How to Ride
Bull History Index
History of Bull
No. 74, July-August, 1990:
THE VERY INTERNATIONAL HISTORY OF A FRENCH GIANT
A century ago, Herman Hollerith's invention of the punched card machine provided the base upon which three great international companies developed: IBM, Powers (absorbed by Remington-Rand, now merged within Unisys), and Bull . Such was the importance of Bull, both within the overall history of these technologies--first with tabulating systems and then data processing--and in the worldwide and French industrial arenas . The story of Bull is also the story of people, their ideas, their passions, and their personal commitments, all of which were too often sacrificed on the altar of political and financial intrigues, and from which one would hope to learn lessons that might today serve European interests in new technological battles . The text presented here is an extract from the works of Pierre-E . Mounier-Kuhn, instructor and researcher at the CNAM, and first historian to have attempted to gather the complete history of Bull.
FROM OSLO TO PARIS (1919-1931)
On July 31, 1919, Fredrik Rosing Bull (1882-1925), an engineer at the Norwegian insurance firm Storebrand, filed for a patent on a "sorting-recording- adding machine using perforated cards." In August, 1921, the prototype was presented to the board of directors of Storebrand, which agreed to use it. When the head of the statistical services department of a Danish insurance company showed an interest in his invention, Fredrik Bull began producing additional models of his electromechanical statistics machine, and gradually perfected them with a series of improvements. He signed a contract with the Oka company, directed by Reidar Knutsen, which took charge of the manufacturing and marketing costs. Ormestad, a precision machine shop in Oslo, built the machines.
The trade press provided Bull with plenty of publicity. Half a dozen machines were delivered to a number of companies between 1922 and 1925. Their success was due to both the technical quality of the machine (notably its simplicity), and the fact that its availability ended the domination of the Hollerith (IBM) system, bringing prices down, and giving customers a choice. The last factor would long remain one of the keys to Bull's success: users were happy to escape the IBM monopoly, and in particular they appreciated the ability to buy their equipment (IBM required that customers rent its equipment). Success led Fredrik Bull to apply for patents in sixteen industrialized countries, including the United States and Japan. But the inventor died of cancer on July 7, 1925. At the beginning of that year, he had asked K. A. Knutsen, the brother of the president of Oka, to continue his work.
Knut Andreas Knutsen ("KAK") (Oslo 1888-Paris 1983), a hydraulic engineer, had helped Fredrik Bull set up and install the machines sold by Oka. Beginning in 1925, he dedicated himself completely to perfecting and installing Bull machines. A tireless worker, he organized after-sale service and user training, studied competitors' patents in detail, and traveled throughout Northern Europe to repair machines. His activities provided him with ideas for improvements, which included a number of innovations (the horizontal sorter in 1929, and a printing device in 1930).
At this time, two new actors came onto the scene: one Swiss, and one Belgian. Doctor Marchand, a manager of a Swiss insurance company, having met Bull in Oslo, bought a Bull tabulator and a Bull sorter, which were delivered in 1926. But Dr. Marchand's ideas went beyond his company: he saw in the manufacture of tabulating equipment a way of providing his country with a modern industry, and a way to end American domination of Europe in the field. He contacted Oscar Bannwart, president of the Swiss company H. W. Egli--which at the time was known for its Madas and Millionnaire calculators--and suggested that he manufacture Bull machines.
1931: The Move to France
The principal instigator of Bull's expansion throughout Europe at this time was Emile Genon, a Belgian who sold Elliot-Fisher and Underwood calculators. In 1927, after meeting with Bannwart, Genon bought the rights to Bull patents for ten European countries. In 1928, he persuaded an initially hesitant H. W. Egli to acquire industrial rights for patents awarded to Bull and Knutsen outside of Scandanavia. In December, 1929, the first machine built in Zurich was delivered to Sandoz Laboratories. It rapidly became obvious that the company would have to set up shop in a European country with a wider potential market than that of Switzerland. France was chosen over Germany, in part because of the personal preferences of Knutsen and Genon, and in part because, in France, the competition's situation and the patent law would give Bull an advantage over its rivals. Despite the presence of IBM (operating in Paris since 1914) and Samas-Powers, France was still an untapped market for tabulating equipment in the 1930s. In addition, for Bannwart, manufacturing in Paris would reduce fabrication costs, because money, manpower, raw materials, and business expenses were cheap in France, thanks to the financial housekeeping carried out by Raymond Poincare.
