The Man Behind the Microchip Page 9
When Westinghouse called with an offer to join the Pittsburgh-based transistor group at an immediate 25 percent raise, plus a guaranteed annual 10 percent raise for each of the two years thereafter, Noyce was tempted. They could buy a house in a nice neighborhood near Carnegie Tech, and the work sounded more interesting than what he was doing at Philco. Philco countered with an offer to meet the raises and make the management position permanent. The company also dangled the possibility of a transfer to Landsdale, where Betty, who reminded Bob it was time to “start thinking of a permanent site for ourselves,” thought they could buy “a little house and a big yard.” Bob wrestled with the decision for a week before deciding one morning at 2 AM to stay at Philco.66
In August 1955, Noyce noted, “My current assets are: household furnishings of a four-room apartment, a car valued at $700 with a lien of $232, about $300 cash on hand, stocks and savings of about $650, and a $20,000 life insurance policy two years in force with a cash value of $260. My current liabilities, other than monthly household running expenses of about $400, are debts to the extent of $500. My wife and children [a baby was due in October] live with me and are wholly dependent on me for their support.” How dramatically his life had changed.67
In December, the draft board sent Noyce a belated twenty-eighth birthday present in the form of an “indefinite postponement of induction.” He would not need to worry about the draft. The new baby, a girl called Penny, had arrived safely. The little apartment looked downright cheerful with her mobile suspended from the ceiling and the philodendron Betty had decorated as a makeshift Christmas tree propped against the rabbit-ears FM antenna. But the military work continued, and the move to Landsdale never materialized. When the calendar turned to 1956, Bob knew he wanted nothing more than “to walk away [from Philco] and start over again somewhere else.”68
On January 19, 1956, Noyce answered his ringing telephone. The man at the other end of the line greeted him with two words: “Shockley here.” William Shockley, one of the inventors of the transistor, assumed Noyce knew who he was, and he was right. “It was like picking up the phone and talking to God,” Noyce recalled later. “He was absolutely the most important person in semiconductor electronics.” And he wanted Noyce to work for him—in California.69
3
Apprenticeship
William Shockley held more than 50 patents for electronic devices and by one estimate was personally responsible for nearly “half the worthwhile ideas in solid-state electronics” in the field’s first dozen years. Lanky and lean, with a wide forehead that began overtaking his hairline when he was still in his twenties, he was a direct descendent of Mayflower Puritans. Educated at California Institute of Technology and MIT—where he received his PhD in Physics in 1936—Shockley was a technical genius whose work ethic recalled that of his ancestors. During the Second World War, while working in the navy’s Anti-Submarine Warfare Operations Group, he developed systems that quadrupled the number of successful American attacks on German subs. He was a pack rat and a vigorous note taker who filled notebook after notebook with cryptic, penciled records of the day’s events. He was also a showman and amateur magician who had been known to pause in the middle of delivering a scientific paper to conjure a bouquet of flowers from the lectern. His colleagues sometimes wondered if he slept. When his wife Jean fell ill with uterine cancer in 1953, Shockley, then overseeing a team at Bell Labs, launched an intensive oneman research effort into the disease, analyzing her lab slides at home, annotating complex medical articles, and writing to doctors around the world.1
Shockley had left Bell Labs in 1954, roughly the same time Noyce began work at Philco. It was not an amicable parting. The inspiring story of the “three inventors of the transistor” that Bell Labs had circulated around the world in 1948 whitewashed a nasty internecine conflict within the solid-state group. The point-contact transistor, the world’s first transistor, had been invented by Brattain and Bardeen without assistance or blessing from Shockley, who was their boss. Indeed, Shockley, nearly apoplectic with jealousy, had marked the transistor’s discovery by calling the inventors into his office to tell them, his voice rising with every word, that since the point-contact transistor built on some of his own early ideas, he could write a patent “on the whole damn thing.” When Brattain responded, “Oh hell, Shockley! There’s enough glory in this for everyone,” Shockley took his argument to the Bell Labs patent attorneys. In the course of investigating his claim, the attorneys discovered that the ideas he said were the foundation of the transistor might not be so original, after all. The safe route was not to include Shockley’s name on the patent application.2
Even before receiving this personally devastating news, Shockley had begun an effort to develop a transistor on his own. Within days of hearing of Bardeen and Brattain’s invention, he cloistered himself in a hotel room in Chicago (where he was attending a conference), thinking and writing furiously. A patentable device eluded him, but he kept his musings to himself, fearful that if he shared them with Brattain and Bardeen, he might have to list them as co-inventors of whatever device he planned to bring into being by sheer force of intellect and will. In mid-January, Shockley’s secret stewing paid off, when he realized that a Bell Labs colleague’s research findings meant that the ideas he had mapped out in his Chicago hotel room could be fused into a workable device: the sandwich-like junction transistor. It may be spite’s greatest contribution to American science.
