Oral-History:Karl Ganzhorn
About Karl Ganzhorn
Karl Ganzhorn is a physicist who has specialized primarily in computers and communication technology. He studied physics at the University of Stuttgart, working with Prof. Ulrich Deilinger. He worked for IBM from the mid-1950s to 1986, as scientist, development manager, and director of the IBM laboratory at Sindelfingen until 1963, as head of the German, Austrian, and Swedish labs from 1963 to 1973, as Director for Science and Technology in IBM Europe from 1973 to 975, as Vice-President of telecommunication systems from 1975 to 1978, and as head of IBM Germany’s technical and scientific operations from 1978 to 1986. He also served on the German Science Board, the Federal Bureau of Standards, and other advisory panels.
The first part of the interview covers Ganzhorn's early years and education in Sindelfingen, his service in the German Army and years as a POW in French Algeria, and his university career at the University of Stuttgart. He devotes the most attention to his education in the informal university of the captured German and Italian officers of the POW camp. The majority of the interview covers his career at IBM. Ganzhorn describes his transformation from researcher to product developer to manager, paralleling both the rise of IBM’s computer industry and the rise of the research and commercial infrastructure of IBM Europe. Ganzhorn particularly emphasizes his work on the IBM 360 product line. He also emphasizes the importance of IBM’s research and production to Europe, a relationship embodied by his own service on various scientific advisory boards to the German government. Ganzhorn especially notes the role he played in fostering computer science in the German university. He discusses people he has found to be particularly influential. The interview concludes with a an especial note that the IBM 360 product line was not only important for his own career and for IBM, but also of vital importance in the general evolution of computer technology
Karl Ganzhorn
KARL GANZHORN: An Interview Conducted by Frederik Nebeker, Center for the History of Electrical Engineering, 2 September 1994
Interview #234 for the Center for the History of Electrical Engineering, 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, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. 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:
Karl Ganzhorn, an oral history conducted in 1994 by Frederik Nebeker, IEEE History Center, Piscataway, NJ, USA.
Interview
Interview: Karl Ganzhorn
Interviewer: Frederik Nebeker
Place: Sindelfingen, Germany
Date: 2 September 1994
Family Background and Early Education
Nebeker:
I'm talking with Dr. Ganzhorn at his home outside Stuttgart. This is Rik Nebeker. So you were born in 1921 in this very town, Sindelfingen. What was your father, may I ask?
Ganzhorn:
Of course, yes, my father was a weaver, in the local weaving industry. And aside from weaving, all the people in this area had their own little gardens, just big enough to live from, not to make any income out of it, but to have a little life base. Aside from that, they all went to some factory. I grew up here and went to school, went to high school here.
Nebeker:
Did you go to Gymnasium?
Ganzhorn:
Yes, I was in the local Gymnasium here, where I also finished at the age of eighteen.
Nebeker:
Were there two lines at that time in the Gymnasium?
Ganzhorn:
Yes, there were. There was a Latin line and a natural science line. I took the science line, which in a way I regretted, because I missed the Latin. Later on I felt a little sorry at not being more knowledgeable about Latin. At a later opportunity I tried to make up for it a little, but there was not too much which I could do. But I chose science because of interest; I had medical inclinations and science inclinations.
Nebeker:
Did you build a crystal radio for example as a boy? Play with circuits?
Ganzhorn:
Not so much. The same question was me asked in the diploma examination at the University of Stuttgart, when I volunteered in an extra field in communication engineering. I am a physicist, but I took communication engineering as an extra field. I didn't have to, but it was interesting for me. There was a famous professor in Stuttgart by the name of Feldkeller, and he asked me the very same question. What caused you to study communication engineering? As a physicist, you didn't need to! He asked me, "Are you fond of building radios?" I said, no, I was not! I wanted to know the principles.
Nebeker:
I see. But you were from an early age attracted to mathematics and science.
Ganzhorn:
Mathematics and to technology, but mainly mathematics. In fact in the beginning I was more inclined to switch to electrical engineering. Well, my very first idea was to build bridges. Become a civil engineer and build bridges.
Nebeker:
This was in your high school years? That was what you planned to do?
Ganzhorn:
This was in high school, yes. Then I thought, "I don't think I am the one to climb up those bridges." So, electrical engineering started to become interesting. Why I eventually chose physics has a specific reason and background; it has to do with the war, and prisonership, and everything. That changed my mind eventually. But I should mention this a little in sequence.
Nebeker:
Yes, well, I'd like to hear about that. What year was it that you received your Abitur?
Ganzhorn:
That was in 1939. Half a year before the war started.
Nebeker:
Just before the war started.
Ganzhorn:
But I had already, half a year before, even before I had the examination, volunteered for the military service. Not because I was fond of it, but because I wanted to get rid of it, as soon and as quickly as possible.
Nebeker:
So it was already compulsory, pretty much, for young men?
Ganzhorn:
It was compulsory, but only at the age of twenty. I volunteered at eighteen to do away with it.
World War II Training and Battles
Nebeker:
How long did you expect to have to serve?
Ganzhorn:
It was two years. Compulsory. But of course, right after I finished the examination I was drafted, depending on my voluntary declaration, and then I got into the war business. Unfortunately. And instead of two years, I was held up for eight years.
Nebeker:
My goodness. Now, were you drafted as a private in the Army?
Ganzhorn:
Well, I had volunteered for communication services. So I was trained as a communication technician. I was actually sitting at a radio station for awhile, and then during the war, for two years, I served in that function —
Nebeker:
A military transmitting station?
Ganzhorn:
Yes, Yes, at first during the war in France, then in Norway, and then I got back to the military academy.
Nebeker:
So you went to France to work at a transmitting station, and then up to Norway?
Ganzhorn:
Yes, I was part of the military organization then. And I worked in the campaign in France, in 1940.
Nebeker:
Was this Army communications?
Ganzhorn:
Yes, Army communications, and from that we were shifted to Norway for a year, and then I was commanded back to Germany to the military academy for officer training. And I got that officer's training as a communication officer. Then there was a switch — all of a sudden we were more or less requested to volunteer for a tank division.
Nebeker:
For the communications?
Ganzhorn:
No.
Nebeker:
Just to become an officer of a tank facility.
Ganzhorn:
So I volunteered for that also, and after a second training, again, in the tank division, I was transferred to the African theater.
Nebeker:
North Africa. When was that?
Ganzhorn:
That was in December 1942.
Nebeker:
I see. When were those battles across North Africa, with Rommel and Montgomery?
Ganzhorn:
Those battles started out in 1941.
Nebeker:
In 1941.
Ganzhorn:
1941. Rommel attacked across the east, towards Egypt, and then they were followed back by the British. Then during the entire year of 1942 there was a battle back and forth along the Mediterranean, to El Agheila and El Alamein, and all these places.
Nebeker:
And when did you arrive there?
Ganzhorn:
I arrived only in December. Our division was pulled out of Russia and was rebuilt. It was pretty much destroyed in the area of Moscow. I joined that division when it came out of Russia, as a young officer. Then the first thing which happened, at the end of 1942, was that we were moved to North Africa, to Tunisia first, to support the Rommel army which was coming back from the other side, and to screen the Rommel army against the Americans, which had landed already in the western part of the African Mediterranean.
Nebeker:
Right. When did the Americans land?
Ganzhorn:
In October 1942 they started to land in Oran, which was at the west entry of the Mediterranean. And then a German division was thrown over Tunisia to screen the Rommel army.
Nebeker:
And you were part of that?
Ganzhorn:
Yes, I was part of that. That war took half a year, against the Americans
Nebeker:
There was this famous battle that the Americans lost early on, at some pass there, the Kassarine Pass.
Ganzhorn:
Kassarine Pass, yes, I was involved in the other side.
Nebeker:
You were involved in that? My goodness.
Ganzhorn:
Yes, I was on the other side.
Nebeker:
Were you up in the front line?
Ganzhorn:
Yes. I was in all that fire, in the middle of it.
Nebeker:
My goodness.
Ganzhorn:
That was Kassarine, that was dramatic but civil. Kassarine was an easy one. We had one which was even worse, called Sid-mud-Siddah, with terrible tank fire. One afternoon when I was an ordnance officer I was with a commander, in the commander tank, and realized that sitting on a hill ahead of us was a vast plain, five miles in either direction, full of burning tanks, German and American and British. Terrible.
Nebeker:
And what was your job with that division? You were an ordnance officer, you said?
Ganzhorn:
Yes, I was what you would call an assistant to the commander. In the beginning I was one of the front officers with a company, and then I was moved to the commanding staff, and also had as a side job the tank's communication system, the broadcasting and all that. But that was a small side job only; it was all fixed frequencies, and we just turned on and that was simple. Occasionally we had to launch a message and such things.
Nebeker:
From the tanks?
Ganzhorn:
Yes.
Nebeker:
How launch a message?
Ganzhorn:
To radio. Send a radio message to some commanding staff or so on. Our commanding tanks were equipped with extra radios so we could not only talk with the other tanks in the group but also send messages out.
Nebeker:
Oh, I see, outside the group.
Ganzhorn:
Telegraphing the messages.
Nebeker:
I see. I'd like to ask you about the radio communications — just how it worked, also in the earlier period. Was that the main way that the German commands were mediated between headquarters and the companies?
Ganzhorn:
Usually not. No.
Nebeker:
They used the telephone lines?
Ganzhorn:
They used the telephone lines. As soon as we were stationary for more than two days, we had the telephone lines.
Nebeker:
And that was thought to be more secure, is that the reason? From the standpoint of being intercepted?
POW Camp "University" & the YMCA
Ganzhorn:
Yes, Yes. At the end of the African war, which was May 1943, like the rest of the German army there I got captured by a British regiment. But unfortunately they had some kind of a commitment to give some of the prisoners to the French Foreign Legion. They insisted on having part of the prisoners kept by themselves. For some reason, they picked a group of us and I was among them — and we ended up in a prisoners' camp, after a few months of moving around in southern Algeria, in the Saharan Atlas mountains, as a prisoner of the French Foreign Legion.
Nebeker:
Did they want prisoners for labor purposes? Is that it?
Ganzhorn:
For labor purposes, and also for prisoner trading purposes. So I got caught there and spent four years in that camp.
Nebeker:
Four years, from 1943 to 1947.
Ganzhorn:
That's correct.
Nebeker:
How were conditions in that camp?
Ganzhorn:
Dry. Lonely. Nothing. It was a camp near an oasis, and there was some water there; there was a little settlement around this oasis. But beyond that it was only salt. No way to escape. You had to walk at least a thousand kilometers through the desert before you could meet anybody. So it was pretty safe.
Nebeker:
What did they have you doing?
Ganzhorn:
As an officer, under the Geneva Convention of the Red Cross, we were not allowed to work under compulsion. We could volunteer to work. But I must say that was seen as — well, by the standards of today it was an almost fortunate situation that we were in. Let me describe it to you because I owe this to the American public. That's why I say it. Hear why. It was a rather isolated camp, and there were about four hundred officers in that camp, German officers and a few Italians. And the first thing we did was try to establish some kind of internal life. We had nothing to do, sitting there, looking out at the desert, day after day. Sunshine, nothing else. We were fortunate enough that we had some elder officers with us. In fact, my own commander was also in the same camp; he happened to be a prisoner of war at the same time. He had been a prisoner of war in World War I before. So he said, "Look —
Nebeker:
"I know how this is done."
Ganzhorn:
"When you go back, life will be totally different in our country, if you go back at all. But if you go back, you'll find a totally different life. The only thing which will then count is if you know something, if you can learn something, if you can use your brain. Everything else you can forget. So, the best thing is, you can use your time to try to learn something; it doesn't matter what. Keep your brain trained." That's what they were insisting on. So we built kind of a little camp university. And then a big problem was that we had no way to write. No paper, and only a few people had saved a few pencils. I was able to save a pencil about that long, which I had sewed into my belt here.
Nebeker:
My goodness. And that was a valuable thing.