In the month of March, 1931, the H. W. Egli Bull company, a French corporation with majority Swiss ownership, was founded in Paris by three partners: the H. W. Egli company of Switzerland; Bull A. G., founded the preceding year in Zurich by Genon, which contributed its embryonic international sales network, and new patents acquired from K. A. Knutsen; and finally, ATEIC (Association Technique d'Etudes Industrielles et Comptables, the Technical Association of Industrial and Accounting Studies), which had distributed American accounting machines up to that time. Under the direction of Henri Vindevoghel, ATEIC would sell Bull machines in France. One of ATEIC's contributions was the Atemeta facility of 9000 square feet in Paris, which was converted to manufacture tabulators, sorters, and perforators, using the modern machine tools with which it was already equipped, as well as manufacturing tools developed in Zurich. In March, 1931, H. W. Egli Bull--with fifty employees--moved into this building, at 92 bis, avenue Gambetta, in Paris. The corporate offices remained in this building until 1983! The second phase of Bull's settlement in France took place in 1932, in response to an attack from Remington Rand, the owner of Powers. From April, 1931, Mr. Rand had been talking to H. W. Egli about buying the latter's rights to Bull machines; Remington directors visited the Paris facilities in July and, in November, 1931, two American negotiators were sent to Zurich.
In the meantime, and since 1930, a group of Frenchmen had also been trying to develop a tabulating-machine industry in France. Its most active members were two graduates of the prestigious Ecole Polytechnique: Elie Doury, a Powers machine salesman, and Georges Vieillard, who used these same machines at the Bank of Alsace-Lorraine. The essential idea of the project have a group of users within a syndicate finance operations. The two Polytechnique alumni met Emile Genon at a time when Genon was looking for capital to permit H. W. Egli Bull's development, and an agreement was quickly reached. But first, Remington Rand had to be dealt with.
The ship that brought Remington Rand representatives back to the United States was scheduled to arrive in New York on December 7. The French thus had the duration of the voyage to act. On December 2, Vieillard telephoned the president of H. W. Egli, and proposed to buy half of his interest in H. W. Egli Bull, promising to send him 50,000 francs the same evening as an advance deposit (all the money Vieillard and Doury had at the time). In order to meet this commitment, a legal entity and funds were needed. On that same day of December 2, 1931, a corporation, the "Tabulating Equipment Users' Syndicate," with authorized capital of 50,000 francs, was established; the first general meeting took place in the presence of the seven people legally required; and the promised 50,000 francs were sent to H. W. Egli.
On December 11, Remington Rand cabled its approval for the purchase of interest in H. W. Egli Bull, and the president of the latter company suggested to Vieillard that he take back his offer. Vieillard refused. And on December 20, 1931, the board of directors expressed its intent to increase capital, giving preference to the Users' Syndicate. The Syndicate gradually took control of Bull away from H. W. Egli, and soon became the majority shareholder. In April, 1932, this was a fait accompli: the board counted eight Frenchman, two Belgians, a Swiss and a Norwegian. The president was Col. Emile Rimailho (1864-1954), a renowned artillery officer who had participated in the design of the 75 cannon, and had later moved into the industrial sector, where he had become a specialist in the scientific organization of work. In 1933, the H. W. Egli Bull company changed its name to Compagnie des Machines Bull (Bull Machines Company), which it has conserved to this day.
Until 1935, Bull--the "Bull Machines Company, maker of statistical and accounting machines"--was in a difficult phase of stabilization.Thereafter it began a long period of instititutional stability and revenue growth, founded on a continually improving line of tabulating-machine products; the coming of electronics both accelerated this expansion and shook its foundations. The tabulating-machine industry between the two wars generated a great many patents. Inventors like Bryce at IBM in the United States, Tauschek in Germany, Foster (the creator of the alphanumeric printing tabulator, announced in 1921) at Powers in England, and Knutsen and Clouet at Bull (as well as innumerable and unknown technicians who made small improvements to the machines), simultaneously motivated by a desire for technical perfection and commercial necessity, brought high performance to electromechanical tabulators and made them highly adaptable to customers' needs. The ever-growing domain of tabulating-machine applications that resulted from this may be be inferred from the vocabulary, which replaced "statistical machines" with "management machines."
The first Bull machines built in Paris were designed by K. A. Knutsen, who had been perfecting Fredrik Bull's inventions since 1925. In particular, he added a major innovation: numeric wheel printing. The central idea in the design of this printer was the replacement of the print bars, which had supported characters in tabulating machines built up to that time (especially IBM tabulators), with wheels. With characters set on wheels that always turned in the same direction, the print assembly was lighter and high-inertia movements were eliminated. In September, 1931, the first printing tabulator was delivered to the French Labor Department's Unemployment Service. It included 60 wheels (numbers only) and printed 120 lines per minute with excellent print quality. No other machine could even come close to this speed.
