Silicon Genesis Europe
Round Table Interviews
electronica 2006 – November 16, 2006
English Translation by Horst G. Sandfort (HGS)
Moderator: Horst G. Sandfort (HGS)
Dr. Ernst Hofmeister – SIEMENS (EH)
Dr.Hans-J. Schuetze – Texas Instruments Europe (HJS)
Dr. Werner Freiesleben – Wacker-Siltronic (WF)
Dr. Uwe Thomas – German Ministry for R&D (UT)
Dr. Hans Rebstock – SIEMENS (HR)
Ing. Curt F. Kesting-Fairchild Semiconductor / ST Microelectronics (CK)
Ladies and Gentlemen,
Welcome to the Round Table discussion, where all participants are in line behind small round tables,
“Silicon Genesis Europe ”.
The Gentlemen assembled today have been referred to as “cornerstones” of the semiconductor industry. I had to revise this and must say “yes”, we may have one or two cornerstones here, however the gentlemen present are solid rocks of the semiconductor industry, as they have been engaged in the early stages and have assisted in the fact that Semiconductors were accepted and applied in systems and technology.
We should note that the semiconductor industry in terms of history represents its 50 th anniversary, however in reality the foundations were laid already in 1750, when Benjamin Franklin documented and published his “Observations on Electricity”. One hundred years later, 1860, Philip Reis demonstrated at the Garnier Institute in Friedrichsdorf the electrical voice transmission, the telephone. 1876, sixteen years later, this invention was economically applied for the first time by Alexander Graham Bell, who constructed a phone and started marketing it. 1896 Mr. Marconi was the first, who demonstrated that on the principles of Maxwell, voice could be transmitted for the first time wireless on a distance over three kilometers. The commercial utilization of this invention started at about 1900, when the special tubes were developed. The tubes not only allowed transmissions however also reception. 1931 Manfred von Ardenne added a highlight with the invention of the Cathode Ray Tube, which signaled the beginning of commercially feasible television.
1942 it was Konrad Zuse, who build the first world wide recognized computer on the basis of Relais, the Z 3, which was followed in 1946 by the ENIAC, developed by John Jesper Eckert and John Mauchly.
And at about that time, in 1947, a new “century” began with the invention of the transistor in the Bell Laboratories. This new century was from its inception accompanied by several Gentlemen who are present here on stage and it is my pleasure to welcome the participants from left to right:
Dr. Ernst Hofmeister, the senior member of this round, who was there, when it all started and I like to direct my first question to him, as I believe, the history books should be corrected somehow, when I refer to the year 1947 as the break through year for the Semiconductor Industry.
Dr. Hans Schuetze, Chief Technology Officer at Texas Instruments in Freising. As soon as we shall hear later from him, how he got involved, we shall see an interesting perspective, how German Scientists became successful in this new industry.
Dr. Freiesleben, to my right, a pioneer who really made it happen, that Semiconductors could be produced. Without the raw material Germanium and Silicon, this would have been impossible. The first Silicon Ingots were pulled inside SIEMENS, however WACKER Siltronic made it a worldwide success with a real entrepreneurial approach.
Dr. Uwe Thomas, a personality who has been instrumental for the development of the semiconductor industry, however not with regard to technology, but inside the BMFT, where R&D funds were decided and made available to push primarily the European, respectively German Semiconductor Industry into a dynamic evolution, in order to achieve a leading position worldwide.
Dr. Hans Rebstock, a member of the R&D team inside SIEMENS, can remember good and bad times and specifically how the competitive environment was perceived and responded to.
Ing. Curt Kesting , to my far right, was one of the first people who joined SGS-Fairchild and served for 24 years of his life in companies associated with Fairchild Semiconductors, a real interesting perspective of life.
The gentlemen present have all a different background, a different set of experiences and memory of history, as I already alluded to. Dr. Hofmeister, was the first transistor invented in 1947 ?
(EH): Yes, however this device had predecessors and you have asked me to also add to the history in general. You forgot to highlight the year 1878, in which Ferdinand Braun developed the Rectifier, the first metal based semiconductor contact and wondered, why the current was delivering a different strength in one direction, than in the other. He did not understand the reason, he thought of magnetic effects or something like it. This was Ferdinand Braun, who also invented the famous Braun's Tube, the basis for the TV in front of which people spend so much time today.
Now to my personal experiences in the Semiconductor World.
I studied Physics after World War II, partially already before this time, at the Munich LMU ( Ludwig Maximilian University ), classical basic physics. And one day there was a sticker on the billboard: Special Presentation by Heinrich Welke: the transistor. Hm, as I am curious, I thought I may go in and listen. There were five listeners who were hearing Heinrich Welke talking about the transistor. This was so exotic and not publicly known. This was the same Professor Heinrich Welke, who in April 1945 – and this is again a correction to the history books- during the last days of the war, delivered by a courier a patent description to the (Munich ) – Berlin Patent Registrar – a Field Effect Transistor.
Unfortunately this one was never demonstrated, however the theory was correct and established. Consequently we can say that Welke was the first Researcher who has laid the theoretical groundwork for the development of the transistor.
Now, after I had listened to the presentation, I have been invited one day to visit the Semiconductor Factory of SIEMENS AG in Munich . As Physics Expert I was so fascinated, I had the feeling, I was looking at the first industrial process generating physical elements. Chemical Industries were long established, however a physics industry was new and consequently in 1945 I joined this industry at the SIEMENS factory in Munich . I started in the Diodes Laboratory, the next step in the development of the Rectifiers, developed by Ferdinand Braun. One needs to state, that during the war Rectifiers were already manufactured in the USA on the basis of Silicon and in Germany on the basis of Germanium. These devices were paramount in the development of demodulators for the Radars.
This was my first exposure to this industry and in my laboratory we developed Diodes. In the mean time we got to 1955, where in the next laboratory to mine, transistors were developed. Actually we had a production line were Germanium Transistor were made, specifically doped Germanium Transistors. This was a very delicate process, where on a very small Germanium base plate Indium elements were added, so that a P-N-P effect was achieved.
