Oral-History:Nathaniel Rochester

About Nathaniel Rochester

Rochester received his BS in Electrical Engineering from MIT in 1941, specializing in acoustics. Soon after he went to work for the Rad Lab. He worked in the RF components group, on components dealing with transmission and reception of radar signals. He ended up in charge of diodes, largely because no one thought they would amount to anything compared with vacuum, so they left someone pretty junior in charge. A lot of basic research on diodes was necessary; in the process he dispensed money to Purdue and coordinated with Bell Labs. He wasn’t qualified to do advanced research himself in solid-state physics, so in 1943, by mutual consent, he left the Rad Lab for Sylvania, where he helped design and build radar sets to Rad Lab specifications—something he could do well. He stayed at Sylvania till 1948, when he moved to IBM. The rest of his career was computer-centered.

About the Interview

NATHANIEL ROCHESTER: An Interview Conducted by Andrew Goldstein, IEEE History Center, 6 June 1991

Interview # 080 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:

Nathaniel Rochester, an oral history conducted in 1991 by Andrew Goldstein, IEEE History Center, New Brunswick, NJ, USA.

Interview

Interview: Nathaniel Rochester

Interviewer: Andrew Goldstein

Date: 6 June 1991

Location: Boston, Massachusetts

Education and Recruitment to Rad Lab

Goldstein:

I'd like to start by asking you to discuss your educational background.

Rochester:

I was a candidate for a bachelor's of science in electrical engineering with a communications option at MIT in 1941. My father ran out of money, so I had to go to work full time for the full last term of my four years. Luckily, I had been carrying a considerable overload of subjects, so I was able to reorganize things a little bit and use a study project as a thesis, still graduating even though I did hardly any more school work. The first job I had was working for MIT in a project that developed the first functional blind-landing system which is still used in commercial aircraft today. It is not the kind they use in the military. I worked on that until March of 1941, I was told by somebody in authority at MIT that I should go down to a particular room. My services were needed on a project that would help the war. So in that way I was recruited into Radiation Lab.

Goldstein:

Was it a supervisor who was familiar with your work on the blind-landing?

Rochester:

I suppose it was probably Don Carew, but I don't remember who did. The first significant event that happened to me after I became a member of the Radiation Lab was that I discovered that every Monday night there was a seminar. I immediately began to go to those seminars. I was accustomed to going to school, so this was just a continuation of it. I thought I'd learn all about radar and that turned out to be correct. The first seminar was just amazing to me. It was a man named Oliphant who came over from England with the first cavity magnetron. When I had studied magnetrons in my courses, it was a vacuum tube that could produce milliwatts of power. The cavity magnetron produced 10 kilowatts — about a million times as much power. This was a complete source of astonishment to me, that anything like this was possible.

Goldstein:

What was your undergraduate work in? You said that you had converted an ongoing project to a thesis. What area was that work in?

Rochester:

That was in acoustics. I thought that would be a good thing to specialize in. The work that I did, which was converted to a thesis was actually published in the Journal of the Acoustical Society of America.

Goldstein:

When you were at MIT, were you aware of the existence of the Rad Lab? Or of radar? Were these familiar to you?

Rochester:

No. I wasn't aware that the Radiation Lab had been started. But it was extremely small then. I would guess something like 20 people were in it. Maybe 50 people. But it wasn't huge. That was at the time I joined it. So it didn't occupy much space. Not much was evident to anybody that wasn't inside the Radiation Lab.

Goldstein:

They introduced you to radar when they recruited you to come work?

Rochester:

I've forgotten exactly what they told me at the time. They probably would not have told me exactly what they were doing because that was secret. Perhaps they did and I'd promised not to reveal it. I do remember what happened after I got into it. I was put in the RF components group, which means the components that deal with the radar signal that is transmitted and received. This was sort of like acoustics to me because the wavelength was about the size of the objects that were dealing with. Electromagnetic waves are more complicated than acoustical waves because they have an extra dimension. But a lot of the things were the same. So that seemed like an interesting thing to me and something that I could immediately jump into and do something with.