It was decided that the wheel technology would be immediately extended to alphabetic printing. A 30-wheel alphabetic printer was developed, which could be added to the 60-wheel numeric printer. This new printer was made available at the end of 1932. The speed had been improved: 150 lines per minute. This speed would not be reached by competitors' machines until 1949, with the development of the IBM 407, wherein wheels replaced bars. In 1934, Bull announced a printer with fully alphanumeric wheels driven by the contents of the card. The first machine was shown at the Office Machines Exposition in October, 1934. This was followed by miscellaneous improvements, but Bull's printer was already a major player in 1934. This same printer remained in production, essentially without modification, until 1968!
Meanwhile, the Company's development encountered numerous obstacles, both inside and outside the firm. The serious financial problems of the Company slowed the rationalization of the manufacturing facilities, the debugging of the machines, and their adaptation to users' needs. A report prepared in 1933 by Georges Vieillard noted frequent mechanical problems, machine malfunctions, machines delivered to customers without adequate testing, etc. In large part, these were the "childhood diseases" of a company in its infancy, learning its lessons in the demanding field of precision electromechanical engineering. Large expenditures were required for maintenance, taking money away from research and development; machines had to be made to work at any price, in order to maintain the confidence of the first customers. This meant that equipment that had already been delivered sometimes had to be taken back to the factory for modification or replacement by updated machines. These obstacles did not discourage the small group of the "users' syndicate," which maintained its confidence and its financial support for the Company. Some of them thus accepted a double risk, being both customers and shareholders.
From 1934 on, the situation improved: technical problems dropped by 70%, compared to the initial deliveries; the first operational profits appeared (more than 800,000 francs); and the Company controlled over 15% of the French market. In 1935, Bull, with more than 60 sites installed, surpassed SAMAS and became the main competitor of IBM in France. This growth was all the more remarkable when one considers that it occurred during an economic crisis (automobile production in France diminished by 35% between 1929 and 1935), and that, at a time when France was said (often in so many words) to be a country lacking in intiative, innovation, and export potential, few companies had the means to equip themselves with punched-card machines. In the mid-1930s, the punched-card industry employed less than a thousand people in France.
The years from 1935 to 1937 were decisive. Innovation required heavy investment, and discussion took place with the French government concerning possible financial support for research and development. Simultaneously, the Board of Directors instructed E. Genon to investigate possibilities for distribution license agreements with various firms in the United States. Among others, he met with Thomas J. Watson, the president of IBM, who offered him a "friendly cooperation agreement." But the company preferred to ask for help from the French government. The government's decision was still not forthcoming, and so Genon, without obtaining approval from the board, sold a majority interest in Bull A. G. (the marketing company for Bull machines), which he directed, to IBM. In this he saw a means to obtain a "cease-fire" in the patent war between IBM and Bull, and the opportunity to "develop Bull's business on an international level with support from an American company." Georges Vieillard, intransigent, gave Genon a choice: Bull or IBM. After ten years of intense activity, much of which proved later proved decisive, Genon left Bull, and Bull management thereafter felt that he had "betrayed" them. New players then arrived on the scene: the Callies- Aussedat family.
The Aussedat Paper Companies were Bull's tabulator card supplier. Since 1932, they had made significant investments in this area, and they were represented on the Bull board by Jacques Callies. The threat of IBM absorbing Bull worried Aussedat, because IBM required that its customers buy cards only from IBM. The directors of Aussedat, which was linked to the Michelin family, examined the problem, and the advice of the "Elder," Edouard Michelin, prevailed: "Just as the takeover of Citroen by General Motors had to be averted, Bull must not fall into American hands." And, since the government had still not acted, the Callies family decided to increase its financial stake in the Company. It took up management of the company via Jacques Callies, a former officer and alumnus of the Ecole speciale militaire de Saint-Cyr, who was named delegate administrator of Bull in December, 1935, and then president and CEO. He continued in this capacity until his death in November, 1948, and was succeeded by his brother, Joseph, who was also a graduate of the Grandes Ecoles, and an engineer at the Aussedat Paper Mills before becoming an engineer at Bull.