This much to my contributions. We are still at the transistor level. In 1952 there was a break through event, when the Bell Laboratories called and opened their process know how for Germanium Transistors in a Symposium. The cost for participation was US $ 25,000.00. There were 35 companies invited. We came back from this trip to our factory with a heavy book. Immediately after our arrival, this was locked up, as it was considered “top secret”, as it contained the complete recepie for the manufacture of Germanium Transistors…however connected to license fees. This much to the early days.
(HGS) Thank you, thank you very much. I accept your correction, as you know more as a real cornerstone of this industry ( laughter)…I am still learning. This is what I wanted to say.
Dr. Schuetze, we just bridged the time between 1945 and 1955. 1955 was to my knowledge also an important year for you. Can you explain to us what happened ?
(HJS): Off course. I am also Physics Expert, as my neighbor, studied at the University in Kiel . In 1955 I finished with my dissertation in applied physics. My industrial career started with TELEFUNKEN in 1956 in Ulm , however not in semiconductors. I worked for 6 years in the R&D department with the Roland Technology, before I joined the semiconductor labs for 4 years. During this time the material we worked with was exclusively Silicon. We studied Silicon structures, specifically, how one can reduce the size of the structures. We accomplished in close cooperation with Zeiss, Oberkochen,
to develop sub micron structures as small as 0.6 micron. This was absolutely new territory at that time. However I did not stay with TELEFUNKEN, as I had to answer the question which was on my mind: shall I stay in R&D, or shall I move into industrial production, to learn more about advanced management, which lead me to Texas Instruments in 1966. There I worked for 21 years. In the first 10-12 years I was in charge of the Integrated Circuits Department. This was the time, when TTL IC's were manufactured in Germany and custom designed IC's for companies like Nixdorf, GRUNDIG and also for SIEMENS were developed and manufactured by TI in Germany. The last eight years, before I retired, I lead the entire R&D efforts for TI in Europe . This far to my story.
(HGS): Thank you, thank you very much…you have already bridged over to the sixties, which does not mean we should not go back to the fifties. Dr. Freiesleben, 1954 SIEMENS started with the mass production of Silicon Devices. Where did the raw material come from. How did you get involved with Silicon ?
(WF): Well, I studied in Munich and New York and wrote my dissertation at the University of Basel in physical / chemical processes. My next step was with Wacker Chemicals, where a delivery of “clean” Silicon had just left for the USA , out of 100 Kilograms of “clean” Silicon, which Wacker had produced.
This was under Dr. Eduard Eng, a friend of Dr. Spenke, who worked at SIEMENS and knew what developments were going on in the USA with regard to Transistors. He said to his friend Eduard: Edi, pay attention – this was about 1952 – transistors also work on the basis of Silicon and not only Germanium. Upon this information Edi Eng, a full fletched expert in pulling ingots out of Silicon, put a team in June of 1953 in place, to enhance the basic Silicon process, licensed from SIEMENS, to an extremely clean Poly Cristal based Silicon Ingot pulling process. This task was not easy, as one could buy metallurgical Silicon, which was worldwide available, in volumes up to 2 tons, with elements of Iron, Calcium, Aluminum of about 1 % mixed in, however for the requirements of the Semiconductor Industry, such a product was not useful. It needed to receive Silicon so clean, that only 100 million of an Atomic Percentage of foreign material could be present. That is a Silicon out of nine times nine, 99,9999999 % clean material. This process has been put in place by Dr. Eng and he did deliver. There was a saying within WACKER at that time: SIEMENS have taught us how to make Silicon, however WACKER have demonstrated what needed to be done with it. In the following years the company became so competent, that it achieved a world market penetration of 45 % in the mid eighties. Several People in this room may remember this. I am now passing the word back to Mr. Sandfort so he can give word to Mr. Thomas.
(HGS): You did already, however Dr. Thomas will explain to us his engagement with the Semiconductor Industry and his first meetings.
(UT): OK, that is not specifically exciting. I studied Physics and started with the AEG Research Laboratories in Berlin . After six months the company decided I had to learn something new and send me to Texas Instruments to learn everything about Microelectronics, which I felt had nothing to do with the field of expertise, we were working in: CAD Systems for the development of Relais.
My real engagement started in 1973, when I joined the German Ministry for Research and Technology. I will later detail this. I would like to make one point. At that time Germany was considered “behind” in the field, which lasted for many years and did not really improve. We were looked at with pitty and I remember a visit by Bob Noyce and also by Robert Heikes of Motorola, who were laughing, when we explained we wanted to catch up. By the end of the nineties, Europe had managed to move three companies among the top ten of the Industry. This was considered “impossible” in the seventies. What the future will bring, is another question.
(HGS): We are a bit nostalgic today and would like to really go back to the early days and focus on the issues from these days, the road blocks, the hurdles which had to be overcome. We keep jumping forward, however I would like to go back and see, how we did finally get from the fifties into the sixties. May be Dr. Rebstock can shed some light into this time period.
(HR): I was born in 1927 and started as was customary at that time as young soldier at age 16 to feed a “Flak” with ammunition. I was lucky, as the majority of my comrades were left behind during the fights at the “Zeloher Hights”, bleeding to death. I was lucky again, as during the war the Radar Technology became paramount. Goehring believed at the beginning, he could win the war with the “Stukka”s, however when during one night 185 German aircrafts were lost over England , people started wondering. We learned that the Radar Technology developed in the USA to measure the territory, were now applied to locate aircraft. The reason why I am telling this is simple: the center for accuracy were Germanium Rectifiers, which had such a sensitivity, that they could be used in aircraft recognition.
I studied Physics in Stuttgart at the Institute for Microelectronics and at the Max Planck Institute for Metallurgy. My dissertation was about SELEN. I was primarily interested in the use of SELEN for High Frequency Applications of Rectifiers. One of the great Pioneers, Professor Seiler, was my Promoter for my dissertation and I was already fascinated by the Semiconductor devices. We had on the other side also a very good theoretical physics expert, Professor Fuss, who taught us the Theory of Quantum Physics. To deal with the hardware, we quickly came to a very important subject – Mister von Freiesleben already mentioned – the chemical cleanliness and also the structural cleanliness.