Seminars and Information Sharing

Rochester:

Let me go back to the seminars for just a moment. They were very important to me. Gathered there were wonderfully capable — in some cases famous — scientists, physicists, mathematicians. These people would give the seminar usually, and they would ask for questions and comments, and there'd be discussion. It was very enlightening to hear the thoughts of people so capable as this in a seminar in which we were all equals. Except some of them there were more equal than others. Not only did I learn at the seminars, but then, having listened to these people, I could make friends with them later. I would have some way of starting a conversation and carrying on. This had a tremendous effect upon my life subsequently because very often when I would meet some engineer or scientist, it would turn out that either he was in the Radiation Lab, or we both knew somebody that was in the Radiation Lab. It was a good way to make a new friend. That was very helpful in making contacts. So that was another of the neat things about those seminars.

Goldstein:

<flashmp3>080 - rochester - clip 1.mp3</flashmp3>

Did you find that this seminar was the most effective way of disseminating information? Did you also speak informally to other members in your group or other groups?

Rochester:

Oh, yes. There was a great deal of personal communication. And then a great deal of communication on paper. I also had a lot of studying to do because there's a difference between electromagnetic radiation and acoustical radiation. The things that had been developed were not known outside of the security environment. But a very lucky thing happened to me then. The magnetron was able to create the transmitted pulse at very high frequency, and it was possible to make oscillators that would oscillate at these frequencies using vacuum tubes. But there was no vacuum tube that could process the echo signal that came back from the target. The only thing that was available that could process it was the diode, which was the same sort of diode that had been used in crystal radios back before they had had vacuum tubes — or before vacuum tubes were very good. And so it was possible to take a local oscillator signal and the returned echo and process them with a diode. Out of that would come the signal that represented the echo, but it was at a low enough frequency to amplify. The people in charge were sure that very soon vacuum tubes would be doing this. It wasn't a very important job to be in charge of these diodes. I was given that job because I was very much lower in academic status than the average person in the Radiation Lab. It didn't turn out that way because it was never possible to produce a vacuum tube that would do what these diodes were doing. This generated a great deal of research on these diodes — the solid state and silicon — which laid the groundwork for inventing the transistor. The transistor and the diode eventually replaced practically all the vacuum tubes. So here, just by accident, I was sitting on top of an important phase of the future.

The RF Group and Diodes

Goldstein:

Who were you working under in the RF group?

Rochester:

It was Salisbury. He was the head of the RF components group. He was a very interesting person to work for. It was valuable working for him. This was a uniform experience in the Radiation Lab. There were such great people around that you couldn't help but get better.

Goldstein:

When you say you were responsible for these diodes before it was recognized that they were an essential component and weren't going to be superseded by some vacuum component, what was the program for them?

Rochester:

There was a good deal of work going on. Some at Radiation Lab. More under contract to Radiation Lab at Purdue or was some university out in the midwest. Their project was completely funded by the government to do research on diodes. It wasn't something that they wanted to set up in the Radiation Lab. I think they weren't fully convinced this was where the future lay. In fact, they were convinced it wasn't where the future lay. But obviously something had to be done.

Goldstein:

Were you trying to improve the operating qualities and characteristics of the diodes? Or trying to design around them in anticipation of vacuum tube components?

Rochester:

No, they were trying to improve — they were trying to understand what it was all about. Diodes were very poorly understood at the beginning. Somebody had discovered that if you take a piece of tungsten wire and press it against a piece of silicon and poke it around different places, you could finally find a place where it'll produce signals that you can hear in earphones. So here you have a radio — a radio receiver — without any vacuum tubes in it. It was just a very simple thing but it did work. The level of knowledge wasn't much better than that initially. The level of knowledge about solid state was very primitive compared to what it is now. More work was going on in this in Bell Labs. They'd been working on this thing a little harder than anybody else.

Goldstein:

From before the war?

Rochester:

Before the war I think they were concerned with using a diode as a telephone component that wouldn't use any power. Instead of using silicon that you dug out of the ground, they were purifying silicon and adding impurities to get the proper semiconductor effect. They were leading in this and the whole thing just progressed more knowledge developed.

Goldstein:

Were you working on a theory of the diode? What were your responsibilities?

Rochester:

My responsibilities were to dispense this money to people in I think it was Purdue and to coordinate with the people from Bell Labs. I was concerned with the design of a circuit that used the diode. But I wasn't really qualified to do research on solid state physics.