Ganzhorn:
- Audio File
- MP3 Audio
(234_-_ganzhorn_-_clip_1.mp3)
That was the most precious thing I could have. Then one day in the kitchen I found a little piece of paper, gray paper like this, about that big. I had also put into my belt a little eraser, and with that eraser and that pencil I was doing mathematical exercises on this piece of paper. There was no communication with the outside world, nothing. After half a year, we were allowed to send one postcard back to Germany. And the first thing I wrote was, "I am alive, I am sound, but I need materials to write with. Send me paper and pencils." Not knowing whether anything could be sent, ever, and in fact it took more than a year before a postal path was established between the camp and the Red Cross through Spain and Morocco. Traveling that path lasted sometimes more than twelve months. In fact I got a Christmas parcel in 1946 which was meant to arrive for Christmas 1945. Nevertheless —
Nebeker:
Did you finally then get some paper and pencils?
Ganzhorn:
Yes, and then something happened that I think that I should tell you. I don't precisely know when it happened, about nine months later, some representative of the American YMCA appeared in the camp. Do you know the YMCA?
Nebeker:
Yes, yes.
Ganzhorn:
And he explained that he was sent by his organization to look after German prisoner camps in North Africa. He would like to know how the camp was situated and what the conditions were. Then we told him the biggest need we had; we had nothing to write with, we had no books, nothing to read.
Nebeker:
There were no books at all?
Ganzhorn:
So intellectually we were drying out. And then something happened and for that reason I still take my hat off whenever I see an American YMCA in the states. About six or eight months later, a few trucks arrived from the desert. Unloaded boxes, and these YMCA people, although we were their enemies at that time, went to the New York Metropolitan Library and extracted there the basic stock of a German university library. They brought four thousand books which they considered the base of the knowledge which should be available at any university. They photocopied these books. Can you imagine what that means, with the photocopying techniques of World War II? That was all with those big machines, those big light bulbs, ultraviolet lights — and each volume of course became that thick, very big. But it was readable and they shipped these photocopies to the African desert.
Nebeker:
One copy of each book?
Ganzhorn:
One copy of each book. That really served us for a year almost. And then the International Red Cross had opened channels to the German Red Cross and then we got German books from them.
Nebeker:
When you say you organized a kind of university there, does that mean there were people who gave lectures and —
Ganzhorn:
Yes. In fact there were not all active officers — the bigger part of the officers had significant professions and many academics were among them. You could learn almost anything, from Sanskrit to Chinese or the Abitur, mathematics, physics, whatever. In the beginning, they were teaching by heart, what they knew, and when books arrived, they started out with the books. So you could learn whatever you wanted. So I learned Italian first, and then started Spanish, and I looked into Latin, as I said. And of course I learned a lot of mathematics.
Nebeker:
Was there a mathematician there?
Ganzhorn:
Teaching mathematicians, also, yes. And I discovered later in my studies that I had learned more mathematics than any physicist in a German university ever would get to see. That was really helpful.
Nebeker:
And that was your main activity, of course.
Ganzhorn:
Yes, and I studied theoretical physics, of course. I also had possibilities to play music; we even got instruments finally. So life was pretty filled with such activities. Except that we were starving all the time.
Nebeker:
Oh, there wasn't enough food.
Ganzhorn:
If you can imagine having a weight of under a hundred pounds during that time, you can imagine how it was. So that was the big handicap. That also limited our ability to really learn. But otherwise, to sum up with that, thanks to these four years of voluntary studies, all around, when I finally came back in 1947, I went down to the Stuttgart University. In fact I applied at three universities because study places were very rare in those days, so I applied in Tubingen, Stuttgart and Heidelberg. And through some circumstances, or because of my history or whatever, I got acceptances from all three. So I finally chose Stuttgart of course because it was nearest to my home, and it was important then that you had a home base, because to live alone in a city after the war devastation was very, very difficult.
Nebeker:
And your family had survived here? There was no bombing?
Ganzhorn:
They'd survived here, yes. There was bombing but they were not hurt. So we were lucky in that.
Nebeker:
Why did it take until 1947 to be returned to Germany?
Ganzhorn:
Hmm. To this very day, I don't know. Mr. Schaal de Grun perhaps knows why he didn't let his prisoners get out earlier, hmm?
Nebeker:
I heard a similar story with someone else I talked to on this trip, who was a prisoner under the Russians and also took a couple of years to get back. Maybe that was also the case for the British and American prisoners of war. I don't know.
Ganzhorn:
Most of them went back a year or almost two years earlier. The British started sending them back in 1945, and the Americans did the same. In 1946 for the Western entente, most of the prisoners were back in Germany. The Russians took even longer; some of them came back only after Adenauer had negotiated back and forth in Moscow about that, after 1950. But anyway, in 1947 I finally got back, and we in that African camp actually were able to get back to Europe, because the Vatican, the Pope, had interfered. The French were claiming they had no transportation across the Mediterranean, that their ships were all sold out and they didn't have room to transport prisoners. So the Vatican rented the biggest French liner, a huge vessel, rented it for two trips across the Mediterranean, and brought all the prisoners back.
Nebeker:
Is that right! Otherwise you may have had another half-year, or ten —
Ganzhorn:
In fact, the last ones which were not on these two trips, they came back only in 1948. Still later. Of course the soldiers there were working in the mines in Algeria. They kept them as workers. Many died there also. But let me finish that story about this university, this camp university. Two things I want to mention still. When I came back and wanted to start at Stuttgart, at the University, I went to the dean of the natural sciences and explained to him that I had done some studies and would not need to start all the way from the beginning of the courses. And he was very liberal. He said, "Look, last week we had the bachelors degree examinations here in mathematics. I will give them to you and you play an honest game and bring the solutions back to me tomorrow morning. Then I will look where you are." I brought it back to him, he took a little look, only two minutes, and then said, "Why don't you apply for the next examination?" So I started my studies with a bachelor’s degree in mathematics. I took it, blindfolded, without any difficulty, of course. Fortunate, very fortunate.
Nebeker:
That was very nice.
Ganzhorn:
That I owe to the American YMCA, to a big part. They saved me some time — I would say saved me almost a year and a half at university. The other thing I wanted to mention is, I told that story once on an airplane to a reporter with the Reader's Digest. She was a little young lady, and after awhile I discovered she was reading French, so I talked to her in French. She asked me where I learned that French, and I said, "Well, at school, in French Algeria," and by that she started to get alerted, and finally she had the whole story out of me. So she suggested I should write the story for Reader's Digest, which I did. And I did it as a reverence to the American YMCA. That's why I wanted to have them print it. But unfortunately some other editor had an interest in making some money out of it. He made a story which I didn't like anymore. But nevertheless it appeared with the title "A University in the Desert Sands" in the late 1970s in Reader's Digest.
Nebeker:
In Reader's Digest. That's great.
Ganzhorn:
Of course, there were many others also mentioned in this article. I had given them names of other people who were there, and they had made interviews and all that.
Nebeker:
Now, did this appear as your article?
Ganzhorn:
No. I was mentioned in it.
Nebeker:
You just provided the information for it.
Ganzhorn:
But I came out too high, and that I didn't like. But nevertheless I got letters and all kinds of sympathy declarations from all over the world after that. I've got a copy in Arabic of the Reader's Digest.
Nebeker:
Oh, yes, of course; that's translated into many, many languages.
Ganzhorn:
A lady unknown to me from South Africa wrote to me that whenever I should come to that continent again, I should be her guest, that kind of thing.
Nebeker:
Did you keep acquaintanceship with some of those people in the camp?
Ganzhorn:
Very few, very few. That had of course some reasons also. In the later years of that prisonership, the German occupation of that camp started to get divided politically, as you can imagine. Those who had the view that Germany was on the right track, and others who were insisting on the old convictions. That split became quite furious and lasted until late after the war had ended. So many personal relationships broke apart on that. Also, people went different ways, and some of them had little particular interest in reviving these times which were kind of awful, and so on.
University of Stuttgart
Ganzhorn:
I had a few very good personal connections. In fact, I had a good friend from the same town, here from Sindelfingen, and we were both interested in the same direction. He was an elder officer already, a few years ahead of me. He was one of the main players in the big El Alamein battle, with high decorations and all that. But we were both studying mathematics and physics together in the same camp, and we then went also to Stuttgart, to the university. Then we made a special approach to studying physics as latecomers, by sharing the load. Load-sharing and time-gaining, we called it.
Nebeker:
You mean to say, working with him to cover stuff —
Ganzhorn:
Yes. We applied for two semesters at the same time for two sets of courses at the same time. One went to the one set of courses, and the other went to the other set of courses. The next week we switched and we exchanged our texts — and so we did two courses. I had a workload of sixty hours of lectures per week. But you know, during those years — and the professors later confirmed this — when we came back from prisonership, that was a generation of leftovers. At a relatively old age — I was twenty-seven then, with no profession — we had to start something and we had no time to lose. So every day was very valuable, and we were insistent to get through these studies in the minimum of time. It sounds unbelievable in modern terms. I finished a complete masters degree in physics in seven semesters. Today they take thirteen here in Stuttgart. One more year for the Ph.D. and that was it. So we got very intent on saving time.
Nebeker:
And how had you decided on physics rather than mathematics, say?
Ganzhorn:
That appeared during the camp time. At first, of course, I was interested in mathematics, all the time, but then I was looking at engineering, and pretty soon found out that the basics of electrical engineering actually is physics. Being interested in basics all my life, I one day decided I would rather go after physics. I did so, and not only physics; I even went to theoretical physics.
Nebeker:
And what did you intend to do? Go into university teaching?
Ganzhorn:
Yes, I was all geared to try a career as a theoretical physicist at the university. It was my dream — since the age of seventeen, I had the dream to get into university, to eventually get into university stuff. I didn't know that it was physics, then. I thought it was electrical engineering first, but I decided to go on to physics. And that point on, having finished physics with a Ph.D. after nine semesters, I was on a good course still. I was kind of old already; thirty was old for an academic career. But by that time I had just started half a year as an assistant at the university.
Ph.D. Work and Offer from IBM
Ganzhorn:
Then, one day, a man appeared, when I left the building at the university, and asked me my name. He knew my name, obviously, and he said he was from IBM and he wanted to talk to me. And on the street he made an offer.
Nebeker:
At first meeting? That very first meeting?
Ganzhorn:
Yes, that first meeting. He said he would like to investigate if I would be interested in building up an R&D laboratory for IBM in Germany.
Nebeker:
Could I ask you what your physics work was? Your Ph.D. work?
Ganzhorn:
Yes, that was theoretical work in solid-state physics. It was called "Quantum Theory and Electron Theory of Transition Metals," it’s hard to translate the title. It was on theoretical electron configuration of transition metals in their crystal structure. The result of it was, to explain from theoretical grounds why the transition metals — a group of metals in between iron and cobalt, chromium, in these groups — why these metals crystallize in different groups and become different crystal classes. That can be explained from electron configurations. In other words — I don't know if I should explain this in more specific terms — solving the Schweringer equation for the electrons of the transition metals is where 3D electrons are [unintelligible]. Solving the Schweringer equation for these electron compositions and investigating — let's say, four 3D electrons — what kind of symmetries are possible in the Schweringer solutions? What are the symmetry classes of these Schweringer solutions? And these symmetric classes then should be also the symmetric classes in which the metal crystallizes.
Nebeker:
Now, it was known of course what structures these metals —
Ganzhorn:
What structures... The crystal classes were known from x-ray investigations of the structure —
Nebeker:
But no one had ever derived —
Ganzhorn:
No one had ever derived it from the electron theory. The electron theory revealed that elements with one and two D electrons should crystallize in one glass. With from three to five electrons, it was cubic-centered crystal class. From five to seven was five, and six was hexagonal. I don't know it by heart now. So I could theoretically prove the crystal classes with this.
Nebeker:
That sounds like a very important result.