At that time this issue was starting to be really addressed, even though ten years earlier it was already discovered that degradation in structures were observed, which quickly were made responsible for the useful characteristics of the material. This meant for Germanium that the sensing of the temperatures was very important, so that atomic copper would not settle in the structural design of the devices. If this happened, the devices were garbage. The processing technology was at that time progressing very fast and I have been working on the enhancements of the structures, specifically the impact of copper crystals. These were my early studies at the Max-Planck-Institute up to becoming an Assistant, before I joined SIEMENS. At that time SIEMENS was very much destroyed, however there was an old research institute in Karlsruhe in a provisional laboratory, called the Main Semiconductor Lab, where we focused on Semiconductor technologies and also included Silicon. One remark, at this time semiconductor technology was worked at within Siemens at several locations. It has been mentioned already, one group was under Spenker – Schottky in Brennsfeld, which build the base for the Schottky effects, first in Theory, leading to the first discoveries for the transistor already in 1935 – 1938. When later in the USA the first transistor was build, it was on the “shoulders” of the work by Schottky, who had described the surface effects of the transistor. Mr. Spenke was his co-worker, who taught us the mathematically very difficult theories in a way so that we could understand them. In Brennsfeld, which belonged to the Schuckert Group (of Siemens) the Silicon became important for the group developing electronic power devices and rectifiers using high power, so the silicon had to be very clean. Therefore the cleaning processes Dr. Freiesleben spoke about were very important and more difficult to apply for Silicon than for Germanium. This was achieved by using high vacuum chambers, which were originally developed by Siemens.
So, Schottky – Spenke developed the technology in their main laboratory as a foundation for the first semiconductor factory in Munich , which opened in 1952. May be one should mention at this stage, that during 1949 a new law had been passed for Germany, that all war relevant technology development including semiconductor technology were banned and I do not know exactly how long this ban lasted leading to 1952. This ban did not allow us to work on any advanced technology other than simple diodes and comparable devices. Starting with 1952 this changed. I joined this semiconductor factory as head of development for Germanium. We developed the famous MESA Transistor, about whose importance we can talk later on. I advanced to Development Manager for Silicon and later to Manufacturing Manager for the factories we opened in Munich , Regensburg , Villach and Singapore .
It has been specifically difficult to work to apply the contacts, as it required a large amount of ladies using microscopes, which we could not employ in Munich , hence needed to outsource this process step. For information just one example: the wires we used to make the contacts to the transistor chips were 5 micron thick. A hair has 6 – 7 micron. We were producing real art work at that time and finally conquered the world with the transistors, for which we had developed specific production processes. Even though the transistors were invented by Texas Instruments and Motorola, however these are details I shall not address at this time, as we were quickly moving to apply TTL technology and later on Memories. This allows me to add one detail, which is not well known. The so called single cell transistor, the basis for all DRAMS, used today, was invented inside SIEMENS by Karl Ullrich Stein, in our Research Laboratory. This means even the large scale memories, developed later on in the USA , which we may discuss later on, have this “little” Siemens invention inside. This for the time being, we can talk later about more details, specifically about the times were there have been semiconductor wars (between the continents), which led to the MEGA project, supported by the BMFT ( Department for Research and Development) which allowed SIEMENS and Philips to catch up with the rest of the world. Thank you.
(HGS) Thank you Dr. Rebstock.
Mr.Kesting, how did you get involved with Semiconductors the first time ?
(CK) Yes, I am here for the first time. I am proud to participate in this elusive round as the youngest member. After my education with the “Bundespost”, the predecessor to the Telekom, in 1966 I joined SGS – Fairchild, which was at that time a joint Venture in Wasserburg and I stayed with these two companies almost through my entire business life. 1970 Fairchild opened a facility in Singapore and 1973 one in Wiesbaden and 1984 ST (SGS-Thomson) opened a Diffusion in Singapore , the first outside Japan .
(HGS) Thank you, we shall definitely come back to several stages of your professional life. Now quickly let's go back to the world, how it presented itself in Germany . For me it is fascinating to hear that already from 1945 to the mid fifties significant developments have been contributed to the semiconductor industry to build the basics, which should have created a significant advantage, held up however by the law imposed upon by the USA , to halt R&D between 1949 and 1952. During this time the progress in the USA seems also not as significant. The real push came with the development of the Integrated Circuit and later, Dr. Rebstock already mentioned it, the technology war times, which had to be overcome, which began with the invention of the Integrated Circuit in 1957, which is nowadays referred to as the birthday of the Semiconductor Industry in Silicon Valley following the American Dream, as several gentlemen left Shockley labs to start Fairchild Semiconductors. Most people however ignore, that during the same time period in another region of the USA also significant developments were taking place, in Texas . There was a company, which still exists today, at which you, Dr. Schütze had engaged: Texas Instruments. What were they all about, the TTL integrated circuits.
(HJS) Yes, I have not been involved in the mid fifties with semiconductors. This started in 1960, when I learned about the invention by Jack Kilby, who had build the first integrated circuit, which later on, as you know, owned him the Nobel Price. I was during that time busy at TELEFUNKEN, as I mentioned already, dealing with sub micron structures on silicon, which did not let me consider anything else, when I joined Texas Instruments in 1966 with the task to introduce the TTL product family into Germany and to dominate the market. That was in 1966 and I did as directed.
(HGS) Thank you – this statement surely triggers several remarks, which we shall pick up every time, when we talk about the product developments, which also required the next step in silicon developments. Dr. Freiesleben do you have any comments ?
(WF) 1958 when I joined Wacker, Gold cost about one US Dollar per gram, whereas clean silicon, which Wacker could sell as the first manufacturer in the world, cost US Dollars five per gram. Had the value of a factor of five over Gold. In 1985 to make that jump, Gold cost US Dollars fifteen per gram, whereas Silicon cost US Dollars twenty cents per gram. My first encounter with the integrated circuit was in 1962. The already mentioned Dr. Eng returned from the USA , carrying the first integrated circuit replacing the functionality of eight radio tubes. We were completely “flat”. How could a piece of Silicon, as small as a thumb nail have the functionality of eight radio tubes ! Our boss asked, what can be the price for this device ? May be one hundred Dollars…now, one tube cost US $ four….Edi, do you believe that Silicon will ever be a business ? With this, I pass on to the people who made it a business.