Goldstein:

Was there anyone at Rad Lab doing that?

Rochester:

To some degree. The people at Bell Labs and the people at Purdue were getting into this much more deeply. We had a wonderful consultant — Fred Seitz who later became Rockefeller University, a very prominent university, in New York City. Prominent not in the number of students because the number of faculty outnumbered the students 2 to 1, but the research was spectacular. Anyway, Seitz was a major expert in solid state. And at that time there were other possibilities of materials to use other than silicon. Germanium was the other one. Germanium was easier to work with, so there was a strong temptation to use germanium instead of silicon. But Fred said that silicon is the very best because it has a higher binding energy so you can do a lot more with it. It can stand a lot more electrical abuse, and it can keep its ability under conditions where germanium would give out. It turns out he was right because now silicon is what all sorts of computers and VCR's are made of. So that was a good forecast.

Goldstein:

You said that you were responsible for dispensing money to Purdue, which you had earlier said was supported by federal grant. Was it done through the Rad Lab?

Rochester:

Through the Rad Lab, yes.

Leaving Rad Lab

Goldstein:

How long did you stay at Rad lab?

Rochester:

It was April 1943 when I left Radiation Lab. The job I was doing was taken over by a physicist who had been working for me and who had doctor's degrees and understood a lot more physics than I did. That was appropriate. What I did was to set up at Sylvania. Sylvania is a company that later merged and is now a part of what is General Telephone & Electronics — GT&E. I set up a group there whose mission was to design and build new kinds of radar sets to Radiation Lab specifications, or perhaps take designs that they'd created and refine them a little bit and build model shop quantities of these radar sets. So that was something I could do much better than I could do the solid state physics.

Physicists and Mathematicians at Rad Lab

Goldstein:

<flashmp3>080 - rochester - clip 2.mp3</flashmp3>

I want to discuss your time at the Rad Lab further.

Rochester:

Bringing a whole bunch of physicists and mathematicians and all kinds of scientists into an engineering project (this was engineering, it wasn't science) had a wonderful effect upon engineers. The engineer is a guy who can design something and build it and make it work. And a big part of engineering is knowing what won't work because usually there's a time limit and there's a price limit for getting something designed and built and ready. So you have to be able to get it done and get it working. And get it to work right now.

It turns out that in the Radiation Lab I became a lot more cautious about using any of this lore about what won't work. An example of that is the radio — the intermediate frequency amplifier. As mentioned before vacuum tubes couldn't do anything with the radar echoes as they came back. But the RF circuit, which involved the crystal, produced a signal that could be amplified by vacuum tubes. But it had to be amplified enormously — from almost nothing at all to a very substantial signal. And at the very substantial signal end of the amplifier, if just a little bit of signal leaked back to the beginning, it would swamp out the signal. Instead of getting the amplified signal, all you'd get is a screech a feedback. And so the engineers had gone to work on this, and one particular had come up with radio receivers with tremendous shielding. Each stage was shielded. These things were enormously heavy and big. However, a fellow who became quite a good friend of mine — Henry Waldman — was a mathematician. He didn't know what couldn't be done. So he thought about all these things, and he decided that you could do it differently. He discovered that by placing the components very cleverly and very precisely, it was possible to build a very small, very lightweight IF amplifier. Instead of weighing 40 pounds, it weighed 6 ounces. There was an enormous difference. His technique of building these amplifiers took over. The older techniques disappeared as soon as he got it to run.

It was a funny story about his getting it to run. I went up to his office one day, and he was sitting there with this tiny little thing about 6 inches long that he'd been working on. At that time they had just finished Building 22, a huge wooden structure which had been put up in very short order because Radiation Lab needed more space. They cleared land occupied by a temporary World War I structure to build this whole thing. Henry looked out the window, which looked out over this new building that was being built, and he said, "Nat, I started this project before those fellows began building this huge building. They're done, and I'm not done yet." And he had this little tiny thing in his hand. [Chuckling]

Vacuum Tubes

Goldstein:

Let me ask you about the work on the diodes. Was there much parallel work in private industry? Was there general consensus that this was a moribund technology?