Ganzhorn:
It was very interesting. It came like a shot, in a matter of a few weeks; all of a sudden, the entire class of transition metals was grouped into their possible crystal —
Nebeker:
And that's because you could see these symmetries in the solutions of the Schweringer equations.
Ganzhorn:
Yes, of course. You used the group theoretical approaches. I'd transformed group-by-group theoretical metals, and on that purpose of course my mathematical background went off. I had transformed group theoretical approaches and matrices in their possible symmetries. In other words, I had transformed the matrices until they had a straight diagonal and no second diagonal anymore, and then you could derive a symmetry, possibly one class and the other class. Those were also the symmetries which were [unintelligible]. I thought then that that was a very nice beginning, and I went back to the professor and said, "Look at this way; this is the way I want to start out." And he said, "What?! Write it down, that's it!" So I got through very fast.
Nebeker:
Was it because of that dissertation work that IBM was interested in you?
Ganzhorn:
No, they obviously had learned through some investigations that I was interested in theoretical physics of solid-state metals.
Nebeker:
So they were looking just for some solid-state physicists?
Ganzhorn:
Yes, a number of people in my situation had finished a physics degree — although I shouldn't say it — but not with excellence. They were looking for somebody who was kind of promising, I'll say.
Nebeker:
Sure.
Ganzhorn:
And so they got to my name somehow, and they offered me that. I knew that IBM had built punch-card equipment and such things. I wasn't very spontaneously fond of that. But later of course I thought about it. They had indicated IBM was interested in building a competence in solid-state physics because the transistor had just been invented two years before, three years before, and they believed it would certainly play a role. They wanted to build up a competence in time, and not only in America, but also looking into the international scene for resources and such things. That's another story which I should mention in awhile. I learned about all that in Europe. But somehow I became known to them and they made the offer.
Then I thought, "This is an interesting offer, a company which is well situated, long established here, and they want to go in a new field. Why not?" Of course everybody was advising me not to do that. A good German physicist either stayed at the university or went to Siemens, nothing else. But nevertheless, I thought I should accept it. Even more so, as I asked them, "Where do we want to establish this R&D laboratory?" They said, "Probably in the Sindelfingen area." I said, "Oh, my God, that's where I come from!"
Growth of IBM R&D in Germany
Nebeker:
And so at that time they didn't have this R&D.
Ganzhorn:
They had a mechanical designing department, where they designed some punch-card equipment, additions to existing machines, others on new machines. But that was all mechanical engineering design.
Nebeker:
Well, some of it probably was at least electrical and probably electronic. They had —
Ganzhorn:
They had no electronics at that time. They had a lot of relay techniques, electrical relay — the punch-card equipment at that time was all electrical relay controlled.
Nebeker:
Just electrical.
Ganzhorn:
Quite sophisticated controls, by the way. And in fact, on the history, many of the stock programming principles, later data processing principles, had their sources in punch-card equipment. When you go carefully through it, you find many sources of the idea of how a program should be built in a machine, without ever creating the breakthrough which came from other sources later on. But it was very close. You find program steps — even the name "program step" — on punch-card plotboards.
Nebeker:
Well, I know that they had these machines with plotboards that allowed you to change what was added to what and so on.
Ganzhorn:
Yes, there they are. That's an interesting investigation — in fact I am interested in some historical work in that field, also. We have some activities going on here in this respect and you can show the developments in the late 1970s. In the electronic computer, systems architecture had already been there in the punch-card equipment under different names. It was never discovered as such.
Nebeker:
That's interesting. So let's see, we had gotten to the point where you were offered this position.
Ganzhorn:
Right. I accepted. And I started building up a laboratory. Originally it was oriented towards research, but soon, more and more, we got into at once development and product development. Around 1960, even in the 1950s already, there was a decision at IBM that the IBM European laboratories should focus on specific areas or fields.
Nebeker:
Applied development fields?
Ganzhorn:
Development fields, with one exception. The Zurich laboratory was dedicated from its very beginning, which was in the years of 1953, 1954, to basic research. That also later became part of the IBM research division.
Nebeker:
How is it that the group that you helped to build up, the Sindelfingen group, moved more to development rather than to research?
Ganzhorn:
Well, that was a strategic decision in which we were participating. I should now go back one step, namely, to the question: why did IBM even establish laboratories in Europe? There were some rather basic views and thoughts about this. One of course was, IBM was doing business with a worldwide product line, a common product line, and wanted to mobilize competent resources from all the major countries in which they were doing business. They were using different educational systems and different cultural backgrounds also in their intellectual resources which did development and research.
Nebeker:
But without any intention to specialize products to national markets.
Ganzhorn:
That was not initially intended. The basic idea was to make use of the various competencies of the world. Another reason, of course, was that IBM believed, rightfully so, that a major business on international grounds in major markets of the world cannot be maintained on the basis of imported competence. You need the full competence of an industrial company on the spot. Right in, as part of the local market environment.
Nebeker:
I can certainly understand that for scientific applications of the computer, you'd have to have people at the highest level at IBM Germany to get the computers used. Is that also true in the industrial?
Ganzhorn:
Oh, yes, but it's also true in the commercial field. Commercial customers have a different relationship to their suppliers of data processing equipment. If they have the full competence of this field readily available, in their local environment —
Nebeker:
Well, it's obvious you'd need the —
Ganzhorn:
They might not always need it, but you need to have more knowledge than you actually use. That was not the basic reason, but one of the reasons. Another reason was that during those years, there was a substantial trend in all Europe that scientists started to move to the United States. It was called a brain drain. IBM was dedicated to counteracting that, to offer interesting job opportunities for local scientists and engineers on the spot. So that they could make use of many of these talents, of different educational origins, right where they had grown up. And that in turn proved to be a valuable motivation.
Growth of Sindelfingen Laboratory
Nebeker:
Can you tell me about how it went here at the Sindelfingen facility?
Ganzhorn:
Yes, we started to build up an R&D kind of little group. There were two physicists in the beginning. In fact, I was not the first one — I was the second one who was hired. The first one was a man named Dr. Walter Sprich. He is known in the electrical engineering community as one of the pioneers of character recognition. He did early character recognition research, by that time. He was very advanced in his thinking. In fact, he launched some patents which even IBM didn't recognize as valuable assets to them early enough. Later, when they recognized it, some of the things were already beyond the deadline.
Nebeker:
Oh, so they missed opportunities to patent.
Ganzhorn:
Missed opportunities to file some of those. Twenty years later, the principles were important. He was ahead of his time.
Nebeker:
Is that work he did here at Sindelfingen?
Ganzhorn:
Yes, he did it here at Sindelfingen. Prior to that he was associated with Heinz Nixdorf in the northern part of Germany. The two together had built a little calculator, an electronic calculator. That's how he came into the field and got interested in character recognition. Nixdorf wanted to keep him, but he wanted to concentrate further on research-type work, not so much productivity. So he joined IBM and did that. The other physicist who was hired was me, and I was told I had a very specific charter to start out with. The first day, the technical director of IBM Germany let me into an empty room with an empty, clean desk, and said, "Now, please, make physics for IBM. The rest is up to you." That's the nicest kind of obligation you can get! So I could look around and find out what might this company use. And started out with several things, of course. We were following up very early on the evolution of transistor technologies, how they could be used, how they could be used for digital purposes — digital circuitry. We were looking into photo work, like optical sensing of punchcards and such applications. But very soon after — about two years — we more and more focused on electronics for computer circuitry, circuitry and memory, basic computer elements.
Nebeker:
So — designing the logic circuits?
Ganzhorn:
Designing. Yes. And we had one interesting visit in 1953, I think, from Mr. T. J. Watson, Jr., the founder of IBM. He came to Germany, and looked at what was going on. I had to show him the young efforts rapidly being built up, what had occurred. We described to him what we intended to do, and then he turned around and told the German general manager, "Give these young men what they want."
Nebeker:
That was nice to hear.
Ganzhorn:
It was nice to hear. From then on, we were given what we wanted. It was fine, and we could really build up a laboratory, totally freely, to our own views and visions.
Nebeker:
And did you get a group working with you?
Ganzhorn:
I was free to hire people. I didn't even know my limits. The only thing was, it should be good people. That was the only scale on which we were committed. So whenever I thought there was a good young engineer or physicist who would fit that group, we hired him or her, and we built up. In about eight years, the group was about two hundred people, electronic people.
Nebeker:
Altogether at Sindelfingen? Or your part?
Ganzhorn:
Altogether, it was about four hundred, but we also had taken over the mechanical design by that time already. That was merged in 1958. All the existing mechanical design work and the electronic development merged into one group, which was then called the development laboratory. Similar developments took place in England, and in France, and also in Sweden and the Netherlands. In France and England, the groups were similar in size, and in Holland and in Sweden they were smaller.
1958 Meeting & Lab Specialization
Ganzhorn:
But until 1958 we had such small development laboratories or groups; then in 1958, we all got together, all the people in charge of these groups, and I was in charge of the German group —
Nebeker:
In charge of the entire operation here at Sindelfingen?
Ganzhorn:
Yes. In 1958, I became what they then called a development manager, manager of a development laboratory. And I was also asked to integrate the mechanical design into this development organization. In 1958 all the managers of these different development groups around Europe were called together in New York. We had a closed shop meeting with our director there, the head of our subsidiary of the IBM Corporation, which was dealing with everything outside the United States. We had instituted a director for development already. In one session we agreed on a scope of responsibilities. We said, "Let's go after product development only at work, at once advanced development and product development. In various fields." So the German group, this one, offered and was also given the mission to produce small and medium data processing systems, and the corresponding semiconductor componentry, plus some input/output technology, specifically printers. So we had small systems, semiconductor components, and printers.
Nebeker:
Now, were these things you'd already done?
Ganzhorn:
Well, we had kind of oriented ourselves already in this direction. And in fact what we called the missions of these laboratories were pretty much preconceived in our minds already when we were at the meeting.
Nebeker:
But before that meeting, each of these development groups could pretty much do whatever they wanted?
Ganzhorn:
Yes, whatever they wanted. They had decided locally. But then we agreed to arrange ourselves in such a way. The British group went after larger computers. The French group went after communication-base systems. The Netherlands group had document-handling systems — check readers, character recognition and such things — and the Swedish group went after process control and control systems. So it was a nice arrangement where we had different directions of the information processing field settled in different laboratories. Each focused on one particular field. This didn't mean that nobody could touch any other field anymore. The ideas were free; it just could happen that when something nice was developed in one laboratory that didn't quite fit that mission, we moved it over to another laboratory, together with the people sometimes. They would travel there. They would fit better, eh?
Nebeker:
And do you think this resulted in much better results, of all these groups as a whole, to direct them in different ways?
The IBM Series 360
Ganzhorn:
I think so. One of the results was, of course, that in the next few years, starting after 1960, IBM developed that unique Series 360, a whole product line of computers and data. It has a historic element in it in terms of information science at large. These were four or five machines, in different sizes, all the same nature, the same architecture, and so on.
Nebeker:
Maybe we could digress here a little bit. The 360 system is, of course, one of these watersheds in the history of computing. Can you tell me briefly the role these European laboratories had in its development?
Ganzhorn:
I was just going to do that.
Nebeker:
Good.
Ganzhorn:
The British lab was developing — the model 40, it was called, of the 360 line. The German lab was developing the model 20, the lowest part of that line. We developed it.
Nebeker:
All aspects?
Ganzhorn:
All the way out until we released it to the factories. And as one example, the model 20 was released in 1964 to five different plants in the world, simultaneously. One was in Sindelfingen, one was in Endicott in the United States, one was in Japan, one was in Canada, and the fifth one was Brazil, I think.
Nebeker:
And the model 20 was designed here?
Ganzhorn:
Designed here, and also built here. Built at the same time. And controlled from here throughout its life. So we controlled the technical status of the machines in all the factories over the world. It was called 'engineering control,' which every product of that sophistication needs all the time. The same happened for the British laboratory. They controlled the 40 for a certain amount of time. The French lab went a different route. They developed modems and switching systems for PBX's, which was not part of the 360 line at that time. The Netherlands laboratory developed input/output equipment.