(HGS) There is one more remark with regard to the statements by Dr. Schuetze, Dr. Hofmeister ?
(EH) Less with regard to Dr. Schuetze. We have spoken about the backwardness in Germany and all of Europe with regard to the developments in the USA . It has already been adhered to, what caused it. One should not forget, that in 1949, when in the USA the Transistor production already was in full swing, Germany had not recovered from the Currency reforms in 1948. Befor the currency reform it would have been necessary to address…everything should have been addressed….except after the currency reform the industry addressed primarily “classical industries”. Why Semiconductors ? We needed engines to generate current, we needed equipment to produce steel etc. For that matter, one had not recognized the importance of the semiconductor industry, as one had no need for these exotic devices, specifically not prior to the re-establishment of the basic industrial requirements. And at that time a lucky and not so lucky thing occurred: the “Wirtschaftswunder” ( the economic miracle). The classical industries boomed in Germany , everything was directing upward. Nothing could go “backwards”. Nobody seriously looked at the chip developments and several people voiced, that these devices could better be bought from the USA , then made locally. This has proven completely silly, as the key technologies for the future must be available “in house”. As a consequence some of us stood up and started a catch up run. However there were several hurdles. There was a drive by the unions condemning the chips as job killers. This was picked up by the Media and broadcasted intensely. This resulted in a less dynamic acceptance of the new technologies during the sixties and seventies, as experienced during the first industrial revolution. The first industrial revolution was the steam engine 200 years ago. 100 years ago, the second revolution was the discovery of electricity and now we have the third industrial revolution with the micro circuitry and information technology. This third industrial revolution was welcome by the experts, however not understood by the masses, actually leading people have tried to stop any developments in this field.
I collected headlines from newspapers: Stop Micro Electronics – We are running into a catastrophe –
Everything with regard to microelectronics had a negative touch, even the first computers, the Personal Computers specifically, were condemned. A member of the presidium of the main Union, Education and Science, referred to this as the “technological Hiroshima ”. Robert Kung, a well respected theological Professor voted against, as did Barbara Mettler-Maybohm, Professor in Hanover : first die the verbal communications, then dies the sole. Everything that this new technology was bringing to life, was condemned. The most shocking to me, when I was reading an article published March, 27 th , 1987 by Jutta Dittfurth: “the novel information and communications technologies are designed as destructive planning instruments to control and rationalize human beings and the nature. In private and industrial life, the I and C technologies are designed to kill more jobs, de-qualify, control, manipulate and cause cultural degradation. The computer has to be removed from schools”.
This has been the public controversy in the seventies and eighties. You can imagine that it has been no fun or just due to these controversies been fun, to continue the developments within SIEMENS.
New challenges were coming up, when we planned a new factory in Regensburg . We were asked to stop production on Saturday afternoon, no work at all on Sunday, this was “verboten” ( forbidden).
When we received visitors from Japan or the USA , they were shaking their heads. Why can you not work on Sundays ? We were not allowed to do so, as we were unionized by the “IG Metall” and as such, different to the “IG Chemie” where steel cooking belonged. We could not work on Sundays. Also the Bishop from Regensburg got involved – no way, to work on Sunday.
We finally got tough and demanded Sunday work to be permitted, as we explained the difficulties with shutting down Temperature controls, Gas supplies on Saturday and trying to bring them up again on Mondays. This would result in yield losses up to 60 / 70 % on Mondays from losses less than 20 %, when work could continue. This had to be fought with the Church and Union Administrations. Also we wanted grants from the Government and you may imagine that even though the responsible individuals were in favor, we did not succeed. May be Dr. Thomas can refer to this, as we did not succeed. The overall opinion was opposed to accepting this devils stuff. Also the user industries did not really support us, as they argued, we can buy this stuff from abroad. Balkhausen wrote during these times a significant book: “Elektronikangst” ( fearful of electronics). To catch up with technology developments elsewhere in such an environment was not easy. However we did catch up. How, this will hopefully be addressed later. The characteristics of the times were as described, exactly the time as described in the book by…I do not remember right now…the new world…I may remember soon and will let you know (A. Huxley: Brave New World). Yes, in any case the initiative should have come from the public, however it was never there. This came later, almost too late. How we recovered any way, will be addressed by Mr. Thoams and M. Rebstock.
(HGS) Thank you. I do not want to contradict you, however as I scratch my head, I am reminded that during the difficult period you describe between the late fifties and late eighties, several non German companies started a wave of new foundations and manufacturing facilities within Europe. For Example: Fairchild-SGS , Texas Instruments. These companies were faced with the climate you described and fought the challenges. Dr. Schuetze, two short questions: how was the beginning, Was a Golden Cage build for you ?
(HJS) No, No, when I joined Texas Instruments in 1966, I was working the first nine months in Dallas, where I participated in the development of integrated circuits in N-pack – plastic packaging. This was no sugar licking, however after a few months, when I received the message from our first factory in Freising, which had opened the same year on April one, that SIEMENS had placed a large order for TTL circuits, under the condition to be manufactured in Germany, I packed large suitcases with the equipment and my experience I returned to Germany, where nothing was prepared. We had to start in a former horse barn, where we assembled the first integrated circuits. This has been dramatic, however we did it. With regard to the competition described by Dr. Hofmeister, I have honestly not noticed any of that. We were excited by the idea to move TTL circuits successfully into the German market. We had no German competitors, however we had to deal with foreign ( USA ) competitors on the markets. However we had developed technological advantages and a concept, how to market high technology products successfully. I developed the three pillars concept. These three pillars were:
Pillar one: Product Innovation, pillar two: Product Positioning, pillar three: Product Support.
You heard me say “product” not “integrated circuit”. To sell an integrated circuit is simple, if it can be done. However to bring a product to market and making this a success, is something completely different. An innovative IC with long term perspectives is mandatory and needs a complimentary data description , however not on “flying papers”. In reality a book. We had the TI TTL Databook available. We organized workshops for our customers so they could study the new technology. Really like at school. Thereafter we delivered the TTL Cook Book to support generic applications. I do not know, if anyone does remember this. Finally we delivered application specific design support. This entire spectrum I call: Product Innovation.
The positioning of the product in the market is depending upon three factors: first: customer recognition of the supplier, customer recognition of the market opportunity, customer recognition of the product.
