This is a fascinating story of the birth, growing pains, and maturation of one of the prominent technologies of the twentieth century, the metal oxide semiconductor (MOS) transistor, which is the basic component of most integrated circuits built today.

A detailed study of the bumpy evolution of MOS technologies from 1945 to 1975 is delivered by the author, analyzing the role that companies like RCA, IBM, Fairchild, and Intel have played in making the MOS structure the building block of the digital age.

Eight chapters make up the book. The first chapter narrates the period from 1945 to 1963, during which the MOS transistor slowly emerged as a low cost alternative to the bipolar transistor.

In the second and third chapters, Bassett describes the MOS work at IBM from 1951 to 1967, presenting the relationship between the research and components arms of IBM, the large scale integration program established at IBM, and the nonlinear recognition that the MOS technology had at IBM during those years.

In the fourth and fifth chapters, the author analyzes the essential role that Fairchild Semiconductor and wizards like Frank Wanlass and Lee Boysel had during the 1960s in the evolution of MOS technology. Fairchild is the birthplace of the modern semiconductor industry. The silicon planar process--essential for fabricating stable components and integrated circuits--was perfected at Fairchild. The work of Fairchild’s Grove, Deal, and Snow has set the theoretical foundations for modeling the MOS structure. However, Fairchild failed to establish a leadership position in the MOS integrated circuit market.

The MOS work begun at Fairchild was continued at Intel, which was founded in 1968 by Gordon Moore and Robert Noyce. The sixth chapter of the book focuses on the evolution of Intel from 1968 to 1975. Intel improved the MOS technology, perfecting the silicon gate process that allowed high yield production of stable MOS components. The firm launched a range of successful MOS products, from different types of semiconductor memories--random access memory (RAM), read-only memory (ROM), and erasable programmable ROM (EPROM)--to microprocessors.

The author notes the contrast between the information transfer policy of Fairchild and Intel. Fairchild had licensed its most important technology, the silicon planar process, and had extensively published in technical journals the innovation made within its research and development labs. In contrast, Intel did not patent the silicon gate process, and patented mostly inventions, which could be identified by examining the chips. In the 1970s, Intel made the judgment that it would be better off by keeping critical technologies as trade secrets rather than by collecting royalties from licenses.

Chapter 7 presents the expansion of the MOS memory technology at IBM from 1967 to 1975, accompanied by the slow adoption within IBM of MOS for digital logic. While the MOS memory was an effective alternative to magnetic cores for IBM mainframes, the firm did not find a compelling advantage of MOS logic over bipolar logic at that time.

The last chapter of the book discusses the birth of a revolutionary MOS product: the microprocessor. In 1971, Intel introduced the 4004 product, a four-bit microprocessor with 2300 transistors operating at 108 Khz. As Gordon Moore noted, with the introduction of the microprocessor, microprocessor makers gained a significant influence on the design of a computer system, at the expense of the value added by the computer makers.

The story of MOS technology reveals the multitude of factors that influenced the evolution of this technology, such as scientific and technical leadership, the relation between the research and manufacturing arms of a semiconductor company, and the vision of the decision makers about the direction of the semiconductor industry. However, having the right vision was not enough for becoming a big player. Bassett characterizes RCA’s vision as being “right, but on the wrong time scale.” While RCA envisioned the advantages of the MOS transistor, it was frequently unable to produce successful MOS products.

I highly recommend this book to anyone interested in technology and research.