Nebeker:
For the 360?
Ganzhorn:
Readers, check readers, document readers for all this line. And the Swedish laboratory had some contributions in the data collection and process control at the end.
Nebeker:
This was all in some sense coordinated from above.
Ganzhorn:
Yes. The basic coordination, of course, went from the United States. So the architecture of the model 20 had to be the same, like the architecture of the other models. That was centrally directed, from Poughkeepsie and Endicott. But we were part of the group, of the architectural group, so the negotiations went back and forth on every bit we designed.
Nebeker:
And you were the director here?
Ganzhorn:
I was the director of the German laboratory at that time, yes.
Nebeker:
Were you yourself involved in the design process?
Ganzhorn:
Not anymore. By the time we designed the 360-20, the German laboratory already had four hundred people. It was a substantial group. So I had the responsibility for this project and the responsibility for a printer which went along with the 360 model 20. We also had responsibility for the development of semiconductor components. In the beginning they were not those which we developed here. The early components of the 360 line were developed mainly in Poughkeepsie. And we shared all of this development.
Systems Management & the Model 20
Nebeker:
We're returning to your own career here. That was quite a change, from when you came out in 1952, as you thought, as a researcher, to a short time later, directing a very large group. You must have liked management.
Ganzhorn:
Well, I don't know. In fact, one of the motivations why I chose to go into theoretical physics in my earlier days was because I thought that was a clean field where I didn't have to have too much to deal with people. I got trapped on that one. Ten years later, or twelve years later, I had four hundred people. And as the director it became much worse!
Nebeker:
We'll just ask about your personal feelings here. Were you unhappy to be getting away from research and giving all your time to management?
Ganzhorn:
No, not at all. In fact, I was fascinated by the fact that something must come out of research. You must be able to translate something into something productive. To see something grow, which eventually can even be sold and make money. That was a fascination by itself. I also discovered an interesting human task — to try to lead other people to achievement. That was the fascinating part which caught me and never left me, actually.
Nebeker:
How did the development of the model 20 go for you here? Were there major worries or problems?
Ganzhorn:
There was one problem, of course. Nobody ever had done such a program. So many interdependencies had to be observed. The system of this model 20 was not the one machine. It was a whole group of machines. There was a cart equipment, there was a printer, there were —
Nebeker:
Tape decks.
Ganzhorn:
Tape units.
Nebeker:
Disk units.
Ganzhorn:
There were disk units, there was a central processing unit. So a whole group of boxes had to be developed, and had to be linked together. Some of them had been developed in other places, or in the States. The architecture of the system as such had to be agreed upon, was in development while we were already trying to apply it. So there was a permanent wrestling about how we should do this. We were traveling back and forth to Poughkeepsie and to other places to negotiate bits and details of the design, so that they would be compatible with each other. So that was the governing highlight of this entire development: to coordinate a huge widespread technical development effort. It must come together and fit together — and not only fit together, but we must be able to build it all over the world and to serve it all over the world. So it had to be built in such a way that it was no longer dependent on one particular engineering group. That was a new event.
Another thing — and that needed something which we called systems management — was how to bring many, many functions together in the proper action, and how to manage this, so that in the right moment everybody who had to contribute something was there at the proper time and with the right solution and had agreed on everything which had to be agreed upon. This systems management we had never done before, and we had tough fights, in fact, inside the company, about who would do this systems management. Our American friends said, "Look, that's much too important; we can't give this away." We finally found a solution, that even Mr. T. J. Watson, Jr., helped us with. He said, "You have to learn it once. We will give you an experienced American systems manager, and bring him to Germany. He leads the group in Germany, but we do it for Germany." By their assigning one man here, a Mr. Wooding, he came here and became one of our best friends, and he really not only led this systems management from here, but also trained some of our most capable engineers to do this.
Nebeker:
Approximately when did he come here?
Ganzhorn:
He came here in 1963 or 1964, until we released the system to the factories, and then we took over. And the following generations of systems we developed all by ourselves.
Nebeker:
I can see that there are major concerns in developing part of this whole line of 360s — everything has to be compatible. I don't quite see how it affected the design development process, since it was being manufactured in five different factories. Did that cause major problems? Did you have to take into account that the same operations would have to be done in Canada or wherever?
Ganzhorn:
Yes. First of all, when you manufacture in different countries, you have different resources for supply components. And that already needs a careful understanding of the system so that these things don't become incompatible. But, more than that, in the early release phase of any such technical system, before it goes onto the market, there are so many changes, almost day by day, that it takes a major logistic effort to control these changes through the factories, so that the engineering control levels were always the same. That's a really military type of effort — the logistics.
Nebeker:
I see. So if you had your design group in the same town as the factory, then it's of course much easier to organize all of that.
Ganzhorn:
Yes. Right. This was the first time that a product was released from Europe to Japan, to a Japanese factory. We had Japanese engineers here, and one of them was sitting all day in our technical meetings where technical issues were discussed and decided upon and developed. We went home; he went back to his office here in the laboratory and got hung up on the telephone with his Japanese friends during half the night, and translated all of that to the factory. I must say, in honor of the Japanese, the smoothest cooperation occurred with the Japanese factory, in spite of that distance.
Nebeker:
And all of these machines coming out were essentially the same machine, from these five different factories.
Ganzhorn:
Yes. Basically the same machines, with the same instruction sets, with the same types of functions.
Nebeker:
Perhaps different individual components?
Ganzhorn:
Of course, different — lots different — signs, different characters in the printers and all that. But, functionally, they used the same programs, and basically, a customer engineer in Japan should be able to maintain a machine in some other country also. Except for the language-type things, which were not the same, of course. For everybody, that was kind of unique, to do such an international-type technical development. Of course, you can think of many products which were developed in one country, later transferred to another country, and then translated and rebuilt in that other country. But from the beginning, when you release it to a laboratory, when the factory sees it for the first time, that's a totally different approach.
Nebeker:
And simultaneously released to five places —
Ganzhorn:
Yes, and simultaneously, has to build up in each.
Nebeker:
Well, that's very impressive. And it worked.
Ganzhorn:
It worked. We sold more than twenty thousand systems that worked.
Nebeker:
Well, you said the Japanese factory or the Japanese production went the smoothest. Were there any particular problems you'd care to mention with others, or even with that one? What were the very difficult points?
Ganzhorn:
Let me put it a little ironically — we had more work to get our friends in the Sindelfingen factory in the same shoes. They were very close to us.
Nebeker:
So it looks like at the same time this is going on, in 1963, you were made the director of IBM Laboratories Germany, Austria, and Sweden.
Vienna Group and Language Definition
Ganzhorn:
Shortly before the model 20 or 360 line was released, we reorganized the European laboratories — in terms of management only — and I became in charge of three laboratories. I also had to take the Swedish laboratory and a small laboratory in Austria which, I must say, I had founded together with a man from Austria in 1958 already as a subgroup of the German laboratory in Vienna.
Nebeker:
Oh, this was your initiative to start another group in Vienna?
Ganzhorn:
Well, not only mine. I knew there was a very good man and a very good group. The leader of the group was an engineer at the technical university of Vienna. His name is Heinz Zemanek. I knew him before, already, and I talked to my director in New York, and said, "Look, there's an interesting group in Vienna." I then made sure that I could get a hold of them. They said, "Well, try it." So I got in touch with Heinz Zemanek, and he had other bridges also to IBM already. So he was inclined to consider this, and at the end, the so-called Science Group Vienna was a subsidiary of the German lab. But very soon we said, "No, look, that must be a lab on its own." So we made it a very small but effective Viennese laboratory. Heinz Zemanek became the lab director there. But I was then given direction for the German, for this Austrian group, and for the Swedish laboratory in 1963 already.
Nebeker:
I see. And it was, I assume, still the case that each laboratory had its own mission?
Ganzhorn:
Yes, the missions remained, essentially. In fact, they have survived for a very long time. The German lab still does the same today.
Nebeker:
And what was the mission of the Austrian group?
Ganzhorn:
Initially they had some hardware, there, also. Their mission was speech recognition. They did some speech recognition work. Not actually recognition — it was called the digitizing of speech, or something like that. They could digitize voice, store it, and record speech from pre-stored digital information. That was called a vocoder, which translates digital information into speech. That was initially done in Austria, and on that basis IBM developed a product, which is not a recognition device. It was a —
Nebeker:
Voice synthesizer?
Ganzhorn:
Yes, a voice synthesizer, something like this. We had it on the market for a few years. Well, you had digitized speech stored, and you had voice answers from the machine, from this storage. So it was not recognizing the speech.
Nebeker:
Right, right. It was producing the speech.
Ganzhorn:
That was the only part which was developed in Austria, still. They previously had developed an academic-type computer called Milestone but that was before IBM's time. Of course we were aware of that, and we knew they were good engineers. That's how we got alerted to this group. But then they focused on software. They were given a very interesting special task. IBM in the early 1960s developed a new programming language which was called PL1.
Nebeker:
Oh yes, I remember that.
Ganzhorn:
Programming Language Number One. The Vienna lab was given the very difficult task of establishing a formal definition of this language. Formal definition. So that in mathematical terms, each turn of this PL1 language could be described in formulas. So it was an exact definition. They were able to do this, but it turned out to be so complex that it was not useful at all for practical applications as a definition of the language.
Nebeker:
What was the intention of doing this task?
Ganzhorn:
The problem with all these artificial languages through the 1960s up to today, is that a language is described in terms of vocal information, normal text-type information, and this is always a little indefinite —
Nebeker:
Ambiguous.
Ganzhorn:
Not ambiguous, but not fully precise.
Nebeker:
Not precise.
Ganzhorn:
Subject to interpretation. And then, the informatics people — the programmers, in the earlier days — took this definition and wrote a translator or a compiler, we called it, which compiled this into a machine language program. But the translation — the definition of this compiler was essentially the understanding of the programmer which he had in his view from the text which was right there. So in the long end, the compiler defined the language. What came out at the end, and what the machine was doing, that was the definition. To a certain extent, that is still the way an artificial language is defined. We felt that was not a good way. Of course, information technology today has means to do much better definition. But the real formal mathematical definition that we had a definition in the mathematical sense — this and nothing else! — that's very difficult to make. It could be done, but it turned out, if you do this, this PL1 language turned out into a book of three hundred and sixty pages. It was not handy at all anymore. It was very difficult to read. So it was very soon recognized that that's not a way which can be applied in practical terms to define a language.
Nebeker:
I see. And it was the Austrian group that did that work.
Ganzhorn:
The Austrian group did that, but that competence was developed on that occasion, and it later entered the computer science under the terms "Vienna definition of language," Vienna definition method.
Managing Three Research Laboratories
Nebeker:
How was it, directing these three laboratories? What were your tasks in those years from 1963 on?
Ganzhorn:
Well, number one, basically and principally, we were responsible for anything which happened. But that responsibility also rested with the local lab director. So my task as a director of several laboratories in Europe was to sit down with the lab director and his managers to negotiate what they can do with their competencies, how should they develop their next operating plan where they defined their new programs. It was a give-and-take in many sessions, discussions; who should do what? How can we divide the work between different laboratories?
Nebeker:
I see. So it's a case of greater control, greater management of the national laboratory.
Ganzhorn:
It's a planning-type operation, and it's a control-type operation. Each time I went there I reviewed the programs and went through each program with individual program managers. Where do you stand, what are your schedules, what are your costs? What are the new events which might have to be changed, which might cause changes in the program? Such things. Eventually all these parts were linked together on the European level in one total European operating plan.
Nebeker:
Was there a new director, then, of the German laboratory here, under you?
Ganzhorn:
Yes.
Nebeker:
Was there a single director for IBM Europe?
Ganzhorn:
Yes. He was sitting in France. He chose to have his headquarters in Nice, where our French laboratory was located, so he was close to one of the laboratories.
Nebeker:
And then under him there were —
Ganzhorn:
Under him we had two directors.