The customer recognition of its suppliers has been very important to us. That results foremost in a commitment from the highest levels of management down to the lowest level with regard to the product. The statements made were measured and we have paid great attention to the validity of our statements. This was part of the success story of TTL and also the availability of resources. The first important resource I already referred to, the request by the customer to manufacture in Germany . The second important investment was into quality assurance and timely delivery. And the last resource which helped us very much at that time – when there was no Internet - was a Satellite based e-mail system in side Texas Instruments which connected 44 manufacturing sites around the globe. Meaning, we had internally our own Internet. The customer recognition of the trends in the market was extremely important, not only via PR activities but more by the development of the prices. A customer, who wants to build a product into his systems, must be able to rely on the price forecasts for the IC's, when he introduces his product into the market two years after he started the design. These volume forecasts were available from the beginning of any project for ten- or hundred-thousand units and something else. We had the knowledge, that with the doubling of the production volume, the prices were falling by 70 % to 30 % , which we called the learning curve principle. The customer knew exactly, when he was in volume production within three or four years, what his cost would be and he was able to plan.
Customer recognition for our TTL products we valued very highly and made sure we had first class technology products delivered. Each single series should not become a fly by night event, but be a part of a series of products we developed by which we warranted continuity for an entire product spectrum. This long term engagement with a supplier became an important base for the positioning of the product. This much to the positioning of the product. Now to the product support.
The designer in a development lab desired a permanently good contact to his supplier. For this we had a product marketing department (PMI's) – product marketing engineers, who have been trained to excellence, so that the customer, if he so desired, always had the same person to deal with. We even had red phone lines, when necessary. This was completed by our sales force in the field, the sales crew, as we called them, which was responsible to establish a close relationship with a customer, developed the logistics support and last not least to warrant the recognition of the customer. These three were the key factors to our success. If you allow, I would like to add a historic remark. 12 years later, in 1980, Intel in California started its “Operation Crash” to establish its microprocessor, the 8080. And this Operation Crash is exactly, what I just described. By the way, very well described in the book “Marketing High Technology” by Bill Davidoff. Thank you.
(HGS) Thank you Dr. Schuetze.
I like to pick up two remarks. The one is the statement, that Dr. Schuetze and his crew were ready to deliver pricing forecasts for the products.
Dr. Freiesleben, did he put you under pressure by these remarks or was they customary to you.
(WF) We delivered Silicon to all chip suppliers and we were under equal pressure by all of them. We invested 14 % of our revenues into R&D in order to deliver architecturally clean silicon and to deliver surfaces, which in comparison would allow only one piece of a rice corn on the size of a soccer field as impurity. In 1980 we had already four thousand different grades of Silicon products, various crystals and various diameters and twenty five different geometric requirements, twenty five different electrical requirements and just to mention pricing: the metallurgical silicon cost per kilogram US $ 1.00 ( one), the poly-crystal very clean cost per kilogram US $ 50,00 (fifty), the prepared sliced silicon cost per kilogram US $ 5,000.00 ( five thousand ); the polished slice cost US $ 9,000.00(nine thousand) and the epitaxial slice cost per kilogram US $ 25,000.00 (twenty five thousand). These prices were under permanent attack by the chip suppliers, even though the cost never exceeded a 4 % portion of the total manufacturing cost.
We once changed the Oxygen content for a slice to Intel accidentally. Each customer had his own formula. However, Intel got normally 300 die out of one wafer and due to the error they had 400.
In our eyes, we delivered a real added value. So I asked Gordon Moore, please provide at least US $ six to us. He answered impossible, but see the purchasing person, I have nothing to offer. And the purchasing person said : no way.
This much to “cost”. Thank you.
(HGS) OK. Thank you. We hear clearly, you have been under permanent pressure. Please allow one more quick question, please pick up the microphone one more time. It may sound like barriers to growth, when prices permanently drop. However you did follow the overall trend, reduced cost, and developed the next generation larger ingots and wafer. The company kept growing in spite of this, or did I say something incorrect.
(WF) I must make one remark. Wacker had in the mid seventies a market penetration of 25 %. Our American customers had the opinion, they did not like to depend upon deliveries from a factory which was only one and a half hour away from a potential tank aggression from behind the “Iron Curtain”. You must invest in the United States , or we must cut you as our supplier. Then, after it become known that we were willing to invest in the USA, our US competitors reduced their prices by 28 % to keep us off and to make our project unattractive. I took the project on and my first action was to increase our prices in the USA by 12 %. I believe this was the only action which took place in history, however I did succeed , I argued: if you do not want to continue to receive deliveries of ultra clean silicon, keep cutting the prices, however if you really need this product and are dependent upon the deliveries, please pay and allow investments into research as well. This way I shall be able to continue to support you. Thank you.
(HGS) Thank you. – May be we can swing back a little, as we had heard from Dr. Hofmeister that the German Industry, also the European Industry, you specifically mentioned SIEMENS and PHILIPS, had the feeling to be falling behind and needed a turn around to catch up. Dr. Rebstock, Mr. Thomas, both of you were leading the chase. You may share the answers.
(HR) May be I should make one or two remarks with regard to the early developments. As mentioned by Mr. Hofmeister, I myself had to deliver speeches in the Academy of the Protestant Church , that a Christian could contribute to the manufacturing of Microelectronics and I had to drink the processing water, which was leaving the factory into the rivers for demonstration purposes. The discussions were about the Fluor content, which was forbidden substance to us even though diagnosed lower than in the drinking water in the USA .
This was one important point. Also I would like to go back to the time before the real battles started.
