The DRTE Computer (page 1)
The story of the DRTE (Defence Research Telecommunications Establishment of the Defence Research Board) computer has its "raison-d'être" in the story of transistor and the diffusion of this new electronics concept in Canada. This story is a good illustration of the symbiotic relationship that developed between solid state electronics and computer technology. The DRTE computer group also presents us with an interesting case study in the effectiveness of government research laboratories in the diffusion of leading edge technologies. The invention of the point contact transistor occurred at the Bell Laboratories in New Jersey. Though any invention takes a extended period of time to create, December 23, 1947 is given as the official date of birth of the transistor. In 1956 Bardeen, Brattain, and Shockley were awarded the Nobel Prize in physics for their pioneering work on the transistor. The transistor is one of the most significant discoveries of our century. The transistor opened the door into a whole new realm of solid state physics research, and radically transformed electronic circuit design.
The development of the transistor, like the development of atomic bomb and nuclear energy, marked the beginning of a new relationship between science and technology; theoretical science had become an essential and integral part of technological development. The transistor could have only been developed by a team of theoretical physicists familiar with the intricacies of quantum mechanics. The image of the individualistic, empirical inventor working from his little workshop was now replaced by a multi-disciplinary team of theoretical scientists working in well equipped laboratories.
In the early 1950's the transistor was not universally acclaimed as an electronic revolution. "As late as 1953, there was still suspicion among some leading British physicists that the transistor was no more than a piece of good publicity for Bell" (Braun,1982:52). In the U.S. many people in the electronics industry approached the transistor with caution, and in some cases with open hostility.
One of the factors contributing to this attitudinal resistance to the transistor was the way Bell Telephone promoted the transistor. From the very beginning, Bell Telephone had publicized the transistor as a solid state replacement to the valve. Most discussions in the early 1950's, concerning the merits of transistors, usually came down to analogies between valves and transistors. This kind of thinking diverted the engineering community's attention away from the entirely new sort of electronics implicit in transistor technology. It was only with a new generation of engineers, who were not rooted in valve technology, that the true potential of solid state electronics was fully realized. This illustrated by the fact that most of the technical and market innovations concerning the transistor were carried out by firms that didn't have a tradition in valve technology. ()
The early germanium point contact transistors were extremely unreliable. In 1952 the more promising germanium junction transistor was produced, but it also had important limitations. Going fiom laboratory transistor to commercially producing large quantities of transistors proved to be quite a formidable challenge. "It was difficult to design a transistor to give even an approximation of the characteristics required, and even harder to produce two transistors with the same characteristics" (Braun,1982:49). ()
There was also the question of economic feasibility. As a replacement to the valve, the transistor proved to be too expensive. "..In late 1953, the best transistors cost about $8 or about eight times the price of a valve". Leaving the problems of quality control, reliability, and performance aside, the price of the transistor didn't make it a very competitive rival to the valve. In this regard Braun(l982) presents a very interesting observation made by an electronics expert of the period:
"For a decade or so, from 53 to 63, we had no choice but to go with vacuum tubes because they did a better job and up until that time they were cheaper. You could get a perfectly good vacuum tube for 75 cents" (Braun,1982:50). In fact, in 1957 tubes were still outselling transistors in the U.S. by a ratio of more than 13 to 1 (Braun,1977). It was only in 1959 that transistor sales started to match tube sales.
Another factor that slowed down the diffusion of transistor technology was the complexity of solid state theory. Many electronics engineers trained in valve technology were not equipped to deal with the new set of problems raised by the transistor (;. Solutions to problems raised by the transistor were often tackled in terms of valve paradigms. This only served to hinder the development of the transistor. )
In most of the established electronics firms that took up transistor development, the development was conferred to their valve divisions. In these circumstances the transistor was seen as a rival to the valve and a challenge to a lifetime of accumulated experience in valve technology developed by the division's engineers. These are the kinds of conflicts that can arise when one generation's knowledge and experience is threatened by the arrival of new paradigms. Max Plank once observed that new theories gain dominance because their opponents eventually die off. In discussing the development of the transistor, it is important
"...to differentiate between the ability to produce a new device in the laboratory and the ability to produce the thing in scale at competitive prices. Now the former was very much easier. The people who made money really did the difficult task which was producing these things in large quantities at very low prices". (Cited in Braun,1982:54)
Many of the difficulties associated with the transistor in the early 50's were overcome in the course of that decade. Perhaps the biggest single advance during that period, in solid state electronics, was the introduction of silicon as the basic semiconductor material (. The biggest obstacle to using silicon was ensuring the purity of the silicon crystal. Along with the introduction of silicon, there were many other advances made in transistor manufacturing techniques. In time, the problems of quality control, reliability, and performance associated with the early transistors were overcome. The 1950's also exhibited a dramatic increase in the number of companies commercially producing transistors in the U.S. In 1951 there were only 4. During the years 1952, 1953, 1956, and 1960 the number rose to 8, 15, 26 and 33 respectively (Braun,1982). )
Transistor technology was slow to diffuse across the border into Canada. Canadian industry was being left behind. It was not until the end of 1959 that plans to build a commercial transistor plant in Montreal were announced! In the early 1950's Canada did not possess the expertise required to manufacture transistors nor to exploit the potential of the transistor.