Nebeker:
Two directors.
Ganzhorn:
With three laboratories each. The other director was a Frenchman. He had the French, the English, and the Netherlands laboratories. And I had the German, the Swedish, and the Austrian.
Nebeker:
It sounded like before this 1958 meeting there wasn't this kind of structure.
Ganzhorn:
No, no. There was a very loose kind of effort in every country to build up a technical competence. They were pretty much free to choose where they'd try it first, in which direction. More and more our American organization became alerted to these ongoing efforts. In 1959, for instance, we for the first time organized a technical symposium of all our European laboratories — here in Stuttgart, by the way — where we presented all the European development efforts. We had invited many of our American colleagues, whom we knew, of course. That way we were gradually developed into a resource, and in 1958 the corporation decided to include us into their working force as a token.
Nebeker:
What coordination was there with the research and development going on in the U.S.?
Ganzhorn:
There was very tight coordination. In fact, the German lab aside, we had very close connection with the laboratories at Endicott, and also with laboratories in San Jose, California. We were in close cooperation because those two laboratories were developing other parts of our systems. Such as printers: the entire card input/output equipment which was linked to the 360 model 20, came from California. So we had California engineers here as coordinators, to bridge this. In fact, in IBM it turned out that every laboratory outside the United States had about ten percent of foreigners all the time on their premises for coordination purposes.
Nebeker:
To return again to the 360 model 20 development, was it the case that there was no major failure, no unit that failed to function as it should?
Ganzhorn:
In the 360 line?
Nebeker:
In the 360 model 20. I mean, you're trying to get it to work, as you just said, with card readers from San Jose —
Ganzhorn:
Of course there were many detailed, technical problems —
Nebeker:
I'm sure there were problems along the way, but everything got solved without too great a delay?
Ganzhorn:
As I said, pretty soon the product was flying, was very well accepted on the market, and we sold more than twenty thousand systems. Which was a huge number during those days. In fact it was the most widespread computer of the 1960s.
Nebeker:
What were the tasks of these three laboratories you were in charge of, in the years following this work on the 360?
Ganzhorn:
The German laboratory essentially went on the same track. They still build systems today, except they build systems today which are much larger. Reach up higher in the line. The German lab also came out in 1967 or 1968 with the first integrated circuit in IBM, a special circuit for line switching. Then we went into the major product line in semiconductors. Some of the early memory chips were developed here. A most famous one was a chip which had two thousand beta at that time, which was the highest number we could get, and that chip went into the trial 370 line; that was successful. The trial 370 memory was equipped with that chip. Also, my charter in 1958 was to build up a semiconductor competence with a view that we might be able one day to build a factory for semiconductor components here in Germany. That happened. In the late or early 1970s — I'm not clear about the timing now — there was a semiconductor factory built in Burbingen not next to the laboratory, two miles away. So there was a match between components and manufacturing.
It so happened that some of the developers' semiconductor components which were developed in Burbingen went into production in Fishkill or in Burlington in the States, and the German semiconductor plant took developments from the United States and built them here. That depended on the resources. A plant must be managed according to different rules, as compared to a laboratory. A laboratory develops one generation after the other. A plant has to be managed in such a way that it is always loaded, utilized in the best possible way, so they bring in whatever they can. That's why we never had a one-to-one match on products. That wasn't all foreseen that way, but the technical competence was an offspring of the laboratory in many ways.
Nebeker:
I see. You didn't have charge of the manufacturing plant.
Ganzhorn:
No. When the manufacturing plant came on, the manufacturing organization was a huge organization — still is — in Europe itself. By that time I had so many international responsibilities already that I couldn't.
Nebeker:
This PL1 work, I suppose that was done at the Austrian laboratory.
Ganzhorn:
The commission work was done in the Austrian laboratory.
Nebeker:
In the late 1960s?
Ganzhorn:
Synchronous to that a compiler was developed in the British laboratory. They had worked together, of course, and more and more during those years, the software development spread and grew up very rapidly. So they added another department in Germany, also, as a programming department, which today is the biggest part of the developmental source.
Nebeker:
Could we digress at this point? I'd like to ask you to summarize how it's gone for PL1 since then.
Ganzhorn:
Well, I'm not a specialist on PL1 markets. We had it on the market for a number of years, and it was used. It was essentially a combination of FORTRAN and COBOL, to bring both together, and as such it was used in many places. Today I think it's only still used where those programs exist. But I don't think it's even maintained anymore, these days.
Nebeker:
How long was it a product that IBM was pushing? Do you recall?
Ganzhorn:
Oh, I'm very uncertain about this. I will not say a date.
Nebeker:
Okay. What was the Swedish laboratory doing, in the middle and late 1960s?
Ganzhorn:
They were doing what we called then 'data acquisition.'
Nebeker:
The same work that you described earlier with the —
Ganzhorn:
Fading away a little from process control towards data acquisition, how can data be acquired? Input-type work. How can it be transmitted? And at the same time, they also started some software work.
Nebeker:
What sort of work?
Ganzhorn:
At one point in time, they were — let me see, what was it? I'm a little at a loss for this. At one point in time, they were at PL1 also, part of that. During my time, at least, they had part of PL1 development also. They also had part of the Iowa-type compiler, but that never flew, really, very few customers on that. But the scientific community wanted it. They did a number of compiler works, technical languages.
Value of Bottom-Up Competition
Nebeker:
I wanted to ask you what it was like, doing the job that you had to do there, which was to oversee and direct the work of these three laboratories. Was there a continual friction between what the company as a whole would like to see, and what individual labs wanted to do?
Ganzhorn:
Well, there was not so much friction — it was more like competition. Which laboratory could make the best possible offer, in terms of a product development program?
Nebeker:
So there was a lot of a sort of bottom-up proposing, "We'd like to develop this."
Ganzhorn:
Yes, that's correct. Mostly bottom-up. Bottom-up competition.
In fact, sometimes we had discussions on whether we should give them all instructions on what to do. But the subject is so complex that we recognized time and again it's much better to let it grow bottom-up and make a choice.
Nebeker:
I see. That's very interesting. But there must have been some areas that got left out, and the company knew, "Well, we've got to push this kind of software, or we've got to — "
Ganzhorn:
Oh, yes, there were many discussions among the development management, that we should do this or that, or here was a product area where we were weak, and we should get something started. And the moment that such an argument came up, one of the other laboratories of course became aware of it. They were all aware, essentially, of the entire product line — that, in fact, by the way, was one of the very big strengths and attractions for IBM for engineers, namely that we had a totally free information flow inside the company. You could call any engineer, day or night, here and in San Francisco, anywhere in the world, and ask him what he was doing. He was not committed, but it was common sense that information flow would be free and open. In spite of the fact that often they are competing among each other for the same product areas and for new product proposals. Sometimes we lost, and sometimes we had a chance and did make the best proposal which fitted the company's imagination, how to approach a market segment, and then we got the development charter for it.
Nebeker:
Now, when you say that there's a competition, is the incentive for wanting to win that competition that the national laboratory would have greater resources for hiring people?
Ganzhorn:
Yes, to a certain extent, we had to, not only to justify it but to make up for our operating plans for the next years. If we didn't have attractive programs, there was no growth.
Nebeker:
I see. I mean, one might imagine, at least with some R&D groups, that they would be happy to do what they were interested in, as long as the company as a whole would permit that.
Ganzhorn:
That was not good enough. You had to make your work interesting for the company and useful for the company, also. But of course, there were areas, we also always had at once development areas in the laboratory where people were free to develop ideas. But the lab director had to make sure, all the time, that he had interesting programs which were of value to the company.
Nebeker:
Did you ever feel that your task was to tell the lab director at one of these laboratories, "I think a lot of that work is not particularly valuable to IBM”?
Ganzhorn:
Oh, yes. Well, sometimes it went all the way to the dismissal of that director. They can't continue.
Nebeker:
So there certainly was a good deal of control from above.
Ganzhorn:
Yes, the challenge, I would say, was not only to control what was going on, but also doing a permanent investigation and challenging, "Is that really the best the company can have?" So it was very challenging, and sometimes — today I can say this openly — the competition among the laboratories to do the best job was so fierce that outside competition had no chance to win against that race. Those were the good times of IBM.
Nebeker:
Do you think that was a conscious policy, to have this internal competition?
Ganzhorn:
It was a conscious competition, actually, very conscious. But it was controlled competition, let's put it that way. We had the competition up to a certain point, when the ideas were brought up, compared to each other, compared to what this lab could do and by what time, and what the other lab could do. Where the different principles, different new ideas, which had higher potential on one side — which was the more progressive approach?
Market Timing & Technological Breakthroughs
Ganzhorn:
When you say "progressive," it's a very delicate thing. You cannot be too progressive, because then you are too annihilistic in time scale, because you don't have all the technologies available which you need at a certain time. So you can only have a limited amount of novel-type innovation in a certain time frame. Otherwise you will not make the market. The biggest example of course is the 360 itself. The entire 360 technology was based on individual transistors, on modules so-called, and the circuitry was printed circuitry. It was called solid-logic technology. Later on we called it — what was the name? — high-point technologies. Where we had components added to the module. We came out with that technology in 1964; in 1959 the first integrated circuitry had already been demonstrated.
The entire electro-technical world was laughing at IBM when IBM came out with that old-fashioned technology. Silk-screening, a printed circuit, and a module of this type of technology, inserting the transistors, and having resistors on top which could be calibrated by sandblasting, calibrated the transistors. That was such an old-fashioned technology. The entire world laughed at it. The integrated circuits were at a situation where the yield from the production lines which were trying to be built up were either two percent or three percent or five percent or zero percent. You never knew, it was totally unreliable. Not prepared for frequent use at all. No base for verified production for use in the unknown. Impossible. IBM had to choose a technology which they could produce in 1965 in mass production. Controlled to their specifications. That was possible with this old-fashioned technology. And while the world laughed, we sold the machines by the thousands on this technology. In 1968, the breakthrough in silicon technology appeared. But the silicon processes became reproducible and controllable in the manufacturing lines. It was precisely in 1968.
Nebeker:
There was a sudden breakthrough in that technology.
Ganzhorn:
It was a very sudden breakthrough, that all of a sudden, these two processes in the manufacturing lines could be controlled in such a way that, at the end of the line, after two hundred steps, there was still a yield, a reproducible yield. And then, of course, the entire world of electronics changed. IBM was the first one at that time to change. By that time, we had a whole line of 360 machines — thousands and thousands of machines — out in the field, with magnetic memories, and in 1968 when we saw that integrated circuits could be reliably manufactured, the company immediately switched the policy and said, "From now on, we will no longer develop magnetic memories; we will go after integrated circuit memories." Silicon-type memories. It was a very early decision. That made the 370 line fly. The German lab was in the middle of this, with one of the major chips, the first [inaudible] chips.
Nebeker:
And that decision, in retrospect, was made at the right point in time, do you think?
Ganzhorn:
Was made way up.
Nebeker:
But I mean, in time, that was the correct time to decide to go to the integrated circuit? It couldn't have been an earlier successfully?
Ganzhorn:
No, it was the first time this became clear that this was a way to go. Before, we never knew; would it take another five years, another twenty years, before that silicon could be controlled? Before it could be manufactured? Nobody knew that.
Nebeker:
But if someone in 1966, say, had foreseen that we're just about to be able to produce these chips reliably and started designing computers and all the circuitry, maybe they would have had a jump on IBM?
Ganzhorn:
In 1966 it was in already, and for the logic of computers — the way was already going, although nobody could produce it in mass production yet. But many people believed — we all believed — that the silicon route should be followed.
Nebeker:
The question was when.
Ganzhorn:
The big question, a question which the manufacturing people wanted to know, and the marketing people wanted to know, was: "When will you be able to ship?" And until you could answer this question, a new development could not be accepted in a factory.
Nebeker:
Are there examples in your experience where IBM made too large a jump, tried to move too early to some new technology?