Siemens had already build a fully transistorized computer in 1954, when IBM did not have any fully transistorized computers. For those, who are interested in history, these were Germanium Transistors and Rectifiers, which were manufactured inside Siemens at their first Semiconductor factory, set up by Dr. Sieberts. Dr. Sieberts, by the way, a person, who has been already with SIEMENS, when the Battle Ship “Bismarck” was equipped with electronics, allowing its firing power to be quadrupled. Dr. Sieberts was almost shot during the war, when the technology once failed to function. He was the one who started the Semiconductor factory in Munich , which was based upon Germanium to build the first transistors for the computers; later transferred to Silicon Transistors, followed by TTL circuits to be later replaced by Microprocessors and Memories. In between we experienced another phase at which the semiconductor factory at SIEMENS AG's capacity was split
to support the entertainment industry for the open market and the internal requirements. As an example, we had to develop a fast technology, ECL, for the computers and Logic systems with TTL, which Texas Instruments had successfully launched and finally reigned. At that time there were in total six different technologies and we had to make them all. We had inside SIEMENS AG a specific problem, as the Industrial Division realized a the noise sensitivity for the Texas Instruments TTL devices, when applied in an environment of heavy machinery. We developed a noise immune TTL family with eight to ten devices, which were immune to noise, however history had decided in favor of Texas Instruments as one had learned and would rather add three capacitors in the system to avoid noise reaching the meanwhile inexpensive devices from Texas Instruments. Our own direction was not accepted in the market and too expensive. As a result, we had to buy big volumes of TTL devices from Texas Instruments. May I add one more event which fell right into the same time frame with regard to the entertainment industry. In Central Europe and specifically in Germany we had not enough Medium Frequency Radio Stations. We had to support primarily UHF and VHF and in that field there was a battle for Germanium Transistors between PHILIPS and us, where we could build technologically better Mesa Transistors than Texas Instruments. We had better noise immunity and amplification effects, as we were using technology from our tube manufacturing experience. Texas Instruments were applying steam through a hole from two sources. We had a moving fixture from Tantalum and with this we had a worldwide leading market position for about ten years. We had in this interim time also made TTL devices, however never in very high volumes. Then followed the start of memories. After the significant invention of the Planar technology, which has been referred to already, and after the SPUTNIK went into Orbit and the USA pumped Billions into research, the battle really started. I already spoke about the invention of our Dr. Stein, the dynamic RAM, which lead to the memory time and the rise of INTEL, which quickly controlled this market. Right after this the Microprocessors showed up. Right now we have too little time to address this effect for us. Than the Japanese came in, in a style, which surprised us all. I could follow this, as I had educated a Gentlemen from Fujitsu in my laboratory for Germanium transistors, who I asked: how come, you can make these low prices, he said, we have studied your approach and decided to develop new and enhanced processes not in R&D, however in the factory while we watched very carefully the yields, while we increased the volumes. I just yesterday read in a newspaper, that Professor Guenther Schuh is the inventor for Lean Innovation, which means continuous development by increasing the production volume. We have been very angry, that the Japanese manufacturers were beating us on pricing. We were driving innovation be increasing memory capacity from 16 K , 52 K and so on. Finally we were with regards to manufacturing cost below the Japanese, however price wise above. It took us many more years until we understood what really happened. In Japan the prices for buildings, the stock market were so high, that all Japanese Semiconductor Companies NEC, Hitachi, Toshiba did not have to write down their semiconductor manufacturing equipment and investment cost, as their valuations grew and they considered Semiconductors strategic and buried the cost elsewhere as they were big corporations making many different things. They were able to create prices without recognizing the depreciation within their semiconductor organization.
Yesterday Dr. Freiesleben told me that one square meter of land in the center of Tokyo cost at that time five hundred thousand Deutschmarks.
So we and also the American manufacturers were bleeding and the economic basis for a successful semiconductor industry in Europe started to evaporate. We stopped making certain products. However to survive, we in Europe organized harmonization teams between Philips, Telefunken and Siemens, ensuring compatibility between devices we made. This was a very interesting development for the future, as we would only be competitors with regard to quality, reliability and yields, leading to differentiation in pricing. What about Japan ? We organized visits to Japan together with the BMFT
and learned quickly that central organizations like MITI, the Ministry for Trade and Industry, the famous NTT institute had made significant financial contributions to develop new manufacturing equipment for the semiconductor companies. This was an interesting constellation as manufacturing equipment and environments became very costly and more complex. Today one cannot have a piece for semiconductor manufacturing for less than one million Euro. At the time when MITI subsidized the industry heavily, we were still making most of our equipment internally.
As a result we received support from the BMFT, as it became even for SIEMENS almost impossible to
generate the money for the investment cost of a semiconductor fab of one billion. These were the main problems and now we had to answer the question, how and where do we attack the Japanese.
Philips decided not to choose the main products like D-RAMS and SIEMENS decided to compete head on. This was very capital intensive as D-RAMS were driving the manufacturing technology. Ever larger wafer sizes – SIEMENS was one of the first operating 30 centimeter wafers – however we had to realize that we were always one generation behind the Japanese. Then came a turning point. We had a new Head of R&D, Professor Beckurts, who was unfortunately later killed, and a new Board member, who almost always changed in a recession, which occurred in our cyclical industry every four years, Dr. Franz. With him the major projects started, like the MEGA project together with the BMFT, in which all the major contributing companies big or small, who were making equipment for our industry were getting together with PHILIPS, SIEMENS and TELEFUNKEN to aim at jumping over one manufacturing generation to achieve the leading position.
I already mentioned that SIEMENS had several semiconductor sites. A central research lab and separate manufacturing facilities. We pooled all the leading people together in a team to push this MEGA project together with the BMFT so we would at least reach the status of the Japanese industry. However we had to deal with a defect, known as H2 O2, which had to do with cleanliness and cost us nine months of progress. This did not pass unrecognized. It was discovered that our H2 O2 did not work the same way as in Japan . And the clou was, that our H2 O2 was cleaner than the Japanese, which was using “stabilizers”. So we had to import H2 O2 from Japan , as our local support industry was eagerly working to produce more clean H2 O2 as we did not immediately understand the effects of the stabilizers. This cost us nine months. However then we started, as can be seen today inside the facility at Dresden at Infineon, to catch up. I must say, without the support from the BMFT this would not have happened. I wanted to say this, as it sounded earlier pessimistic for the European Semiconductor Industry, however the problems are again visible for this industry everywhere.
Mister Thomas ?