Rochester:

No, there wasn't a consensus that vacuum tubes wouldn't be needed to do it. Everybody assumed that the modern thing was vacuum tubes because they had such a good track record. Until this began, no serious work had been done on silicon or diodes. It was just something that Edisonian-type experimenters would get a piece of silicon, and it came from some mine or something like that or it looked right, and they'd poke around and find a good spot without any science at all. Now, solid state physics is one of the major branches of physics.

Goldstein:

So did you work much with other researchers? Were you aware of much work in the same area? For instance, there was work at Purdue.

Rochester:

Purdue. Oh, yeah. There were other companies, and I don't know just who did what. But I've named the places where the main research was done. But there was excitement at the Radiation Lab. I'm going to come back to diodes in just a moment. It'll relate to this. In the first place, we were at war. And most of us — I, at least — took this very seriously. We were very motivated and very patriotic and inspired by this we were competitive and we wanted to win the war. This was our particular contribution to it, which turned out to be pretty large. One day a friend of mine came into my lab, and he said, "Nat, I have a really important need right now for the very best diode" — crystal he called it; that's what we called them all then — "that you've ever made, that's available and that you can get for me or make for me or anything." So I got him this choice one that I'd set aside as a remarkably good one and I gave it to him. That was on a Tuesday. The next Monday seminar was a real winner because these guys had gone out with the very first attempt at anti-submarine warfare, radar-guided airplanes going after submarines, and had sunk a submarine. So that was really exciting! It was exciting because I had provided this diode that had helped sink this submarine. A submarine costs more than the total money that had been spent by the government on Radiation Lab til that moment. So it was a paying proposition. There were many other exhilarating times like this when wonderful things were done, things that couldn't have been done before, or things that stopped what the Germans or the Japanese were doing. That was one of the tremendous rewards of working in the Radiation Lab.

War Spirit Before War

Goldstein:

Was this patriotic motivation intensified in December of 1941? You began at the Rad Lab in early 1941 before the U.S. had entered. Was there that sense of patriotism initially?

Rochester:

We were at war. And Franklin Delano Roosevelt was at war, but he was guiding his country into war. He couldn't take a strong position because otherwise he would have produced all sorts of reactions in Congress and elsewhere. He had to wait until the Japanese attacked Pearl Harbor. Once they attacked Pearl Harbor he had no difficulty getting everybody else to follow him. But we were already in it. We were at war, and there wasn't any question. We were dedicated to it. We knew what was going on; we could see that things were desperate in Europe and that France had fallen. England came very close to falling and still was in trouble. England needed more help from the United States than it was getting. We were giving everything we could toward it. The Radiation Lab was not an isolated organization, but it was the most important one of these organizations that was fully at war. So Pearl Harbor was a help, [Chuckling] and it was stupid because they had some primitive radar there at Pearl Harbor which detected the Japanese airplanes coming. But the people in charge of the military there were so sure this couldn't happen that they simply ignored it. They said, "Don't bother." So that was another first for radar; it didn't work out very well.

Anti-Submarine Radar

Goldstein:

You say you supplied the crystal for the anti-submarine radar. Was that what your work in RF circuits was applied to or was your work applied to other projects?

Rochester:

The same type of circuit was used in all of the microwave radar sets. There were 10-centimeters, 3-centimeter and 1-centimeter radar, and they differed quite a lot, one from another, because you'd have to construct the radio frequency circuits so that they'd fit these. But usually those different models could be adapted quite easily from one radar set to another. This is called the first detector.

Goldstein:

You mean the detectors you were working on?

Rochester:

Yeah. The first detector converts radio frequency to intermediate frequency. The second detector, which I had nothing to do with, converts the intermediate frequency into the signals that you would display in a vacuum tube. That's why it's called the first detector. All those first detectors were about the same. The design kept improving. More things were learned about microwave circuits, and so everything kept getting better.

Goldstein:

I'm interested in the assistance that the Rad Lab received from industry in designing these first detectors. Was there any effort on the part of private industry to work on these in a mutual effort?

Rochester:

Oh, yes. All the people who were making radar sets had to have them. They all would look at them, and see what they could do to improve them. All the information in the Radiation Lab was available to the contractors who were properly cleared so they had all the data we were working with. They had good people, too. I certainly can't give you a detailed account of exactly who did what, but a lot of people did a lot of different things. To a very considerable degree, information would flow from one contractor to another. There was some feeling that we shouldn't give away our secrets to our competitors because the competitors were bidding against them to get contracts. But there was a pretty good flow of information.