The team that was to manage and expand the Company for nearly thirty years was now in place. And so was the form of family- style capitalism that guided it until 1964 (with the Callies family holding 55% interest in the company), with its guiding philosophy that characterized the Company: a feeling that each employee is important, a kind of social paternalism, a feeling of being "the best," and an intense concern for customers. Like Michelin, whose corporate culture seems to have influenced Bull, the company cared little for official titles and rigidly defined organizational charts. A good example of this is provided by the the case of Roger Clouet, a young employee recruted by the company around 1933. With a considerable gift not only for mechanical design but also for adapting machines to customers' needs, Clouet effectively became the technical director of Bull in the 1940s, where he trained some of the new generation of engineers recruted after the war. Management, from the Callies on down, created an image of managers driven by a sense of duty, paying their own way, "arriving every morning at eight o'clock in a car they bought themselves, without a chauffeur, and without a reserved parking place." An interview with Joseph Callies (in Enterprise, No 117, November 30, 1957) clearly summarized the family spirit of the Company: "like pioneers, we take care to avoid bureaucracy and impersonal treatment. We don't publish official organization charts, in order to make it easier to adapt structures and to make promotion of the best employees easier." To this was added a bonus system to provide compensation for personal effort. Callies insisted on "the innovative character of our activities" (the average age of employees was 35 at the time of the interview) and "patriotism" with respect to the Company (the most heinous crime, never actually committed, would be to quit Bull for IBM) and France: the fact that Bull was a 100%-French firm was a frequently-used argument, especially in negotiations with government customers.
From the beginning, the unique constraints of Bull's profession were visible as well: free trials, rental agreements, the possibility of purchase with monthly payments, the seasonal nature of sales, the slowing of sales when a new model was to be announced, and the return of hardware.... The principal difficulties of the Company continued to be financial, linked to a certain under-capitalization. In addition, the government had still not approved the aid to the Company that the latter had requested on several occasions.
In the mid-1930s, Bull, having earned its right to exist, was solidly established. The Company had gained a substantial market share, not only in France, but also in Belgium, Switzerland, Italy, Argentina, and Scandanavia. Its industrial capacity was significant, and it had several hundred employees working in 25,000 square feet of facilities. It produced around three systems per month and regularly increased its production capacity.
Bull's tabulator, having outgrown the problems of its youth, was the fastest on the market. A perforator connected to the system, capable of punching an entire card at a time as well as all results, increased performance even more. The 80-column card with rectangular perforations was adopted. The printer was improved by adding a paper eject function, and a ribbon replaced the inking roller. At the same time, the subtracting tabulator (1936) was perfected, which opened up the accounting market for Bull--including the banking market, after development of a special interest-calculating tabulator. Soon, banks represented a third of Bull's French clients. In 1939, Roger Clouet designed the BS 120 tabulator, with independent cycling, which would be one of the prime movers of Bull's expansion for twenty years. These technical advances inspired two remarks. The sales experience of the Company acquired up to that time demonstrated that it was not sufficient to sell the few machines invented by Fredrik Bull and K. Knutsen; a full catalog of machines was necessary, including "auxiliary equipment"--punches, calculators, etc.--which prepared or accelerated the work of the tabulator. The Company was thus pressed to diversify its production. In addition, the appearance of Bull machines on the market had intensified the competition, and had forced competitors to increase their R&D, and Bull in turn had to keep up the pace. Result: Customers were able to acquire better and better equipment at lower and lower prices--the price of hardware had dropped by 30% to 40%, and that of cards by 25%.
The war, despite the confusion that it caused, didn't erase the results of preceding efforts. Requisitioned in 1939, the factory on Gambetta Avenue was forced to move to Lyons on June 11, 1940, on the order of the Ministry of Armaments. Work resumed on August 8. Thirty-seven employees were prisoners (more than 10% of personnel). The occupation imposed restrictions: difficulties in obtaining raw materials and finished products, electrical power outages, air raids, travel restrictions, special procedures made necessary by the overall penury and by the demands of the occupation authorities, and ruses with these same authorities in order to avoid conscriptions for forced labor, or to benefit from German orders and study requests, while delaying the actual responses as long as possible. Despite all this, the Company's revenues continued to increase. The new BS 120 helped to support this expansion. After the Liberation, a number of Bull directors, known to be marechalistes because of political conservatism and loyalty to veterans of the First World War, were subjected to investigations meant to purge collaborators, and they survived unscathed; true, the Company had been obliged to deliver machines to the Germans, but it had also protected Jews and labor-force insubordinators.
After Jacques Callies' death in 1948, CEO Georges Vieillard (1894-1974) reaffirmed his authority. Accepted at the esteemed Ecole Polytechnique, Vieillard performed brilliantly as an artilleryman during the war, and began his studies only after the Armistice. After 1920, he held positions in a number of companies, in particular at Real, a French distributor for various American companies, where he became head of the calculating and accounting machine department. He developed a repair shop that became the largest in Paris of its kind (80 workers), and designed several calculating machines (one of them is described in an article in La Science et la Vie in 1922), although he continued to specialize in scientific methods for organizing work. Vieillard is typical of a generation of alumni of the Polytechnique (theoretically a military school) who, being "more and more reluctant to take up a military or government career, and preferring instead the private sector (more than 1400 alumni chose the private sector between 1919 and 1924), have renewed the upper management of the larger corporations." His contribution to Bull's expansion is considerable, partly because of his financial expertise, and partly because of his leadership ability. After his trip to the United States in 1948, he made sure that Bull always had a few American "allies" to help in the struggle against IBM. From 1950 to 1960, this meant Remington-Rand; later came RCA, and finally General Electric.
