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

This is a remarkably detailed narrative of the development of MOS (metal-oxide-semiconductor) technology. A singularly important technical and commercial breakthrough, the MOS transistor replaced (Ross Bassett goes so far as to say overthrew) the bipolar transistor invented by William Shockley in 1948. Bassett's account of the development of MOS has important implications for understanding the relationships among science, commerce, new product development, and research and development (R&D).

As Bassett explains, the significance of MOS is that it is scalable. That is, it makes possible putting extremely complex circuitry in an extremely small space by etching electronic pathways on tiny and inexpensively manufactured wafers of silicon. The bipolar transistor was a single device. Each individual transistor, representing the most basic element of an electronic circuit, had to be individually inserted and soldered into a circuit board that connected the transistors to create the complex electronic circuits that perform the basic functions of computing: bytes of memory and rudimentary logic. The MOS technology allows a large number of transistors to be combined in a circuit on a single device, which might represent tens or hundreds of thousands of transistors. A single wafer, embedded in a plastic case of a few inches represents the equivalent of a great many circuit boards the size of a small room.

Scalability has profound implications for both manufacturing costs and processor speed. Costs are reduced because, rather than requiring large numbers of individual components, a complex circuit can simply be drawn (now using computer-aided design) and the drawing then photo-reduced to microscopic dimensions and, in effect, photocopied onto the silicon wafer. Assembly involves just a few rather than thousands of electronic components. Speed is increased because, although the electrons moving through the circuits move at close to the speed of light, they are moving over vastly smaller distances in these circuits. In fact, "Moore's Law" that the cost of computing is halved every eighteen months is a function of the continuing ability to increase the density of circuits.

Bassett offers a story of the industrial R&D process that is at odds with a great deal of conventional wisdom. The development of MOS technology is not a Thomas Edison-like drama of the lonely inventor struggling for a breakthrough. Nor is it the story of a community working on a well-focused Manhattan Project. Rather this is a story of dozens of people working in different teams on different projects held together by what was, at best, a loose awareness of each other's work. Bassett weaves several threads together as he describes the interplay between research and product development in several well-known and not so well known enterprises. It is precisely this lack of an inventor-hero or focused project team that makes Bassett's explanation of value to students of industrial research.

Bassett makes a contribution to the general critique of the dominant paradigm of industrial research and development that has emerged over the past twenty years. Associated with the 1945 report to President Harry Truman by Vannevar Bush, Science—The Endless Frontier, this paradigm has informed the government's science policy for the past half century (see Bruce Smith, U.S. Science Policy since World War II, 1990). Bush articulated what has become known as the "linear model," which "conceived of industrial innovation as proceeding from basic to applied research, and then to development and commercialization" (Wesley M. Cohen, Richard R. Nelson, and John P.Walsh, "Links and Impacts: The Influence of Public Research on Industrial R&D," Management Science 48 [Jan. 2002]:1).

The most important element of the "consensus" was the primacy of basic research, which pursues its own internal scientific goals. Innovation (changes in technology applied in industry) would follow from it...