Ganzhorn:
In this company we saw jumps which were reaching too far, in the sense of one bridge too far. I would say, during my time, I am not aware of a strategic decision which reached too far. Of course, there were a number of products where we tried something and undertook something and it turned out it was much before its time. One of the very prominent examples comes from software, a software issue. In 1955, I heard a presentation which later on was also implemented in one machine, the so-called NORC, the Naval Ordnance Research Computer for the United States Marines, I think. There was a software device which was called a look-ahead device. It was called a device although it was a software principle. How can you write a program which looks ahead on its own work? We had a program consisting of many program steps, and the machine looks at it, several instructions before it does it, in order to find out whether it should rearrange it or so — some kind of its own work plan. So this look-ahead function is kind of a basic principle, and was tried out in the NORC first, as a first approach; then it was not forgotten, but pushed back for a long time. It only occurred about twenty years later in the more modern programming systems. Such things happened many times. But early approaches had been made.
Of course, there were also many approaches which were undertaken three times and it turned out the time was not right or the technology was not supplemented by enough different things which were needed. A prominent example is the entire low-temperature technology, or actually, technologies. IBM made three approaches over the decades to do something with low-temperature physics. Each time we had to stop it again because it could not be implemented for a technically viable product. But there are good technical reasons why, and today we know why. This was apt not to be competitive against this powerful silicon technology. One must recognize when one talks about electronic development of the last four decades that there is a unique event — namely, that's silicon. That silicon fulfills all the positive hopes which you can possibly put into an electronic material, more than anything else. It's such a coincidence of positive effects that no other technology ever had a chance to win against it.
Nebeker:
Are there examples of the reverse, in your experience at IBM, waiting too long to move to some new technology?
Ganzhorn:
Well, of course, I'm probably not the most qualified man to talk on that. But the fact that IBM during the last six or eight years has had to reduce its forces and lost market share, as we say, does have reasons, which were not marketing-type reasons; they also had reasons in the product lines. Certainly IBM has probably missed some of the opportunities which were there. It's commonly known and can be read in every newspaper that IBM has much too long focused on large mainframes and such things. IBM took steps on the personal computer (PC), not the first ones, but in time. But unfortunately, when IBM should have taken the second step on PCs, it has not. Others took that step earlier.
Nebeker:
What was the involvement of your laboratories, the laboratories you were directing or overseeing, in the development of the PC?
Ganzhorn:
Not too much, not during my time anymore. The first PC in IBM was developed in the early 1980s, and I left the company in the mid-1980s.
Nebeker:
I would have thought that it was maybe 1980 that they came out with their first.
Ganzhorn:
Their first one, Yes. But the German lab was not involved; that was done in the United States.
Nebeker:
Perhaps we can cover now your subsequent appointments.
Ganzhorn:
Yes. There are some other interesting things to discuss.
Dir. for Science & Technology, IBM Europe
Nebeker:
Yes. In 1973 you became Director for Science and Technology in IBM Europe. What was that position?
Ganzhorn:
That was a position overseeing the technical and scientific resources of IBM in Europe. That was not a line responsibility except for the scientific centers, which were small groups, software groups, in many countries — France, England, Germany, Italy, Israel.
Nebeker:
Separate from these national laboratories.
Ganzhorn:
Yes, they were developed later when the software development became viable. I had to oversee those; I also had to oversee all the technical resources in terms of advising the president of IBM Europe where to focus, where to put something, where to develop, where not to develop. What a country would need in order to be competent enough to maintain its market in the local country. If there is a good scientific resource — like in Israel, for instance — how can it be made useful to the entire organization? Such things. But I didn't have line responsibilities in terms of that I was responsible for everything which was going on.
Nebeker:
I see. I meant to ask you earlier how IBM Zurich fit into this system of R&D.
Ganzhorn:
IBM had always a research division, since the early 1960s. And that research division was totally separate from the development laboratories. In the ratio of, I would say, nine to one, development forces, and research forces.
Nebeker:
Was Zurich the only?
Ganzhorn:
Oh, no. The main research laboratories were in the United States.
Nebeker:
No, I meant in Europe.
Ganzhorn:
In Europe they were the only one. One must also say that this became a little fuzzy, because the scientific centers did a lot of research work in the software field. But Zurich was the only research laboratory which actually did physics, solid-state, and technical work — communication-oriented work.
Nebeker:
I see. What were the relationships between IBM Zurich and these other IBM laboratories in Europe?
Ganzhorn:
In Europe? Oh, we were friendly, and we exchanged our information and programs.
Nebeker:
Was the work at Zurich tied closely?
Ganzhorn:
No, no. Occasionally there were efforts to transfer one of the researchers out. We had some of this work also done — in the communications field for instance — where work from research was transferred to a development laboratory. That was done.
Nebeker:
But it wasn't any kind of close coordination.
Ganzhorn:
Not a forced cooperation, determined cooperation. In fact, Zurich was in a way held as IBM's most free research resource. They could very much do what they thought would be interesting for IBM. I recall a comment which the chief scientist of IBM made, Dr. Piori, when the Zurich lab was established. He said, "We want to have a small group in Europe which does basic but high-quality research. Only the best researchers which you can find in Europe should be there. Quality is number one. It should be high up. And if out of this laboratory, in about thirty years, comes a Nobel Prize, that's fine. That's good enough. Then it fulfills its purpose." He was dead right. It was exactly thirty years when the Nobel prizes came.
Recognition of IBM Europe as European
Nebeker:
Is there anything that comes to your mind, when you think of this period, 1973 to 1975, when you were the director for science and technology in Europe? Any particular event?
Ganzhorn:
Event, I wouldn't say, but an effort which was remarkable: during those days, there was a strong effort from the European Community to establish itself as one of the major world economic forces. We made all the efforts to demonstrate that part of this European economic resource was IBM in Europe. During these years we had many meetings and discussions with Brussels, and successful meetings, and during those days — largely thanks to Jacques Maisonrouge the president of IBM Europe then — IBM Europe became a European entity. It was respected as a European-type entity which connected or translated the computer to the European community. I think that was not a one-time, new situation, but during those years, in my mind, that was the most remarkable thing which happened, that IBM was recognized as an intrinsic European scientific and technical resource.
Nebeker:
I suppose that outsiders might always assume, you know, "IBM Germany or IBM Japan, oh, that's this U.S. company."
Ganzhorn:
Or they are considered as a sales subsidiary.
Nebeker:
Yes. But what you're saying there was that it became clear to people at large that IBM Europe was a European operation with its own identity.
Computer Science in German Universities
Ganzhorn:
Yes. In fact there are many events which underline this. You were asking me about my own career. I was one of the first — and probably the first in Germany — who gave a university course on computer systems.
Nebeker:
Oh, is that right? Where was that?
Ganzhorn:
At the University of Karlsruhe.
Nebeker:
When?
Ganzhorn:
Started in 1960.
Nebeker:
My goodness. And what was the course?
Ganzhorn:
Basic Principles of Data Systems — Data Processing Systems. That's a translated title, but the content of the lecture — I gave this lecture for twenty-seven years. Not the same lecture, of course, but the same contents.
Nebeker:
But the same course?
Ganzhorn:
Essentially the same title. And in fact, when you look about the buildup of the competence in data processing, information processing, in Germany in the scientific community — of course, there were some scientific institutions that I'll always mention — but by far the biggest contributor was IBM Germany. Already during the early 1960s, IBM Germany had a view that they must not only be present, as a scientific, technical development and marketing and manufacturing resource in Germany, but they also must be known as such, and they must be actively involved in the local community. So very early in those years, we were already trying to encourage some of our people, for instance, to take university charters and also to accept university chairs. Sometimes I very much encouraged, to say the least, some of the members of the laboratory to leave the lab and to accept a chair at the university.
Nebeker:
Maybe this is a good point to digress a little bit and talk about computer science or the academic side of computing, and how that developed in Germany, and, of course, your involvement with it.
Ganzhorn:
I was considered as one of those who had helped to establish computer science in Germany. In fact, the German government has given me a high decoration for that, even.
Nebeker:
You said, of course, this course that started in 1960 was —
Ganzhorn:
Was one contribution. But also, in 1968 and 1969, the German general manager of those decades and I were both promoted the same day into the German general management. We worked together for over twenty years, very closely. We made several basic recommendations, based on very well taken studies, to the German government, on what to do to establish informatics in Germany. That helped a lot to get it off the ground.
Nebeker:
This is a huge topic, I know, but maybe you could summarize your recommendations and what in fact happened.
Ganzhorn:
We recommended that in all technical universities and many other universities, spread over the entire country, a competence in computer science and programming must be built up to teach that as an engineering science. And we had given numbers, how much and in what directions; we had described the contents of the lectures, what should be taught at universities, to what degree, in what volumes, and we made a definite promotion program of recommendations to the government. Those recommendations in fact were also accepted by the government. So we were instrumental, to say the least, in helping the German scientific community to build up.
Nebeker:
Who was this other physicist?
Ganzhorn:
The general manager.
Nebeker:
And his name?
Ganzhorn:
That was Walter Boesenbauch. He was IBM Germany's general manager from 1964 to 1983 or so.
Nebeker:
And the result was that informatics became a specialty in technical schools?
Ganzhorn:
We were not the only ones, of course, who were pushing. But we, as the biggest market representative in this field, of course, had a voice. It was heard, in spite of the fact that we were an American subsidiary. But we were considered part of the community. In fact, I've served for two decades on advisory committees of the government. Later on, in the 1970s, I even became a member of the German Science Council, which is a council of twenty-two German scientists who advise the President of the Republic directly. I was there for three terms.
Nebeker:
So you had then informatics established as a specialty.
Ganzhorn:
As a discipline.
Nebeker:
So there were institutes for information processing, data processing?
Ganzhorn:
Chairs for computer science, for information processing, for technologies, and so on.
Nebeker:
And you said this was more than at the technical universities?
Ganzhorn:
It was not limited to the technical universities, not at all. For instance has an informatics faculty; Tubingen has one now. Most universities these days have something like that. But we pressed very hard, that this should be established as an academic discipline. That was the idea.
Nebeker:
Now, especially in Germany, there's always, I believe, been this distinction between the engineering disciplines and the scientific disciplines.
Ganzhorn:
And humanities, Yes.
Nebeker:
The things that are taught in the technical universities. How did informatics fit into this academic structure?
Ganzhorn:
It did not fit; it was spread over both. I would say it was heavier on the technical universities, of course, but not limited to them. Tubingen for instance had very early efforts.
Nebeker:
There were also of course debates about the subject matter of computer science or informatics — how much mathematics, how much?
Ganzhorn:
Oh, that's a long debate! Like in your country, also. The mathematicians were always claiming it was a subset of mathematics, actually.
Nebeker:
And the electrical engineers wanted it to be a part of electrical engineering.
Ganzhorn:
That's right, and eventually the clever computer scientists said, "No, information technology or informatics is neither one." Today mostly there are informatics faculties.
Nebeker:
So not within electrical engineering and not within mathematics.
Ganzhorn:
Oh, yes, there were only one or two universities which kept it together, both with electrical engineering. Erlungen and Darmstadt. Those are the two which still have it together with another faculty. Which is not bad, by the way, because the worst thing a country can do is make all universities the same.
VP of Telecommunication Systems
Nebeker:
Maybe we should return to your career with IBM, unless there's more you wanted to say at this point on that matter on establishing informatics.
Ganzhorn:
Yes, we were touching it when we talked on the European situation, in which I was involved from 1973 to 1975 as Director of Science and Technology for Europe, particularly. It was a really heavily intended approach, to establish IBM as part of the European Economic Community, in the European market. You might ask why I stepped out in 1975. That was due to a specific event. By that time, all of a sudden an old friend of mine, Mr. Evans, Bob Evans, the IBM vice president at that time, came to Europe. Some of the laboratories reported to his division in the later years. By the way, I should say in the late 1960s, more and more it turned out that we must associate European laboratories to individual product divisions in the corporation. IBM had become so big that we had a division for communication products, a division for large systems, a division for small systems, and so on. Then came this anti-trust type effort where we split the company for awhile into two groups, and one was a general business group. The other was a data systems group, but that was removed once the anti-trust thing was over. But the split remained, unfortunately, to a certain extent.