From the end of the war through 1961, the Company enjoyed extraordinary growth and prosperity. In 1948, Bull surpassed IBM in the French market, with 385 systems installed. In sixteen years, the number of employees had increased by a factor of ten. This was mostly internal growth, attributable to sales and product development, along with absorption of some of Bull's subcontractors. This period was simultaneously the golden age of the tabulating machine, and the beginning of the conversion to electronics. One of the most striking features of this dynamism was its international nature. From 1947 on, export activity, which had been interrupted by the war, resumed with a vengeance. Over the next fifteen years, Bull's international network, which was already in operation in the 1930s, would undergo considerable expansion, and would become one of the great strengths of the company. Thus, in Belgium, SOMECA, which represented Bull A. G. in 1930, became the Societe belge des machines Bull in 1942. In Switzerland, Endrich A. G., a Bull partner since 1930, became the subsidiary Bull Lochkartenmachinen A. G. in 1947. In 1949, an agreement was reached with Olivetti, creating an Italian distribution subsidiary, Olivetti-Bull. CMB sold part of its interest in this last company in 1962, and relinquished the remainder in 1965. Between these dates, Bull had been largely responsible for the spread of EDP in Italy. In the Forties, Bull set up shop in the Netherlands, Germany, and South America. Bull also addressed the Anglo-Saxon markets, but did not establish subsidiaries in these markets until much later, since its competitors were on their own turf in these countries, and since exclusive-market agreements with Remington-Rand and British Tabulating Machines had been signed (these two companies were thus excluded from France). The exclusive-market agreements were both an advantage and a handicap, since North America represented 80% of the worldwide information-processing market at the time. In 1956, the Soviet market opened up to Bull. In 1960, Bull entered the People's Republic of China. In October, 1962, Bull signed an agreement with Mitsubishi giving the latter exclusive sales rights in Japan (Mitsubishi also acquired a Gamma 60 at this time, followed later by Gamma 10 systems). In 1963, Bull Corporation of Japan opened its doors.
The number of employees in Bull branch offices and subsidiaries increased sixtyfold in sixteen years: there were 4,000 employees in 1964. The Bull sales network at the time covered more than 42 countries. Between 1950 and 1965, Bull earned between one third and one half of its revenues from export business--exceptional performance for a French company. This success was due in part to the Company's policies (notably those of M. Sanson, export director from 1945 to 1964), in part to its multinational origins in the 1920s, and in part to the measures taken by the French government to protect the French market and encourage Bull's export activity. In 1963, Bull held 1/3 of the French market, and 10% of the European market. Tabulating systems contributed to the overall phenomenon which, after the devaluation of 1959, produced a surplus in exports over imports; in 1960, for the first time, France exported more punched-card equipment than it imported (2,397 systems exported versus 1,457 imported).
This opening of international markets helped Bull deal with the great technological innovation of the Fifties. The classic punched-card machine was already part of an outdated system, built upon electrical principles developed around 1880, and upon precision machine components, developed in the nineteenth century for typewriters, sewing machines, armaments (Remington), and for clock mechanisms in even earlier times. Bull management realized the potential of electronics early in the game, around 1948, the same year that IBM delivered its first tube-equipped Model 604 in the United States. At this time a change occurred in the Company's technical recruiting policies. Up to then, Bull had recruited technicians schooled in mechanical and electrical areas (tabulating systems, electric clock systems, calculators) and a few engineers with widely varying educations, such as former officers and a few university graduates. Everyone initially spent six months at "Bull University," where they learned the principles of tabulating systems and punched-card systems. From 1949 on, the Company began to hire many engineers (particularly from the Ecole superieure de mecanique et d'electricite Sudria, and from the Institut Polytechnique de Grenoble), and a new type of specialist started appearing at Bull: the electronic engineer, a graduate of the Ecole superieure d'electricite or the Ecole superieure des telecommunications. Some of them came from companies where they had already acquired experience in EDP; for example, from SEA or Univac. With them they brought a new technical culture, and work methods imported from other professional environments. One of the first things B. Leclerc did when he arrived at Bull in 1949 was to equip his lab with the electronic measurement instruments to which he had grown accustomed while working for the telephone company.