Goldstein:

When you first arrived there, was the group in full swing? Did they already have a model that they were working on?

Rochester:

Oh, no. The magnetron, on which it was all based, was just at that very instant being brought to the United States from England.

Goldstein:

In what month?

Rochester:

The time when I heard about it first was the Monday after April 3rd, when I went to a seminar on the subject. Oliphant came and showed this thing off and gave a lecture on it. But before that there was a considerable history of what had gone on. You can't have a radar set until you have something to transmit it out. There were a lot of other unknown things: the first detector and the transmit receive switch. There were a whole bunch of things that had to be done. Those were all being worked on.

Goldstein:

When you showed up to begin work on this first detector, did you start from scratch or was there some other existing component that you could use as a model?

Rochester:

There was a design of the package in which the diode went, which was in use at first. Then some of the people that were working harder on the circuit came up with a better design sometime in 1941. That one was used all the way through the war. It was used for K-band as well as X-band. This was all beginning right then. I was there on the ground floor, but a lot of people were doing the work.

Goldstein:

Can you tell me something about your personal interaction with industrial suppliers or manufacturers?

Rochester:

I could make more contribution to the war effort and develop better in directions that I wanted to go if instead of staying in Radiation Lab and working on these diodes, I would transfer to private industry. My work was getting increasingly involved with theory and understanding experiments. Scientific experiments were building up a big body of knowledge, which I could read about but I couldn't contribute very much to as it was being developed because I didn't have the educational background. So with the approval of my manager, Salisbury, I changed from Radiation Lab to Sylvania Electric Products. I wanted to do that. I thought it would be a good thing to do, and it turned out to be very good. What we did in my shop was mainly K-band radar, which could have turned out to be something very important, and there was no way we could have known that at the time, but it didn't. What we did was to develop these radar. There were two main products. One was a complete Navy radar. The other was an RF head. This is the thing that takes the signal that comes from the antenna. It's up at the top of a mast somewhere, and it sends the signal down to an indicator. There's a circuit that actually puts the picture out for the people to look at. But it has to be up on the top of the mast because there's so little signal that comes back that you have to use it right away. You can't transmit it down through a wire and lose any of it for then there's not enough left. So this is why it was an RF head. It was an equivalent thing in the radar set for the Navy.

Goldstein:

Was this while you were at Rad Lab or was this at Sylvania?

Rochester:

This was at Sylvania. This was what we built. They would give us a lot of ideas. They would tell us what they wanted done, and they'd solved a whole lot of the engineering problems, but not all. We would have to do the things they hadn't done. If they'd done something wrong, we'd make it better and then see if they'd agree to it.

Goldstein:

Was Sylvania constructing these for both the Army and the Navy?

Rochester:

It was the RAF and the U.S. Navy. These were the two main products. There were a bunch of other products that were smaller. These were the things that we did the most work with.

Transfer to Sylvania

Goldstein:

What were the circumstances of your going to Sylvania? Had you been working with them at the Rad Lab and realized there was a position for you there?

Rochester:

The people at Sylvania in the Special Products Department were mainly concerned with incandescent lamps and Fluorescent lamps. But the Special Products Department also took care of all sorts of screwball things that would come up. This fell in that category. They were making the transmit receive switches that I mentioned. When the magnetron sent out a pulse, you have to do something to protect the receiver from that very first, one-microsecond pulse will destroy the receiver. Then when the echo comes back it has to be received with no attenuation — no unnecessary attenuation — so the transmit-receive switch was a pretty tricky thing. I don't know where the type that was in use then had been invented, but it certainly was improved at Sylvania. So I knew about that.

Goldstein:

While you were at Rad Lab you were aware of that?

Rochester:

Yes. A very close friend of mine, Chris Peck who I later employed, was one class behind me at MIT. He was a good personal friend of the son of the man who ran Sylvania. So I had connections through that. It was a very agreeable place to change jobs.

Goldstein:

Did you contact them and apply for a job and were offered one?

Rochester:

I don't recall the details at all, but it was a very smooth transition. It was hardly like submitting an application and having it reviewed. On both sides there was a friendly cooperation in trying to win the war.