During those days we associated the laboratories to individual product divisions — not organizationally in the beginning, but technically. And also budget-wise, more and more the budgets came from those divisions directly. In this way, we could control better which money went into which part of the product line. That was very reasonable, and that was also one of the reasons why in the beginning of the 1970s we dissolved the European laboratory organization, and linked each laboratory directly together with a division. In 1975, the president of the communications division of IBM, the division which built all the communications-oriented products, from modems to multiplexors and line switching equipment and all that — a large product line — came over and said, "I need you as a vice president of telecommunication systems." So I took that job, and that was a job out of White Plains, New York. So, again, the question was, should I move to White Plains? And on that job I had responsibilities for the French laboratory, which was communication-based, and I maintained my responsibility in Germany. That was one agreement from the beginning, that I would maintain my position. In 1963 I became a member of the general management at Germany, and that's a legal position in Germany for IBM. I was an officer of IBM Germany, which I remained until retirement in 1986.
I maintained that, and along that line, I also maintained the personnel and the local responsibility for the German laboratory, inside the German organization. So development in Germany always reported to me. No matter the division they were linked to. But, at the same time, I had this vice president job from the United States, and the question was, should I move over there? I said, "Look, with a laboratory in France, one in Germany, with all this in Raleigh, North Carolina, in Kingston, New York, upstate, again I would be traveling all the time. It doesn't matter where my office is." So, reluctantly, they left me here in Germany. But by that time I was never at home. I had an office in headquarters, but in the office there was always a flight ticket. In the morning I often didn't know that in the evening I would be sitting in a staff meeting in New York.
That was to see in which way, in terms of principles, of technologies, and of products, the computer field can be linked with the communication field. By that time, computer systems were already widespread systems, with large teleprocessing lines, a lot of internal communication, which was as complex as the usual telecommunication type. Specifically, I detailed products in the area of switching systems, PBXs, a lot of threads, multiplexors, many lines which came into a computer system. That was the basic mission: what must be done in order to link computer systems with communication systems. One of the products was a big software product, IBM's first big software product which came under my responsibility also: system network architecture, called SNA, was the first networking system. I didn't develop it, but it came under my responsibility then.
Nebeker:
Where was it developed, mainly?
Ganzhorn:
That was developed in Raleigh. A large group, a lot of engineers there on that system.
Computing and Communications
Nebeker:
Was IBM early in recognizing how closely tied communications and computing would be?
Ganzhorn:
If I may be frank, part of IBM did, but not everybody. IBM had some difficulty in recognizing that communications happens between products, not in products. And in that way the development efforts were difficult to associate and link cost-wise to the products. It was a new situation for IBM, which they were not aware of, before, and therefore these complex software systems — well, it was sometimes even an uphill race within IBM. To convince everybody that this was needed. It was the same thing as ten years earlier, in 1964, when it was difficult to make everybody understand that a computer was not a box of iron and silicon and germanium, but there was a software needed to run the computer with, so-called operating systems. When the 360 line came out, there was a shock in the company that you also needed an operating system 360, which cost a hell of a lot of money. People were very afraid of that cost. So those were the evolutions which had to be introduced into IBM. Starting in 1975, that was just about the time that the communications world had opened. And IBM needed to get access to this field.
Nebeker:
Also, software, such as this network software, required a definition of new markets. It's not a traditional product. How did you like that task?
Ganzhorn:
For me, it was an incredible challenge. It was most interesting; I liked it from the technical point of view. It was complex — the most complex software systems are still the communication operating systems, very complex. But I liked it, and I also had some background in communication technologies from before.
As I mentioned earlier, during my university time, as a side issue I took four semesters of communication engineering, which helped me a lot then, of course. So I knew the basics. That's helpful when you go in a laboratory and you understand what the people are talking about.
Nebeker:
Yes. So in those years you were traveling a great deal, talking to different groups developing different communications products.
Ganzhorn:
Yes. And also making plans for the corporation, for the total corporation for the communications issue in information technology. There were, of course, very important discussions all the time, and very important programs. Decisions in IBM: what to do, how far to go into communications, how far not to go —
Nebeker:
Drawing a line.
Ganzhorn:
Yes, drawing a line, but it was strategic-type management work. But in 1978, three years later, more and more it turned out that this job couldn't be done on a traveling basis. It needed a different organization with much more centralization in the headquarters of IBM at Armonk and at White Plains. And by that time, I suggested that — I was getting close to sixty then, also — this job needed a different calibration and somebody else should take it up. I also believed that this must be taken up by an American in America.
Nebeker:
Why is that?
Ganzhorn:
Because you needed so many internal U.S. facets and components and inputs. The Federal Communications Commission played a major role in this, and for a foreigner to enter its circles, it is very, very difficult. The Federal Communications Commission world is such a tricky world. If you want to be successful there, you must have lived a lifetime in that community, to know to whom to talk when, and how to get something. So I thought, "IBM cannot do this with a foreigner, because that is an insider's area. This must be done with all the possible resources you can find." Synchronously, it appeared that the President of the German Federal Republic called me onto the Science Council, and then I thought, "That's probably the right time to transfer the job to somebody else."
Nebeker:
How time-consuming is that job on the council?
Ganzhorn:
On the Science Council? It was not a full-time job, but you could spend any time on it.
Nebeker:
You felt you couldn't continue this vice presidency.
Ganzhorn:
No, I felt that both were incompatible. But that was not the major reason. It was more the opportunity to say, "Well, now is the time to make a point and make a change."
Nebeker:
And what became your IBM responsibilities?
Ganzhorn:
Then I went back to Germany and said, "Look, I have served on the international scene for more than fifteen years. Isn't that enough?" And then I said, "I am content running IBM Germany's technical and scientific operations in Germany, and that's it." I was doing this external work also on the German Science Council. So I focused on German issues later.
Nebeker:
And is the position you then held until retirement?
Ganzhorn:
I held that position from 1978 to 1986 and in 1986 I retired at sixty-five.
Science Council and Planck Institutes
Nebeker:
Can you briefly describe your work on the Science Council?
Ganzhorn:
Yes. The German Science Council was instituted in 1958, I suppose, as a council to advise the German government, and, from case to case, also to advise the county governments. On plans for scientific education, universities, for research programs, and for research institutions. One major part was that each year the Science Council of Germany had to issue and to extend the university plans for Germany by one year. How to build which university, where, what to extend there, what new fields should be taken up where, how they were going, what about these institutes which were there, are they worth being supported or should we consider a different approach? To oversee all the universities, all the special scientific institutions — we did not oversee the Max Planck Society.
Nebeker:
Is that because it's entirely independent?
Ganzhorn:
That's an independent organization. We did not oversee the German Research Foundation; that was in a modern facility itself. The Farnhofer Society was outside. But aside from that, everything which had a notion of science was subject to be judged by the German Science Council.
Nebeker:
Very important.
Ganzhorn:
Very important, Yes. We advised the government how much money should be spent on science, on university building, on building up new faculties, new disciplines. And on where it should be spent, where we should build, where we should not build...
Nebeker:
Now, in the United States, we have the situation that the advice of scientists and engineers often doesn't count for a lot in Congress, that Congress will decide nevertheless to do something different. How is it in Germany? When the Science Council recommended that money be apportioned a certain way, was that usually the case?
Ganzhorn:
It was done. Yes, it was done. I do not recall a single situation of the entire history of the Science Council where a recommendation expressed — or delivered, we didn't express, we put it in writing and officially presented it to the government, of course, each time — I do not recall one single recommendation against which anybody tried to do something else.
Nebeker:
Is that right? That means that this is —
Ganzhorn:
Not everything was approved by the German Science Council. But nothing could be done once the German Science Council had decided otherwise. This is still the case today. The German Science Council played an immense role — after my time, after 1989 — in absorbing the Eastern scientific communities of East Germany. There was a huge crowd of twenty-six thousand scientists hanging around in institutions which were dubious and kind of weak. The Council had to judge every little institute, whether it was worth being kept, or maintained, or built up, or closed down. And they did that. With no essential contradictions — in fact they got high praise at the end, that they mastered this job and cut it down to I think one-fifth of its former size.
Nebeker:
Well, I suppose if it's the case that you have this historical practice of honoring the recommendations of the Science Council, then a lot of lobbying activity would be directed to the Science Council, one would think then, rather than the political and other input coming in at the Parliament. It would be directed at the members of the Science Council.
Ganzhorn:
You would think so, but strangely, no. There was not much lobbying around the Science Council. It was so that sixteen of these twenty-two members were selected scientists from the universities, or from the Max Planck Society, so they always tried to get the very best and most mature-type scientists from these institutions. Six came from the economy, from industry. And I was one of these six, of course. These six should have a substantial scientific background, but also should have essential experience in the world of economy and industry.
Nebeker:
And in your view this has worked well for Germany?
Ganzhorn:
I think so, yes. I think it has worked very well.
Nebeker:
It's a non-political way of deciding certain questions.
Ganzhorn:
The science policy was made on non-political grounds in many respects.
Nebeker:
You served until 1987, I see, on that. Do you recall any particularly contentious or difficult decisions the Council made in your years?
Ganzhorn:
Oh, yes. The Council worked, of course, in subgroups, on many subjects simultaneously, all the time. And I was in many. I was in one for research, and I recall one which I had to direct, that was on academic education. We called it "the elite promotion," the promotion of young elites.
Nebeker:
How to provide positions for very promising young people?
Ganzhorn:
How to make sure that promising young people are not hampered by any system, but are brought up in the best speed and capability and so on. We had to work on a recommendation for that, what should be done in universities. So we worked on this recommendation, and we had to go through several reiterations before the recommendation actually flew. But based on that, a number of things were instituted. The basic thing was to overcome a prejudice in Germany which we were very alert to, in post-war years: that was the word "elite." "Elite" was a word which was considered suspect. Those were the people who misled others.
Nebeker:
Or the improperly privileged —
Ganzhorn:
A privileged class. In that recommendation, we did not use the word "elite," but we made sure that an elite could grow.
Nebeker:
You have also been on the advisory board of three Max Planck Institutes.
Ganzhorn:
On one, I still am. In fact on two of them I was chairman.
Nebeker:
Which Institutes?
Ganzhorn:
One was the Institute of Physics, the Heisenberg Institute in Munich, which is very famous, one of the most famous ones.
Nebeker:
Yes, I visited that, in fact, some years ago.
Ganzhorn:
I was the chairman of the Direktorium for that for a number of years. And the second one is Chairman of the Institute for Metal Research in Stuttgart. But there is a new institute in Stuttgart also, on solid-state research. That was only created in the 1970s and I was on that board also. Then I was on the proscenium of the German Institute of Standards, a dean. I was on the Federal Bureau of Standards, as you would say, in your terms, for Germany.
Nebeker:
The PTB?
Ganzhorn:
The Physikalische Technische Bundestadt in Braunschweig, I was on their Direktorium also. So on a number of these high-level German boards I was serving.
Communication Standards
Nebeker:
Did you have any particular area of standards that you dealt with yourself or made recommendations on?
Ganzhorn:
Oh, yes. Of course I was very much focusing on communication standards, together with computer standards. But one idea during my time had to be pressed very hard, namely that standards in communications was not something which you would establish eventually when the technology was old enough to standardize. In communications you need the standards first, before you can develop the technology; otherwise you cannot do anything without any standard. So that's a turnaround in the standards philosophy; it was not commonly known. Not aware of enough. You had to know German traditions in communications also. But that became very obvious when the software field entered communications. We were pressing very hard in those days. The active, developing productive developing forces in Germany must be part of this standardizing process, and they should be actively involved in the beginning of anything, in the development of new fields, so that the standards are a live product, being immediately developed as technology allows it. So that when the standard once is created, then companies can develop products.
