develop a special curriculum for electronic engineering quite early, in 1955, or so. The faculty asked me, even though I was not yet a member of the faculty, to develop this course. Six years later, a new law established this faculty. What we had done is to develop new courses within electrical engineering. We admitted not more than 25 students. But they were the best students. And all of them have had wonderful careers in industry. Many of the people in Olivetti and other places came from this school. I remember in those times that it was a lot of work, but teaching 25 students is one thing. Teaching 250 is different. Much more fun with few people. You don't need to have examinations. We were also teaching digital technology in the early '50s. In '62, the new branch of electronic engineering was established. We made a mistake. It was rushed. Everybody wanted to be in electronic engineering. But we did not open it to everyone, even though that was against the law. But the minister knew that we were doing this on purpose. We admitted them a few at a time. When we opened the new program, we had another tool in our hands, which was very strong examinations. So we only took the best students, and we succeeded in controlling it.

We had three branches in the beginning; computing, telecommunications, and control. Did you interview Mr. Costoloni

Morton:

No.

Dadda:

He must be 85 or 90 now. As these three branches developed, control became the place where bioengineering came. We also established a number of courses in management, for engineers. It has been very active. Costoloni was a key man in two areas. In the mid 30s, he was an engineer at the Companie Generale de Electrizitat  — General Electric you might call it. He was sent to the United States to learn the technology of mercury rectifiers. It was basically a device for converting AC to DC, especially for traction. Before the war, there was a very large conversion from steam to electrical trains. He worked in the States, he came back, and he built his new machines in Milano. After the war, he was again sent to the States to learn about control. In both cases he went as a professor from this school. And he was the man responsible for founding the control programs. So, strictly connected to industrial activities.

By the way, when I mentioned Viciati, he was connected to Magneti Marelli. This was in the field of radio links.

Morton:

Did anything become of that system? The radio link system

Dadda:

This technology was applied in a number of links, but at the same time, just after the war, the technology which "won" was frequency modulation. So it has not been very important. But there is some of it left, because the applications of pulse modulation in transmission became an important issue in telecommunication. We were well prepared. The problems are similar to switching technology, in which you want to multiplex messages on a cable. Using PCM. Frequency modulation was used for multiplexing up to 2500 channels on the same link.

[inaudible passage]