Robert Noyce

Robert Noyce: The Mayor of Silicon Valley

In his final year at Grinnell College, Robert Noyce stole a twenty-five-pound pig from the mayor's farm for a fraternity luau. He was nearly expelled; his physics professor, Grant Gale, argued his case down to a single semester's suspension. It was the same professor who had obtained two of the first transistors out of Bell Labs and shown them to his class. Both interventions mattered. The suspension kept Noyce in physics; the transistors decided what he would do with it.

A Minister's Son from Iowa

Noyce was born on 12 December 1927 in Burlington, Iowa, the third of four sons of the Reverend Ralph Brewster Noyce, a Congregational minister, and Harriet May Norton, herself a clergyman's daughter and, in Tom Wolfe's description, an intelligent woman with a commanding will. He grew up in Grinnell.

He graduated Phi Beta Kappa in 1949 with a degree in physics and mathematics, and with the Brown Derby Prize — awarded, in the college's phrasing, to the senior man who earned the best grades with the least amount of work. That combination of facility and lightness followed him. At MIT, where he took a PhD in physics in 1953, fellow graduate students called him Rapid Robert.

He went first to the Philco Corporation in Philadelphia as a research engineer, then in 1956 to Mountain View, California, recruited by William Shockley to Shockley Semiconductor Laboratory. Shockley had a Nobel Prize and an ungovernable temperament. Within a year Noyce and seven colleagues, Gordon Moore among them, had resigned — the traitorous eight — and founded Fairchild Semiconductor with backing from Sherman Fairchild.

Metal Over Oxide

The problem that made Noyce's name was interconnection. Jack Kilby at Texas Instruments had demonstrated in 1958 that a complete circuit could be built from a single piece of semiconductor material, but his version required tiny wires soldered by hand between components — flying wires, unsuited to volume manufacture.

Noyce's answer came out of work already in hand at Fairchild. Jean Hoerni had developed the planar process in early 1959, building on Mohamed Atalla's silicon surface passivation and thermal oxidation methods, which left a chip with a flat, oxide-covered top surface. Challenged by the patent attorney John Ralls to think of further applications, Noyce realised that if the surface was flat and insulated, the connections could simply be printed onto it: diodes, transistors, resistors and capacitors diffused into the silicon, linked by aluminium lines deposited on top. Fairchild adopted Kurt Lehovec's reverse-biased p-n junction method to isolate the devices electrically within the wafer. Noyce filed his patent for a semiconductor device-and-lead structure in July 1959; it issued on 25 April 1961, and the first working monolithic circuits came off the line in May 1960.

Fairchild and Texas Instruments then spent years in court. The rulings went largely in Noyce's favour, but the companies settled with a cross-licensing agreement, and history has, correctly, recorded both men as co-inventors of the integrated circuit — Kilby for proving it could be done, Noyce for making it manufacturable. The planar approach he exploited remains the fundamental method by which chips are made today.

Intel and a New Kind of Company

In July 1968 Noyce and Moore left Fairchild to found Intel, and Noyce served as its first chief executive until 1975. The technology mattered, but so did the culture he built, which turned out to be the more portable export. Noyce abolished the trappings of corporate hierarchy — no reserved parking spaces, no executive jets, no private offices for the leadership — and ran the company on the premise that engineers respond to responsibility rather than supervision. Nearly every Silicon Valley firm since has copied some version of it, often without knowing where it came from.

His personal standing in the region became semi-official; he was known as the Mayor of Silicon Valley, and in his later years he took on the industry's collective problems, leading SEMATECH, the American semiconductor manufacturing consortium formed to counter Japanese competition. He held fifteen patents. He flew planes, hang-glided, dived, and read Hemingway. He married Elizabeth Bottomley in 1953, had four children and divorced in 1974; he married Ann Schmeltz Bowers later that year, who had been Intel's first director of personnel and became Apple's first vice-president of human resources.

The honours were substantial: the Franklin Institute's Stuart Ballantine Medal in 1966, the IEEE Medal of Honor in 1978 for his contributions to the silicon integrated circuit, the National Medal of Science and the Faraday Medal in 1979, election to the American Academy of Arts and Sciences in 1980, and the National Medal of Technology from Ronald Reagan in 1987. In 1989 he entered the US Business Hall of Fame, with a keynote from President George H. W. Bush. He died of a heart attack on 3 June 1990, aged sixty-two.

Why Robert Is Called a Genius

Noyce's gift was for seeing that a problem everyone was attacking head-on could be dissolved from the side. The industry in 1959 was trying to shrink and connect discrete components; Noyce saw that Hoerni's planar process had incidentally produced the thing that made connection unnecessary as a separate step — a flat insulating surface you could print wiring onto. That is a different cognitive move from invention in the usual sense. It is recognising that a solution to one problem is already lying around, unrecognised, next to another.

He was also unusually fast. Rapid Robert was not a nickname bestowed casually by MIT physicists, and the Brown Derby Prize for best grades with least work points at the same quality. Institutions have endorsed the judgement about as heavily as American institutions can: the National Medal of Science, the National Medal of Technology, the IEEE Medal of Honor, the Faraday Medal.

The honest counter-case is that the monolithic integrated circuit was not his alone. Kilby got there first in demonstrating the principle and won the Nobel Prize for it in 2000, ten years after Noyce's death made him ineligible. Hoerni invented the planar process; Lehovec supplied the isolation method; Atalla supplied the surface chemistry. Noyce's contribution was the synthesis and the recognition of what mattered commercially, which is precisely why the court fight was fought and why it ended in a draw. His leadership at Fairchild and Intel may in the end be the larger achievement, and leadership is not the same thing as brilliance.

Legacy

Every microprocessor descends from a design principle Noyce sketched in 1959, and a great deal of how technology companies are organised descends from how he ran Intel. He founded the Noyce School of Applied Computing at Cal Poly, San Luis Obispo, and pressed throughout his life for better science education at every level. In 1990, the year of his death, he shared a Lifetime Achievement Medal with Jack Kilby and John Bardeen — the transistor's inventor, the integrated circuit's first builder, and the man who made it something the world could mass-produce.

Achievements

• National Medal of Technology and Innovation — 1987
• Affiliated with Fairchild Semiconductor
• Educated at Grinnell College and Massachusetts Institute of Technology
• Worked as physicist, inventor and businessperson