Goldstein:

When you worked at Sylvania, did you have business with the Rad Lab still?

Rochester:

Yes. After I left, they issued me a seminar pass which entitled me to do anything and go anywhere in Radiation Lab. It entitled me to go to the seminars, which was one of the most secret things. I had completely free access. I could even use the parts supply rooms and draw stuff out and then take it over to our place and put it in our equipment.

Work Life at Sylvania

Goldstein:

How did the work style differ? You were working on both sides — at Rad Lab and Sylvania. What were the differences or similarities?

Rochester:

There were certainly a lot of differences, but one thing that wasn't different was the commitment to win the war. At Radiation Lab, people would work very hard. If something was really needed by the military, people would work long hours and weekends. You just had to do these things. We had to win the war. We did the same thing in our shop. If we were faced with a deadline and we were having trouble with it, we developed the policy that we'd work until two o'clock in the morning. But we'd always be there at seven o'clock to start work the next day. If you didn't do that, there was no sense in working that late. We were pushing very hard, and we had a team that was really good also pushing very hard.

Goldstein:

When you were at Sylvania, were you doing research or design?

Rochester:

Design. Not much research. I had access to the people in Radiation Lab. I could go to anybody in the Radiation Lab. That was the way the Radiation Lab worked. It was all free and open inside. I knew lots of people. So if we had a question that we didn't know about or if something looked wrong, I could go to anybody in the Radiation Lab and ask him what to do.

Goldstein:

Was that a privilege of your former employment there? Or did all industry representatives and consultants have that privilege?

Rochester:

There were certainly people like me who had moved from Radiation Lab to industry. There were also cases where industry had sent teams of people to work in Radiation Lab and then moved them back to their home territory. So it was very cooperative. No antitrust was allowed. We had a war to win. We weren't going to fool around with lawyers.

Goldstein:

Were you working on the same sort of equipment at Sylvania as at Rad Lab?

Rochester:

No. It was much more general. They had the RF components that I'd been working with before, but we had the whole rest of the radar. That was a new experience to me. There was another experience that was somewhat like this case of the special diode that I'd supplied for the first sinking of a submarine. In the RAF there was some big operation going back, a plan to install one of these K-band radars, which had a lot more definition than 3-centimeter radars, on an airplane and fly it down the Rhine, 100 feet off the water and bomb out one bridge after the other. It was so low, and they would be coming so fast, that the Germans couldn't turn their guns fast enough to shoot at them. They were going to do this at night or in bad visibility, which would make it hard anyway. We really worked hard to get this stuff out to them. We shipped it off by air to England, but we didn't hear what happened. We weren't the people that needed to know the results of warfare. All we had to know was how our equipment was working.

Goldstein:

Was the environment different at the Rad Lab versus Sylvania? What I'm getting at is you were an engineer among largely physicists at the Rad Lab. Was Sylvania different?

Rochester:

I was an engineer among all kinds of people. A whole lot of them were concerned with making and designing incandescent lamps; they knew about fluorescent lamps and they didn't know anything — much at all — about radar circuits. But they gave me a whole lot of support, and we built them.

Goldstein:

Apart from specific expertise was the methodology also different?

Rochester:

Howard Biggs, who ran the Special Products Department, was very adaptable. He could do anything. It was very wonderful working for him. One thing was different during the initial part I was in the Salem plant. There was a successful union drive to organize it. Then we moved to another plant, which was actually set up just specially for the Special Products Division, and that was not organized. Later the union tried to organize that. There were two laws that passed through Congress. One had been passed under the Taft-Hartley Act. Before that there was another act which restricted the employers. The first one was a successful union drive under this previous Act. The second drive was an unsuccessful drive. The Taft-Hartley Act hadn't quite been passed. But it was going to pass, and so that changed the whole complexion of things. Originally it was impossible to defend yourself. You'd go to work, and you'd be handed this sheet of paper, and it would contain all sorts of lies about you. You couldn't do anything about that because employers weren't allowed to do anything about things. The previous act had been designed to promote unionization in a situation where employers had been abusing their employees. Then this other one — the Taft-Hartley — made it a lot fairer. We could defend ourselves a little bit. Nothing like that occurred at Radiation Lab.