To be well positioned in High Tech markets, one needs more breath than usually expected. In principle it started in 1968, when technological step studies by the OECD were made and showed, the German industry, the European industry in total, was lagging in some very important High Tech fields. That lead together with the book written by Jean Jacques Servan-Schreiber “The American Challenge” – some my memorize this – with the preface written by Franz Josef Strauss – just as a side remark – that the Atomic and Science Ministry transformed to a truly Research Ministry with the dedication to become an Industry Support Ministry. This was in complete contrast to the fundamental philosophy of the Economics Ministry. The conflict was finally only dissolved in the mid nineties. The analysis of the early years of the seventies was: we have Micro- Electronics production in Germany of various technologies. There was Heilbronn with AEG Telefunken. The analysis said: we are not really bad. I remind you of the example of the micro processor. Olympia , a part of AEG, had developed a microprocessor, the CP 3F. This one was in principle, to say this respectfully, as powerful as the one from INTEL. The question was, whether the next step, which was already conceptually done, the CP 5F, should be completed or not. The management from AEG Telefunken, which did not know much about semiconductors, if I may say so, stopped this development. In the USA the team which had the know how together with some technologists and financed by Venture Capital would have left and tried to start their own company.
This was the first difference which we saw in the early seventies. And it was not so, that any of the big conglomerates in the USA , who all have made the effort, survived in the Semiconductor Industry.
They all got out. I think of Rockwell for example, who were at that time in co-operation with AEG and many others like Westinghouse and so on. The new start ups, young fast growing companies and off course, the markets. The fast growth in the USA in the microelectronics industry went hand in hand with the fast growth of the computer industry and the growth of the Japanese microelectronics industry was not only due to memories, which could be sold globally, also due to the growth of the entertainment industry, which had a significant boom. In Germany we studied in the mid seventies, whether we could not do more specifically for the so called “Mittelstand” ( mid size companies) with regard to market development. I do not want to go into too much detail, however relate to one program, into which we at the BMFT invested 450 Million Deutschmark. The special program: Applications with Microelectronics, which was aimed at the “Mittelstand” and assisted significantly in the acceptance and change over from mechanical solutions to electronics. These 450 Million, according to a study by the Savings Bank Organization, generated an investment volume in the “Mittelstand” of over 20 Billion. These were real economic factors. The second was, we were convinced in Germany that new start ups could be generated, so we started a program called: technology oriented start up companies. This started in 1982, somewhat later. This worked with a positive trend until the collapse of the “hype”, the euphoria on the stock markets. We all of a sudden had a Venture Capital Industry in Germany in the nineties. We had a stock exchange in Germany . Companies could be started, one of the reasons why we are not so badly positioned in the bio technology field today. The third had to do with the companies, who were making Microelectronics and those were in Germany mainly German companies – in contrast to foreign companies operating in Germany – SIEMENS and AEG, Heilbronn , and there were at the end Europeans like STM and PHILIPS, STM formed out of Thomson and SGS. And the assumption to do something together with these three big ones in a coordinated fashion, specifically in the area of equipment, was a very high focus. We also would only support any activity if the companies showed and committed a direct engagement. You have already heard from Dr. Rebstock about the investment criteria for a semiconductor company during a downturn and in order to understand that, one needed at least one experience of such a cycle at the Board level. However if the Board was changing faster than anyone could experience a complete cycle, this becomes very difficult.
In such a situation we found the SIEMENS Board. The Semiconductor Member changed constantly. The member needed one year to understand the basics of this industry and had another two to three years, before they disappeared again. The exception was Friedrich Baur, who also disappeared all of a sudden. The problem remained with the ministry to educate the Board and tell the members you must invest or we shall not co-invest with state funds. Such a statement needed to be serious and believable. The fact that SIEMENS would not get any financial support from the Government was not so trivial a task, however was credible and lead to investments by PHILIPS and SIEMENS. This was repeated and I am not ready to further elaborate with regard to “JESSI”, which was born in Paris as “Joint European Silicon Sub Micron Initiative” – the “I” stood for Initiative and was changed later. The construction of this was too complex and did not work. This was rectified luckily during the time when the 300 mm project took off. SIEMENS and later Infineon as well as the other partners, like those coming from Freiberg for whom Mr. Freiesleben did make contributions in a fantastic way,
They all made an investment. And as such I can only make a recommendation to the Quimonda people to read the agreements associated with the government support. Having said this, I shall stop here.
Yes, it has been mentioned again, there were challenges, fall back positions, catch up requirements, government subsidies and at the beginning you Dr. Thomas mentioned that finally three European companies benefited from this and made it into the ranks of the top ten world wide. In spite of all of this, we keep hearing about doubts for the future of the European Semiconductor Industry and if I allow this to pass before my eyes, the requirements for success are three fold: a stable management, a clear product planning in close cooperation with the customers on a global scale and thirdly one needs top talent to realize the customer's requirements. Mr. Kesting, you have only been given the chance to give small comments. You are the youngest member in this environment and I would like to hear from you your perspective with regard to management, availability of funds and to do the right thing at the right time. ST Microelectronics, SGS and Fairchild have been permanently “in” and “out” and meanwhile back “in”. Can you give us a focused historical introduction, how ST Microelectronics did not make it to the top position in the world statistics ?
Yes, I would like to come back to what has been said before, we had to fight big resistance efforts in Germany from the Churches and facts like the following: if I can make 8 tubes much more inexpensive , why should I invest a lot of money to develop chips ? A small example: in the sixties there was VHF television and all of a sudden the idea came up, we need UHF, for which no semiconductors were even considered. Mr. Jack Kilby of TI received the first patent for his integrated circuit, however to support the requirements for UHF the Planar technology was essential and this technology was invented by Fairchild on the basis of the requirements by the military and the space programs inside the USA. One may remember the projects initiated under Kennedy and before.
This created an immensely strong pressure on the Microelectronics industry, as no one had considered flying to the moon depending on tubes. This means the engine to start the fast growth in micro electronics came from the requirements of the aerospace industry. For example, Faichild had developed a UHF transistor the famous 1211. This device sold for 100 Dollars. The Military was ready to pay for this. However the boss of the FCC, the institution in charge of the television industry in the USA , decided this was too expensive. So he was asked what price do you need. He replied one Dollar.
