To the Digital Age: Research Labs, Start-up Companies, and the Rise of MOS Technology

2. Back from the Frontier: IBM RESEARCH AND THE FORMATION OF THE LSI PROGRAM, 1951 – 1965

57 TWO Back from the Frontier IBM R E S E A R C H A N D T H E F O R M AT I O N O F T H E LSI P R O G R A M , 1 9 5 1 – 1 9 6 5 RCA and Fairchild did not have the MOS transistor to themselves for long; they were soon joined by others, most notably IBM Research. The MOS transistor’s attractiveness to IBM Research had much to do with the history of this large but relatively new research lab. Between 1945 and 1963, IBM’s research capabilities underwent a profound transformation. What had been a modest group, with few Ph.D. scientists and no pretensions to the kind of work done by Bell Labs or the General Electric Research Laboratory, became one of the country’s most important research labs, with hundreds of top-flight scientists working in a new Eero Saarinen–designed building. The beginnings of IBM’s MOS research program were related to the growth of IBM Research in an ironic way. Since its formation IBM Research had embarked on several ambitious research programs that had failed. Because of its inability to make a direct contribution to the welfare of the company, by the early 1960s IBM Research was in an increasingly precarious position within the company. The MOS program was a pulling back from the frontiers of science, to an area that seemed to have much more realistic prospects for applications in products. The technological position of the MOS transistor compared to the bipolar transistor mirrored the organizational position of Research vis-à-vis IBM’s de- 58 T O T H E D I G I TA L A G E velopment groups. The bipolar technology was the technology of the establishment ; the MOS transistor was the technology of the unproven upstart. Because the MOS transistor was so primitive compared to the bipolar transistor and because its characteristics were in so many ways inferior, IBM’s development groups were not interested in it. But it offered Research a way to begin work in the important new technologies of integrated circuits and the silicon planar process without competing directly with the development groups. IBM and Semiconductor Research By 1963, the year it started its MOS research, IBM was a vastly different company than it had been twenty years earlier. Although revenues from electronic computing did not surpass those from electric accounting machines until 1962, the changes that made the growth of electronic computing possible were put in place in the late 1940s and early 1950s. Both symbol of this transformation and driving force behind it was Thomas Watson Jr., who in 1956 took over as chief executive when his father, Thomas Watson Sr., retired at the age of eighty-two.1 IBM’s transition from electromechanical accounting machines to electronic computing machines brought substantial changes in IBM’s development staff. Prior to World War II the design of new products was done by people Thomas Watson called ‘‘inventors,’’ self-educated men with strong mechanical skills working in isolation from the outside world and each other. It was not until 1944 that IBM hired its first Ph.D. scientist, Wallace Eckert, the initial leader of the new Department of Pure Science located at Columbia University.2 An internal report written in 1950 describing IBM ‘‘Research and Development ’’ gives a picture of IBM’s technical staff as it stood on the threshold of this new era. It opened its discussion of IBM’s technical personnel with the statement ‘‘engineering is the heart of IBM,’’ but the prominence it gave to the fields of dynamics and kinematics was a reminder that IBM was a company based on mechanical technology. It emphasized the experience of the IBM staff, suggesting that experience could substitute for education. Of 619 people in the ‘‘category of technically trained members of the organization,’’ 11 held doctorates; 48, master’s degrees; and 280, bachelor’s degrees. The report noted that while many of the remaining 280 lacked formal engineering education, they had 59 Back from the Frontier ‘‘proved their ability during a long period of experience with the IBM organization as inventors and design engineers.’’ The work of development was done in ‘‘Invention Departments.’’ In an era when the Manhattan Project had shown what large numbers of highly trained physicists could achieve, IBM’s Physics Laboratory had a staff...