Goldstein:

As an employee of Sylvania, were there situations where you would bring innovations developed at Sylvania?

Rochester:

Oh, sure!

Goldstein:

Would the flow be back and forth and equal?

Rochester:

No. More design and technical information flowed from Radiation Lab to us than flowed back because we didn't have the manpower. Radiation Lab was at that point up around a 1,000, 2,000 people, half of them professionals. We had maybe twenty professionals, very few of them anywhere near as good as the average at Radiation Lab. So the flow of information was, as it was intended to be, mainly from Radiation Lab to us. We were serving Radiation Lab and serving the armed services, doing a different kind of thing that had to be done.

Goldstein:

During wartime, were Sylvania's resources taxed to the maximum? Was Sylvania eager to take on new assignments and responsibilities?

Rochester:

Oh, yeah. Howard Biggs was always ready to do anything. He liked things that were new and different. That's why he was head of the Special Products Division.

Postwar

Goldstein:

After the war was over, did Sylvania maintain its operations in the particular technologies that it had been working on with the Rad Lab?

Rochester:

No. When the war ended, there wasn't any market for a group that could design and build model-shop quantities of radar sets. After the war we still had contracts to finish up, but there weren't any new contracts coming in. There was a salesman attached to our unit who had some responsibility to us, and he had discovered a job for us that we could really do well. This was to build the arithmetic unit of a new digital computer, Whirlwind I. It was one of the very first computers. I think we got the contract in 1947.

Goldstein:

Did MIT approach you with it, or some intermediary?

Rochester:

No. It was a different part of MIT.

Goldstein:

Yeah, I understand it was a different part — not the Rad Lab.

Rochester:

I don't know how this salesman found out about this job, but it was under Jay Forrester. They had this computer which actually had about the same power as not a regular computer chip in a personal computer, but the kind of chip that you use where you use it as controller. It has both memory and the computer part all in this chip and this is a thing the size of a postage stamp. Before a thing with similar capabilities filed a very large room as big as a gymnasium. But that was about as well as you could do. Of course modern computer chips, which are much smaller than a postage stamp — smaller than a dime — are enormously more powerful, many orders of magnitude more powerful. It's kind of amazing to think back what it was like in those days. But we did this build this arithmetic unit, and again there was a series of lectures. At MIT they're big on lectures. Someone gave a series of lectures on how the machine was organized and how you program it. Not getting too much into the grubby engineering details about how you move the power around from one place to another and how you have enough power and how the signal gets around and all the kind of thing. How it stays cool enough. I think they let me start taking this course before they let the contract. The fact that I was taking the course helped us get the contract. After I finished this course I came to the conclusion that this was the kind of work I was designed to do. It was just wonderful. I really loved it. While we were working on this project I put together a proposal to Sylvania management in which I proposed that Sylvania make a big plunge into computers because I saw this huge opportunity. Not everybody thought that way in those days, so the management turned it down. So I went over to the Sloan School at MIT — that's the business school — and I looked up American industry, and I picked out two companies that I thought were in a position to take advantage of it. One was GE and the other was IBM. I sent off letters applying for a job. GE turned me down, but IBM offered me a job, and I took it. It turns out that just before I sent this letter to GE, they'd looked into the question, and they'd decided that they weren't going to bother with computers. They weren't going to amount to anything, so why bother?

When I went to work for IBM it was very secretive. Before I went to work for IBM, I had published a considerable number of papers. I published one jointly with Dave Brown, who was working under Forrester in the Whirlwind project. Whirlwind was the name of that computer. They told me when I went to work for IBM that I'd written my last paper that would be published in a journal. IBM changed pretty quickly after that. Not so much because of me but because of the change in the environment. When I got under the security tent and I discovered what was going on in electronics. I discovered nothing was going on. This surprised me a little but not too much. At least they'd offered me a job. My boss thought it should be going on. He was really into using electronics to do the work. Almost all of the rest of IBM, except his little group, was concerned with machines with relays, gears, punch cards and all this kind of thing. IBM was making a whole lot of money in this, so it was a wonderful business to be in. Here these guys were proposing to get into electronics. So there was a problem in getting IBM started. We did a number of different things. Eventually we did in fact get the computer going, which was supposed to be IBM's contribution to the Korean War effort. That's how we sold it to management. After we got this thing out and working, it became evident that IBM had a big future in this. So then we could carry right on.