The arrival of electronics led some Bull employees to establish contact with people and activities that had previously been far removed from their work, namely, those of the scientific community. A National Electronic Calculating Center functioned beginning in 1951, as a demonstration center for Gamma computer systems. Directed by Philippe Dreyfus (who created the French word for EDP, informatique, in 1962), the NECC offered "the services of a calculating center equipped with Bull electronic equipment to companies and laboratories who lack their own high- speed calculating equipment."The first users were the astronomical observatories, the physics laboratories of the Ecole polytechnique and the Ecole normale superieure, and so on. The flight simulations of the Caravelle were carried out on Gambetta Avenue. However, products with electronic components were also designed, in close cooperation with customers, for Bull's main market, information management, and within the framework of the medium known best by Bull and its customers: the punched card. This explains the success of the Gamma 3 (1,200 installed in France and elsewhere from 1952), connected to the BS 120 tabulator.
Technological research at Bull was carried out mainly as part of a race with the competition, especially IBM. New products were developed partly as responses to IBM's innovations. Thus, the Gamma 3 calculator (1952) was a response to the IBM 604. In 1956, after the announcement of the first SEA computers and the IBM 650 in France, Bull announced the Gamma with "drum extension" ("ET" in French) its first stored-program computer. In the other camp, following the success of the Gamma 3, IBM teams in France and Germany started projects that would lead to the IBM 1401. Bull answered with the Gamma 10, among others. In the meantime, the Gamma 60, announced in 1957, was an attempt to compete with the IBM 700 Series supercomputers. In this race to innovate, the machines began looking more and more like computers, and less and less like tabulators. The magnetic drum of the Gamma ET allowed for a stored program. This was the bridge between the calculator and the computer. The relationship between the electronic system and the tabulator that served as an input and output unit changed. The tabulator lost its central role and was gradually replaced by a card reader, a printer, and a sequence of instructions in a program--a spreadsheet! This transition, which lasted about a decade, represented a fundamental change in Bull's products, and in the company's technical and cultural identity, with respect to its competitors, its suppliers, and its markets.
The high point of tabulating system sales was reached around 1960. Ten years earlier, it was unusual to find punched-card equipment in a company of less than 1,000 employees; by 1960, this type of equipment was cost-effective in small businesses of 100 employees. This brought about a spectacular growth in sales and rentals of tabulating machines, and for their new competitors, the computers. But tabulating systems could not follow the trend very long; the technology had reached its maturity between the two wars, and it had now reached the saturation point. Only minor improvements could be made. Attempts to push tabulating machine performance to its theoretical limits failed. At Bull, this happened with the "300 TI" series (1960); comparable results were obtained with the "Series 3000" at IBM, and the "Samastronic" at Powers-Samas. These machines, which were both unreliable and unprofitable, were withdrawn from the market shortly after their announcements.
During this time, Bull enjoyed a period of remarkable growth. The Company began a massive hiring program, especially after 1956, in order to deal with the market explosion and to produce the Gamma 60. Bull had three plants in 1953, two in France and one in the Netherlands. Between 1953 and 1962, seven other factories were built or purchased, five for tabulating equipment, and two for electronic systems. These facilities represented some 1,880,000 square feet of manufacturing space in 1962. The Angers factory alone, dedicated entirely to electronic systems, occupied 680,000 square feet. In 1960, the Compagnie des Machines Bull was the world's second-largest manufacturer of electronic information processing systems. It boasted an installed base of over 800 electronic systems, plus 3,453 "classic" tabulating systems, of which over half were exported. With 14,000 employees in France, and revenues of 16,022,000,000 francs [about $160 M in 1990 dollars--Trans.] in 1959, the Company was still far behind IBM. Big Blue was thirty times larger than Bull, and spent three times the total revenue of Bull on R&D alone; it also enjoyed lucrative contracts with the U.S. government. That did not prevent the Compagnie des Machines Bull from feeling a certain euphoria when the price of its shares, which had skyrocketed, placed it in ninth place among French corporations at the end of 1960.