Nebeker:
Well, I suppose one difference is that in an earlier age, say fifty or a hundred years ago, anyway, technologies were much simpler, so it didn't matter that twenty different companies produced different telegraph systems, and then one could see what worked best, and choose one for a standard. But development costs are so huge these days, that companies want assurance that if they develop a video system, there's going to be a market for it. So you have to, earlier on, establish a standard in order to get the investment. And communications technologies are also special because by their nature they have to be agreed upon.
Ganzhorn:
Of course, Yes. First you have to agree on the language before you can talk with somebody. That's communication standards.
Publications and Patents
Nebeker:
I see here that you have four books and some fifty publications and some fifty patents to your record. Can you give some overview of that work?
Ganzhorn:
I can. Two books I have written myself, and two others I have edited. The two I wrote myself, together with some assistants and coworkers, but I appear as the author. One is on the history of data processing, of information processing, starting out in the old ages. The other one is a textbook on data processing system principles, which is the outcome of my lectures at Karlsruhe. I lectured for twenty years, and then I thought the subject and the principles I had developed and was teaching there were mature and invariant enough to be written as a textbook which would exist for a certain time and not be subject to product development or something, I guess, independent of individual technologies. Then I wrote the book. It's a textbook of some two hundred and thirty pages. It has been used in universities for about a decade as the standard textbook.
Nebeker:
I see. And what areas have your articles been in?
Ganzhorn:
Oh, they have been very widespread, all over communication technology. Reaching all the way from early work in optical data, optical computing devices, to magnetic devices, to principles of computing, right across the entire computing technology.
Nebeker:
Now, did you manage to continue publishing after you became so involved in management?
Ganzhorn:
Yes, to a certain extent, because it was a challenge. As a kind of funny but very healthy instruction, I told all the lab directors who were reporting to me over the decades, (there must be more than a dozen which eventually I had under my direction), that once per year, I wanted to see him on the stage of a scientific convention giving an original paper.
Nebeker:
The lab directors. So you had to do this yourself.
Ganzhorn:
That's right. I felt very uneasy when I discovered this year that there's no major publication coming off my desk. Very seldom did I miss a year. Of course this has to do with a personal inclination; I was always interested in basic underlying principles of the general events. In the later years I was trying to publish kind of summary considerations of technologies, invariant principles which are worth being kept. One publication I recall very vividly. I made it in 1974, and it's being cited these days now as, the principle has been described then already —
Nebeker:
What article is that?
Ganzhorn:
That's an article on basic principles of computing. And in one section I mentioned adaptive configurations of systems, a system self-adapting to its task. I described that already as one kind of systems architecture.
Nebeker:
And that has recently become of much interest?
Ganzhorn:
These days it's being discussed very heavily in many respects. Many respects. A few months ago, a good friend from the United States, from our research division, wrote me a letter saying, "Karl, you know, you were not the first one on that. I've discovered a publication where an American scientist said about the same thing four years before you. But you were both way ahead." And nobody knows that we ever said it!
Nebeker:
Well, at least you're being cited.
Ganzhorn:
Sometimes, yes. But these days you wouldn't care anymore.
Nebeker:
What areas have your patents been in?
Ganzhorn:
In the same areas, many areas, yes. In computer circuits, computer devices. One of the very early patents was in optical computing, how you can do computing with light, magnetic circuitry, magnetic devices. And a whole class of patents was on how to build a computer strictly of magnetic components only. That was by the time that nobody knew — in the late 1950s, there were germanium and silicon — if they would ever make it as technical devices. There were many alternative technologies investigated, and magnetic technologies were the most prominent ones, because magnetics was a field which was well known in many respects, and the company technologies could handle magnetic materials then, reliably. So there was substantial thinking in the computer world about how to use magnetic devices. But it turned out, as soon as germanium was replaced by silicon, and silicon components, at first transistors and later circuits were reliable and reproducible, and the yields in manufacturing were real, then of course magnetics had no role left, because it was all high currents and much higher energy levels and all this.
Influential Books and Colleagues
Nebeker:
That's very interesting. I like to ask people what books have been very influential to them. Are there any that you think of that have served you for much of your career as great resources?
Ganzhorn:
Oh, I would say that the textbook on data processing principles, which I wrote in 1981, has had a substantial impact on the community and [great?] profit I also gained from that.
Nebeker:
I'm sorry, I meant not your own publications, but those written by others that have served you. Sometimes an engineer will tell me that "Whinnery and Ramo's book was the one that guided my work for a long time." Are there any books that were very influential?
Ganzhorn:
I didn't have a specific one which would be outstanding in directing my mind. The answer is that, no, I couldn't identify one particular book.
Nebeker:
Most of the areas you've been working in were so new that there wasn't —
Ganzhorn:
Beyond that, as you can imagine, being responsible for laboratory work which was growing very rapidly, there were so many different fields in the lab, that I had to observe many fields at the same time. And sometimes you've got an input there, other times you've found something interesting which should be brought up in another field again.
Nebeker:
What about outstanding people? Of the many people you've come across, are there one or two or three maybe that stand out as being really exceptional in some way, with management or scientific insight or technical abilities?
Ganzhorn:
Oh, that's hard. There are certainly a number of outstanding people which are outstanding to me.
Nebeker:
Well, I'm interested in people who have impressed you personally.
Ganzhorn:
Well, I determined my own life to a certain extent. But the first one certainly was my teacher in mathematics in high school. He was such a brilliant teacher; he didn't use one word too much, but said everything. I was fascinated by his way of teaching — by the easiness with which we could understand him. Mathematics for me was something which went without saying, and therefore I was attracted by that. The second man who really impressed me was the professor in theoretical physics at the University of Stuttgart.
Nebeker:
Who was that?
Ganzhorn:
His name was Ulrich Deilinger. You might not know his name. He's not one of the famous physicists, but he had an incredible way of leading young people to see the essentials without knowing everything. He could look through complex physical matter — or other things — and point right to that thing which was essential. He has systematically trained all his associates and students and assistants to watch out for that. From him I have learned a lot — to look behind things, to look for what is essential to a device. That really has helped me throughout my career, all the time.
Nebeker:
Was he the one you did your Ph.D. with?
Ganzhorn:
Yes. He was the one who told me to derive the crystal structure of transition metals from electron theory alone: "That's something new, that's worth a Ph.D., write it together." Time-wise it was much too early, I was scheduled for at least two or three more years to do that — but that's it! And it flew, and this work is still cited today. So this ability to look through things, I really learned from him and whatever I could do, in this respect, I think it has helped me a lot, thirty years later, when I wrote that textbook. There is nothing in that textbook which you could trace to a specific computer, but there is everything in that book which you find in any essential computer.
Nebeker:
So you tried to do what he recommended for computing.
Ganzhorn:
Yes, I wrote kind of an abstracted computing technology on computing systems architecture.
Nebeker:
What about people you came across later in your career, who maybe didn't influence you, but who impressed you immensely?
Ganzhorn:
Oh, there were certainly a number of people at IBM, there's no question. IBM had outstanding executive people during my time.
Nebeker:
Are there one or two that you'd single out?
Ganzhorn:
I mentioned one already, Mr. Bob Evans. He was a vice president of IBM, and he retired a few years ago, early retirement. He's now running a small venture company in San Francisco, very successfully, also. From him I learned what it means to be a driving force. He is a man of dynamism which you cannot believe sometimes. He is bursting with energy, and he is doing things — every second something happens, no time to lose. So I learned that from him, to use my time, and to be effective. I say I learned it; maybe I didn't learn it, but at least it was a kind of example, how to design a life. That was very impressive. And of course, we mentioned the name of Dr. Piori earlier. He's also a man who had a very independent vision, through things, through people. He knew immediately whether something was behind a man or not.
Nebeker:
What about researchers that you've come in contact with?
Ganzhorn:
Oh, there are so many, that I'm not sure I can —
Nebeker:
There's not any one or two who seem to just stand head and shoulders above the others?
Ganzhorn:
I have not thought of the question, and you take me a little by surprise. I'm sure there are, above them, which we must respect. Some of them I do not respect, I must also say.
Nebeker:
Who've been nevertheless been widely acknowledged?
Ganzhorn:
Some famous men in science — in physics in particular — I've lost respect for because of other weaknesses. Human imperfections.
Nebeker:
Failings.
Historical Importance of Series 360
Ganzhorn:
I wanted to mention one view which is very strong in my mind. And that is the basic role of the development of the 360 product line in IBM. The basic rule for the evolution of information technology — there's much more in this system development contained than the public knows today and recognizes. Even the scientific community recognizes it. Many of the basic principles which are invariantly left in computer technology and information science today go back to the 360. They were all being created, almost, there — it starts with the eight-bit byte, that a character should have eight bits, a very simple thing. It starts with the idea that IBM, before that time, had a large group of individual computers. Each was completely different in its own terms. Each had to be programmed separately. When they said, "We need a line of products where you can have a program, and if your computer becomes too small, you transfer this program to the next higher computer, and it runs on that one, too," we called that compatibility among our product line. That meant that we had to create a computer architecture.
Architecture is the external appearance of the computer — namely, the interface, or the way you can communicate with the computer, the total set of instructions which you can give to the computer — that represents the architecture. That's what a computer can do. For the first time, these three architects of the line — Fred Hooks, Andau, and Blau, those were their names — said, "We want a whole line of products, one complete family of computers, which has the same architecture." So the user doesn't know which computer he has — he can have a small one, or a bigger one, or a very large one. The performance of the systems is different. The larger one runs a hundred times faster than the smallest one. But the way he uses it is the same. In doing this, IBM has created the basis to separate the programming of a computer from the hardware. That is the key and the clue to creating an information — informatics — industry. The entire software industry today goes back to this concept.
Nebeker:
Separating forever the hardware from the —
Ganzhorn:
Separating hardware once and forever from the function. Separating hardware from function. That's such a basic decision that a billion-dollar software industry lives on it today. That's one of them. Take another one. It's a special thing. You might not know what I'm talking about, but nevertheless: so-called relative addressing. That information in a machine is not addressed directly — by giving it a house number or something — but addressed in a very simple way. We say that address is x points away from a base address. We only give this x a relative address which refers to a base point. And the base point we leave open. The operating system of the machine is programmed in such a way that it can allocate base points, so that the operating system by doing this can move information in machines. It's a simple principle, but by way of this principle, information became movable in machines, another basic principle.
Nebeker:
The 360 was the first time that was implemented?
Ganzhorn:
Yes. And the funny thing is, the principle as such was already contained in many computers, but nobody knew it. Under different names, it was called indexing, it was called substituting address, and it combined both of them — and all of a sudden it became relative addressing, relative to a moving point, to a moving start. When you start at the starting point, when you can move this, you can move the entire information. That's another basic point: being sure we have enough supplies. I could list a few more of these very basic principles. The 360 has created that, and the computer world has, without saying it, absorbed it as a given thing. The entire computing world today lives on these. In fact there are five of these principles. I won't try to develop them now, of course. But that's the basic role which the 360 development in IBM has contributed to the growth of computers.
Nebeker:
Well, thank you very much!
- People and organizations
- Corporations
- Government
- Inventors
- Research and development labs
- Scientists
- Universities
- Profession
- Business
- International collaboration
- International trade
- Computer industry
- Decision making
- Human resource management
- Innovation management
- Project management
- Research and development management
- Engineering and society
- Military applications
- World War II
- Education
- Educational institutions
- International affairs & development
- Communications
- Communication networks
- Computing and electronics
- Circuitry
- Digital circuits
- Semiconductor devices
- Integrated circuits
- Telecommunication computing
- Computer science
- Formal languages
- Data systems
- Data acquisition
- Information theory
- Audio coding
- Speech coding
- Pattern recognition
- Character recognition
- Software & software engineering
- Software architecture
- Materials
- Semiconductor materials
- Transition metals
- Engineering education
- Engineering disciplines
- Computer engineering
- Standardization
- Standards
- Standards organizations
- Telecommunication standards
- News