Position at Rad Lab

Goldstein:

Did you only have that one position at Rad Lab? You were there between '41 and '43. Did you change job titles or the person to whom you reported?

Rochester:

No, that was all. When my change came, I changed out of Radiation Lab. I still kept one foot in Radiation Lab.

Rad Lab and Private Industry

Goldstein:

Now you say there were others who shared your circumstances, who had worked in Rad Lab and then moved on to jobs in industry. Can you think of any examples of the companies they were with?

Rochester:

No. I mean, if this were more recent, I certainly would be able to list them off one after another.

Goldstein:

Did you have much to do with people in any groups or with RCC, the manufacturing arm of Rad Lab?

Rochester:

Oh yes. For example, the radar we built for the Navy had an indicator. So I had to work with the indicator group. And both of them had receivers because the RF head had to take this tiny signal and build it up into a big strong signal before it could send it anywhere. In an airplane the other stuff would be somewhere other than right next to the antenna, where the RF head was, so I had to work with the receiver group. There were a lot of people I had to work with.

Goldstein:

Sylvania obviously was involved in manufacturing these RF heads. Were there other companies involved in production of these?

Rochester:

Oh, yes. RCA, General Electric.

Goldstein:

While you were at the Rad Lab were they working on these?

Rochester:

Yeah. And there was very congenial relations among these people. It wasn't a dog-eat-dog thing. It was, "Let's get in and beat the Germans." Raytheon was also in it in a big way. I never visited Raytheon or GE. I did visit RCA.

Goldstein:

Were these companies manufacturing different RF heads for different applications?

Rochester:

Oh, yeah. They were in the pay dirt end of it. The 3-centimeter and the 10-centimeter radars were the ones that worked out. What happened to the K-band radars, the frequency was chosen without a lot of knowledge that was developed later. It turned out that that exact frequency, had a problem. Eventually they set up a couple of antennas on tops of hills. They sent signals back and forth between these. Radiation Lab people did this. They discovered how much rain and fog and things like that attenuated radar signals. That wasn't known until they measured it. These measurements were made while we were working. It turned out that K-band was right at the peak of an absorption band for water, so the present radars that have high definition are at a higher frequency. I think that the higher frequency was not really feasible at that time. K-band was the big step in making it smaller, at 3 centimeters and 1.25 centimeters. Now they have still smaller. If you go in and look at the display of waveguides, the K-band waveguides are the second smallest ones. The smallest ones are the ones that are now used for high-definition radar.

Goldstein:

Did any other companies — the ones who were manufacturing the S- or the X-band have postwar success with their technologies?

Rochester:

Certainly the staff they had working on radar must have been cut enormously right at the end of the war because that's what had to happen. We were spending a huge amount of money on munitions, and that stopped, so there were big adjustments all around. But that was normal because there was a war to fight, and you had to do that.

Goldstein:

Let me ask if you have any general impressions of the Rad Lab vis-à-vis industry or really any other aspect of the Rad Lab which you'd like to offer.

Rochester:

Working for Radiation Lab was a wonderful experience. Working for IBM was also great during my time of employment, IBM increased in gross sales by a larger total amount of money than any company had ever previously increased its assets. It was just a tremendously wonderful, expanding market, and it was just a wonderful time to be there and a wonderful place to be. On the whole, I feel very lucky. The class before me had trouble getting jobs. There weren't enough jobs for engineers. There was no problem at all for me.

Goldstein:

I can see the chain of events in your life: the Rad Lab to Sylvania. Sylvania got you involved with Whirlwind, which brought you to IBM. Would you say that your experience at the Rad Lab directly affected your career path?

Rochester:

It was just chance. It had a profound effect upon me permanently. Earlier I was talking about the seminars — about knowing how these people worked, these very capable people, accomplished people. There were eight or twelve Nobel Prize winners. Guys like that were my friends. They hadn't gotten their Nobel Prizes yet, so it was just a terrific experience. Wonderful for me, and great fun.

Goldstein:

When was it that you left the Rad Lab for Sylvania?

Rochester:

That was April '43. Then I left Sylvania for IBM in November '48.