In 1962, CMB's profits began to drop. The following year, Bull was in the red, with losses of 85,000,000 francs [about $80 M], and with no hope of improvement over the short term. Approximately 650 persons were laid off in 1964. Overall, it may be said that Bull was a victim of a growth crisis (debt incurred to finance increased research and production capacity, and to finance the heavy investment in technical training that ordinary schooling did not provide; and the initial cost of the machine rental plan) and a technological adaptation crisis concerning its new products (the Gamma 60, a technological masterpiece, was a commercial failure; the poorly designed 300 TI in particular never managed to succeed as a mid-range machine). In 1960, IBM had presented its small management computer, the 1401. The success of this machine surprised everyone--including IBM--and left the competition (including Bull) empty-handed. Each 1401 delivered to a customer replaced one or more tabulators; this computer, easy to program and equipped with a fast, excellent printer, effectively sealed the fate of classic tabulating equipment. The IBM Series 1400 was soon equipped with magnetic disks, a technology that none of the European vendors had seen coming. In order to save its market share, Bull marketed four new computers at the beginning of the 1960s: the Gamma 30 (1961), actually a RCA 301 built under license, and of which 250 would eventually be sold or rented; SEA's CAB 500, for which Bull handled distribution in 1963 and 1964; the Gamma 10, a small "compact" computer using punched cards, of which 1,600 were eventually sold following its introduction in 1964; and finally, the Gamma M40, which was a surprising entry in the industrial automation field by Bull. Twenty M40s were installed from 1963 on, including one at the Feyzin refinery.
This effort was not adequate, or was too late, to keep Bull afloat in the face of attacks from IBM, despite a R&D budget of 13% of revenues. Many factors intervened: self-congratulations founded on the success of the Gamma 3, poorly adapted "family" management techniques, management that was too unaware of the potential of computers and the requirements of EDP, poor links with the university community, a lack of interest in software, and a poor reception for EDP in the French market (including the government); and Bull was caught off guard by the suddenness of the EDP revolution.
Several solutions to end the crisis were proposed and considered, simultaneously. They involved the French government and a number of companies, including several banks and several foreign firms. The government intervened beginning in 1963. Its main concern was to maintain a French computer industry, in order to have readily available computing power for its nuclear energy programs (civilian and military). In 1963 and 1964, under pressure from the French Ministry of Research and Industry, Bull signed a series of cooperation agreements, first with SEA, then with a consortium formed by CGE (Compagnie generale d'electricite), CSF (Compagnie generale de telegraphie sans fil), and the Banque de Paris et des Pays-Bas. The idea was to create a common computer subsidiary. This "French solution" was to include 210,000,000 francs [$120 M] in government R&D contracts during the five following years (this project resembles the Computing Project or Plan Calcul, launched three years later). In fact, the three companies entered into these agreements with the greatest of reluctance. This attempted solution was abandoned in mid-1964.
At the same time, Bull had asked various European banks for loans via its subsidiairies, in order to avoid seizure of CMB itself. The French Finance Ministry refused the necessary authorizations. Bull then looked to the centralized Credit National. The Finance Department delayed the authorization of the loan for months.
The third solution would be an agreement with a foreign company that had both the necessary technology and the necessary commercial and financial means. An agreement with ICT in Britain (later to become ICL) was considered--although ICT had the same problems and was in the same situation as Bull. The most workable solution, particularly in the opinion of Georges Vieillard (without a doubt the key man in the affair), was an agreement with General Electric. Discussions between the two companies had begun in 1962. General Electric proposed purchasing an interest in Bull on several occasions (initially limited to a modest 20% of Bull's capital), but these were rejected by the French government. Finally, in the summer of 1964, as General Electric was preparing to buy into Telefunken, the French government reacted and gave the go-ahead.
General Electric (260,000 employees in 1963, with revenues of five billion dollars) had entered the EDP arena in 1956. It took control of Bull (with 51% of the company) on July 1, 1964. In 1967, General Electric increased its interest in the company and replaced part of the American management team of Bull-GE, claiming that it had failed to improve the management of the company. Those favorable to Bull, the bullistes, had to fight to persuade GE to keep French-designed products in the catalog. They succeeded for the Gamma 10, and for the Gamma 55, a small office computer using punched cards. More than 800 units of the latter model, as well as its disk version, the GE 58, were sold in North America. In contrast, GE management killed several projects proposed by Bull researchers--in particular the Gamma 140, a competitor of the GE 400--whereas an Olivetti project was adopted under the name GE 115. At this point, many engineers quit the company in order to join the Plan Calcul project and CII, created in 1966.
Nevertheless, the contribution of General Electric to its French partner should not be underestimated. First, the contact with General Electric caused a new industrial culture to take shape at Bull; this included the development of modern management methods, communication and marketing techniques, and industrial planning systems, all with a worldwide perspective. Second, Bull engineers had access to all the civil research data of General Electric, including that dealing with time-sharing and communications operating systems--an area in which General Electric, which had participated in military research along these lines at MIT, was a world leader. In 1968, Bull-GE opened the first time-sharing center in continental Europe and committed itself to developing networks, an area that is still one of Bull's strong points. Another consequence of technological cooperation within Bull-General Electric was the perfecting of a "micropackaging" technique, starting in 1968. This technique would be adapted to mass production by CII-HB, and would later be used by Groupe Bull for the DPS 7.