Now the argument at Fairchild was not: if eight tubes are less expensive than the development of such a device we shall not consider this; instead of such a position Fairchild said: OK , we accept the challenge and shall increase the volumes and the yields. This was the birth for UHF in the USA which later also made it to Europe . This much with regard to decision making by management. Fairchild and SGS started as a joint venture. American Management by Objectives, executed by Europeans worked out fine. However, when the first crisis came in the early seventies with a down cycle, this did not hold up. The joint venture broke up and the government stepped in. The revenues at SGS kept going down, the losses climbed. After ten years the Italian Government decided to stop this spiral and engaged Pasquale Pistorio from Motorola. The Italians were fighting the same issues as the Germans, Unions etc. They gave him card blanche and Paquale accepted in '82 and took over SGS. He set the plan to achieve a turnover of 1 Billion within five years. At that time SGS had revenues of 300 million and was a “nobody” on a global scale. He did not make it within five years however he did achieve his target within ten years, however by acquiring Thomson semiconductors from the French Government. This with regard to management and continuity. Pasquale has lead ST Microelectronics, the former SGS and Thomson until about two years ago for about 12 – 13 years. Also within ST there were people who worked there for 40 years – no fast changes – and Pasquale together with his team have moved SGS Thomson, now ST Microelectronics from nowhere to one of the ten largest semiconductor companies in the world and to the top position within Europe today, if I am not mistaken. This means continuity and solidity. The company focused on products and was also lucky. The development of an E-PROM family was a good idea and the production moved to Singapore .
This worked fine until the end of the eighties and all the way through the nineties, as INTEL decided to get out of E-PROMS. This was a big decision in favor of ST, which now jumped to the top position as E-PROM supplier. Again continuity and good management have proven significant factors.
Markets, Management and Finances we heard, must come together and out of this combination new ideas must be created which need to be realized together with customers. We spoke shortly, actually very short about the computer industry, the military technology, however within Europe, based upon the TV and Radio history, the communications and the Telecommunications Industry created an incredible drive and hunger for semiconductors. Where the Europeans not ready to play first fiddle here ?
We did so. We did so with GRUNDIG. GRUNDIG had a genie as head of development, Mr. Mangold, who unfortunately died too early. As mentioned, we were producing normative well defined devices for the entertainment industry, the same by all European manufacturers for all customers and for the same applications, like for example for an amplifier. This was bound to fail. We learned pretty fast and engaged with GRUNDIG to develop about ten VLSI chips, which all were customer specific. We had for quite some time a co-development with GRUNDIG. We only manufactured, what they needed. We did similar things with the Entertainment Divisions of PHILIPS and TELEFUNKEN and other major accounts and we learned, we could dominate these applications for about two to four years. This also worked for PHILIPS, which was stronger attached to the entertainment industry than SIEMENS and may be finally suffered more than SIEMENS, when the industry downturn arose.
As soon as the entertainment industry became too difficult even for GRUNDIG to deal with, we ran down the production and turned towards the commercial markets with Microprocessors, embedded micro controllers and many other devices. The entertainment industry was until the early nineties the engine, which was replaced inside SIEMENS very quickly by the upcoming electronics in the automotive market. This has lasted until today and the tens of years of experience in semiconductor manufacturing with regard to building highly reliable and high quality devices started to pay off.
Today one may wonder if there are not too mane electronic circuits in an automobile. Thank you.
Last remark to which I would like to hear a comment: you just said SIEMENS was the dominant factor in the entertainment industry. At the same time PHILIPS ..(HR) yes OK PHILIPs as well as INTERMETALL, who also played…(HGS) Yes, however were these excellent activities unnoticed or did they lead to jealousy , Dr. Schuetze ?
To this, I can say the following:
A lot has so far been said about the competition between the German and American companies. The American companies had during the period '55 to '75 the big advantage of access to Venture Capital.
'75 with the first oil crisis the money dried up. During the same time the Japanese Industry supported by MITI and the Department of International Trade and Industry pushed forward on to the international markets. In Europe by '84 with the ESPRIT-Program first investments were made to which JESSI was added later. Up to today in about twenty years with a total investment of 60 Billion Euro into these programs no new company like Texas Instruments or INTEL came about. And for this there is a reason, not only one, several. In order to compete against companies like this, one needs as a first point an institutionalized innovation process. I am absolutely sure about this. At Texas Instruments we had the OST System, which stands for Objectives – Strategies and Tactics. This was the backbone for the entire company, a decentralized company structure to force individual initiatives. Thirdly the capital required to finance R&D and new wafer fabs, generated by leading products in major markets and not by government substitutes or similar instruments. Fourth, an employment team, which expresses the spirit of entrepreneurship. This means the “I” always comes second, the company comes first. And that is, what I can report, implemented by Pat Haggerty, the founding father of our company, a set of ideas which are still alive today. And a classical example, that these ideas create success can be measured at something Texas Instruments makes today. Go look at the booth and you will see examples of DSP applications from front to end. This technology was started in 1980, in the eighties, and put to life. Texas Instruments build at that time the first DSP.
We were leading the development and are still leading today. For those who may not understand the background for this. With regard to micro – miniaturization we always look at vertical integration of functions for end equipment. With microprocessors and memories on the hardware side and DSP's on the software side. Today we sell one billion DSP devices per year, a tremendous market. Twenty to twenty five years of continuous investments, standing and know how, which pays off.
Thank you so much. The last word belongs to the senior in this round.
I would like to add a success story to Dr. Schütze's, a very humble, however very significant, which I remembered when Mr. Rebstock spoke about our contacts to GRUNDIG, specifically the Tuner Diode, which was developed by us, meaning you would no longer need inside your radio and TV turning potentiometers as before, just a very small diode to turn to your favorite station. This was a real success story for SIEMENS and I hope we still make this device.
Thank you, and with this I believe we bridged the time from history considerations to today, where “electronica 2006” highlights “Automotive”. We have in between always alluded to the industries of today, whose economic factors would not be accounted for without the inventions of the semiconductor industry. And grace to the people at this round table and many more in this world we have created an industry which started a new age in the fourties and we all can be thankful.
On top of this I would like to be thankful also to the Munich Fair Authorities, who have provided and supported this platform during “electronica 2006” to discuss the history by the gentlemen on this panel.
May I invite you all to an open buffet , which has been prepared across the aisle, so you may take refreshments after you have been listening to us for such an extended time.
Thank you all.