Few Canadians in official circles grasped the technological and economic significance of the transistor. The transistor was the leading edge of a transformative technology. As part of the 1956 Royal Commission On Canada's Economic Prospects, a study was published on "The Electronics Industry In Canada". In this report there is a conspicuous absence of any mention of the transistor's importance to Canada. Nor was there any reference of the fact that Canada had no transistor production capacity. But they took great pains to present the statistics concerning vacuum tube production in Canada since 1929. The report made a token reference to the rise of semiconductor electronics:
"The use of transistors is likely to expand rapidly. The first transistorized home and car radio receiving sets are now being introduced on the Canadian market. Further development will probably create new applications, some of which heretofore have been beyond the scope of vacuum tubes. It is doubtful, however, that vacuum tubes will be rendered obsolete" (Royal Comm.,1956:44).
This was the only mention of the meaning of the transistor to the Canadian electronics industry throughout the entire report. It is clear that the authors saw the transistor as a mere extension of the vacuum tube, a solid state valve. As regards the economic importance of the transistor, the Royal Commission missed the boat.
Norman Moody realized that transistors would be the electronic devices of the future. He convinced the Defence Research Board (DRB) of the strategic importance of transistors to Canada's industrial and military future. Through Moody's efforts a group was set up, and funded by DRB, to explore the potential of solid state electronics. From 1950 to 1960, under the direction of Norman Moody, the Electronics Lab at DRTE blossomed into a solid state digital circuit design group of international stature.
Moody's group functioned as a centre of technological diffusion. However, Moody cannot ever recall DRB ever issuing such a directive. Nevertheless, engineers from industry would come to work within the group in order to learn about transistors. Then they would go back to their companies and promote the use of this new technology. One of those engineers from industry recalls:
"...Moody started this group and he exposed about half a dozen people out of industry to this new technology who then went out and became the apostles or disciples of this new technology and around them sprouted centres of competence" (Rywak, 1984)." ()
The DRTE computer, perhaps the first solid state computer designed and built in Canada (, was initially perceived of as a vehicle for exploring the potential of transistor circuits. In the early 50's computers consisted of thousands upon thousands of vacuum tubes. These computers occupied a great deal of space, consumed large amounts of power and required elaborate cooling systems because of the heat generated by the vacuum tubes. Computer technology was very quickly exhausting the capacity of valve technology. Computers offered an excellent opportunity to demonstrate the promise held out by semiconductor electronics. Computer and semiconductor electronics were two new technological forces that interacted in a very symbiotic way. These two technologies have, since the mid 1950', reinforced each other both technically and economically. )
Initially, the computer was a means to an end. However, in time, the computer took on a life of its own. After it was built some of the engineers on the project came to see the computer as an end in itself. These engineers were less interested in transistor circuit design and more interested computer research. This of course posed a problem: DRB was funding transistor circuit research and not computer research. Thus many of the significant improvements, such as faster clock speeds, a parallel arithmetic unit, and more applications software development, that were envisaged for the machine in 1960 were never implemented because of funding restrictions. ()
The DRTE computer was innovative in its circuitry, its architecture, and its core memory. The basic flip-flop was based on a novel trigger circuit developed by Norman Moody. The computer's architecture was inspired by David Florida's view of how a scientific computer should function. The novel way of wiring up the core memory was based Richard Cobbald's work. Many others made important contributions and their roles will be expanded upon at a later point. Though the DRTE computer was innovative, it never translated into a viable commercial venture. Norman Moody was unable to convince Canadian investors to take up production of the DRTE computer. On the other hand the work by Moody's group played an important role in promoting the understanding and use of solid state electronics in Canada. The work carried out by Moody's group served as an essential stepping stone in the design of reliable solid state circuits for Canada's early communications satellites. Hence it was only fitting that the DRTE computer was used to process the signals sent from Canada's first satellite, Alouette.