In 1970, Honeywell Inc. acquired most of the computer manufacturing operations of General Electric. Bull-GE became Honeywell-Bull, the main subsidiary of the Honeywell Information Systems group. A new CEO took the reins of management in 1972: Jean-Pierre Brule--a former employee of "Big Blue," and one of the very rare defections from IBM to Bull! Brule had spent twelve years with IBM, where he was director of the military division, then director of Defense Systems Europe at IBM World Trade, before joining Bull-GE in 1967 as executive vice-president.
After 1975, Bull began another period of intense growth activity. Most of this was due to acquisitions of other firms and mergers. The first merger was with CII, which had been, since 1967, the major industrial component of the Plan Calcul, and which covered 12% of the French market. This operation caused several problems. From a political standpoint, the integration of CII into the French-American group seemed, in the eyes of many, to spell doom for the crusade to maintain national technological independence that had been led since 1966. Could this merging of firms really be the best solution to ensure economies of scale, or to guarantee the innovative capacity so essential to competitive competence in the EDP market? Wasn't it a serious error, with respect to the European partners of France, to torpedo Unidata, in which CII, Philips, Siemens had participated since 1973? Wouldn't it have been ultimately cheaper to support two competing "national champions," rather than one? Debate still rages on this merger, as has become apparent in recent conventions on EDP history, because of its political aspects, and because of its high cost. The second problem concerned image and dynamism as it related to sales. Along with CII, Bull inherited the role of "national champion": after the kickoff of the Plan Calcul, the government had praticed preferential acquisition and protection procedures in favor of CII. The agreement signed between CII-HB and the French government in 1975 guaranteed contracts worth four billion francs between 1976 and 1979 [about $2 billion], plus significant subsidies (seven billion francs). Nevertheless, a kind of competition still existed: in 1983, the installed computer base of the French government was 53.5% CII-HB, and 18% IBM. This arrangement ended in 1986, when the government of prime minister Jacques Chirac decided to allow unrestricted acquisitions for government agencies and nationalized firms.
Finally, after becoming, or re-becoming, the major French computer company, CII-HB had to solve the serious problem of heterogenous product lines. First, CII-HB had to confront all the typical problems of any industrial merger, for example the definition of a common identity for teams with diverse origins, and it had to introduce consistency into its strategies and products: in 1977, 45% of its products came from Honeywell Information Systems (which itself marketed several different product lines); 32% from Honeywell-Bull, inherited from Bull-GE; and 23% from CII. In order to merge product lines, a Unisys-style program was undertaken in 1976.
Second, the agreements reached in 1975 and the "plan peri- informatique" of 1977 decreed a separation between the "large- scale" EDP assigned to CII-HB, and the "small-scale" EDP (minicomputers and peripherals) assigned to Logabax, Intertechnique, Thomson (SEMS), or CGE (Transac and Sintra). It rapidly became clear that CII-HB would not limit itself to "big" systems, and it began investing in minicomputers and peripherals. For example, CII-HB marketed a Honeywell minicomputer under the name Mini 6 and acquired, in 1978, R2E, which had been producing Micral microcomputers since 1973. In 1983, CII-HB took control of SEMS and Transac and thereby enlarged its activity in minicomputer systems and office automation.
Bull's partners changed again at the start of the 1980s, because of political changes in France, and because of the gradual withdrawal of Honeywell from the EDP industry. In 1983, the French government acquired a 97% interest in CMB. A new management team, headed by Jacques Stern (a graduate of the Ecole Polytechnique, the Ecole superieure aeronautique, and Harvard, and founder and CEO of the Societe d'etudes des systemes d'automation) and Francis Lorentz (a graduate of the Ecole des hautes etudes commerciales, the Ecole nationale d'administration, and former executive vice-president of the Societe lyonnaise des Eaux), was formed to direct "Groupe Bull." The new name emphasized corporate unity and recalled the company's origins. Three major objectives were defined for the period that followed: regaining a worldwide presence; improving competitive competence; and forming partnerships, with participation in European projects.
In 1987, Honeywell-Bull Inc., was created by Bull (42.5%), Honeywell (42.5%), and NEC (15%), from Honeywell Information Systems. Two years later, the Compagnie des Machines Bull took control of Honeywell-Bull, with 65.1% interest. In 1989, Bull bought the EDP operations of the American company Zenith, one of the major microcomputer vendors (nine billion francs in revenues). In a few years, via external growth, the group, initially somewhat "Franco-French" has again become a major world player, with 70% of revenues earned outside France. With 47,000 employees and revenues of 41 billion francs [$US 8 billion], it is today among the top ten EDP vendors in the world.