The conflict with Brattain and Bardeen—which only widened when Bell Labs decreed that any picture of “the inventors of the transistor” must include Shockley—represented merely one in a string of battles and hard feelings between Shockley and his colleagues. By 1951, Bell Labs senior management could no longer ignore complaints about Shockley’s prickly personality, especially after Brattain and Bardeen told the head of the Physical Research Department that they no longer wanted to report to Shockley. Within days, the department was reorganized, with most of Shockley’s old transistor group shifted to someone else.
His career clearly beginning to stall, Shockley took a sabbatical. After a year spent teaching at his alma mater Cal Tech and another in Washington, D.C., where he headed the Pentagon’s Weapons Systems Evaluations Group (a cadre of civilian scientists and engineers charged with advising the military on the weapons planning process), Shockley was desperate for a change. He divorced Jean, whose life he had fought so valiantly to save only the year before. He moved to California.3
And he decided to start a transistor company. He foresaw a day when transistors would power everything from airplanes to televisions, and he felt confident that Bell Labs and its manufacturing arm, Western Electric, which viewed the transistor as little more than a potential replacement for vacuum tubes in the telephone grid, would offer little competition. He spent most of the summer of 1955 in conversations with Raytheon, Texas Instruments, and Laurence Rockefeller, trying to secure half-a-million dollars to launch a transistor operation. After promising starts, all of these efforts stalled.4
In August 1955, however, Shockley found a willing source of financial support in millionaire Arnold Beckman, a fellow graduate of Cal Tech with whom Shockley had become reacquainted during a black-tie dinner at the Los Angeles Chamber of Commerce the previous February. A professor of analytical chemistry whose military bearing and bald pate recalled Dwight Eisenhower, Beckman had successfully made the leap to industry that Shockley longed to make as well.
Beckman Instrument’s first product had been a pH meter that could electronically measure the acidity of oranges. It soon found a number of industrial uses in water treatment plants, plating and anodizing operations, and paper factories. The company’s other major products included a helical potentiometer, used in the nose of proximity-fuse missiles, and an ultracentrifuge employed in the first successful separation of the polio virus. Headquartered in a low-slung, attractive glass-and-steel building in Fullerton, California, Beckman Instruments had offices in a dozen sites in the Unit
ed States, Canada, and Germany.
Beckman Instruments was flush in 1955. With more than 2,000 employees, sales of more than $21 million, and profits in excess of $1.3 million, the company was in its second consecutive record-breaking year and had acquired two small firms within a few months. Since the firm was already organized into multiple autonomous divisions, each with its own general manager and complete operating organization, adding another independent business unit was relatively easy.5
As befitted a former scientist, Arnold Beckman was deeply committed to research and development, the heading under which he classed Shockley’s work. Beckman Instruments regularly reinvested at least 8 percent of sales in R&D, a move its founder called “insurance against obsolescence.” In the mid-1950s, when improved data-processing methods and advances in semiconductor technology seemed likely to shape the future of Beckman’s military and industrial markets, Arnold Beckman decided he wanted a stake in the basic research in both fields. Accordingly, in early 1955, Beckman Instruments launched an expensive R&D effort to investigate optimal methods of digitally interpreting data. Shockley’s proposed transistor efforts would give Beckman Instruments a leading position in semiconductor research.6
The negotiations between Shockley and Beckman were quick and friendly. Beckman Instruments would fund the operation, but Shockley would have complete managerial control. As a director and president of the new division, he would receive an annual salary of $30,000 and options to purchase 4,000 shares of Beckman Instruments stock. Beckman and Shockley estimated first-year costs at $300,000, including a $25,000 payment to Bell Labs to license patent rights for the transistor. On September 3, 1955, they signed an agreement to “engage promptly and vigorously in activities related to semiconductors.” They predicted that within a year the new Shockley Semiconductor division would generate monthly sales of $30,000.7
Beckman would like to have situated the semiconductor operation in Fullerton or at least in the greater Los Angeles area, but Shockley wanted to start the company further north—in the San Francisco Bay Area. He had grown up in Palo Alto, near the Stanford University campus, and his mother, whom he adored with a devotion that unnerved some of his acquaintances, still lived there. He was an avid mountain climber and outdoorsman, and the topographical diversity of the Bay Area intrigued him. Palo Alto still had the feel of the semirural college town Shockley recalled from his boyhood. Wood-shingled houses dominated the oldest neighborhoods near the small downtown built around University Avenue, a street that was called Palm Drive at the entrance to the Stanford campus. Slightly farther south, where the houses took on a distinctly Mediterranean cast, walnut and olive trees lined the wide boulevards named for famous poets and local families. A half mile beyond, south of Oregon Avenue, lay scattered dairy farms. Palo Alto had one telephone exchange and shared a municipal court judge with the neighboring town of Mountain View.8
The once quaint rural town was rapidly becoming a small city in 1955. Between 1950 and 1960, its population more than doubled, from 25,000 to 52,000, and its acreage tripled. Developers and builders, who had opened housing tracts among the dairy farms at a rate of three per year since the end of the war, enticed returning veterans and young families to the area with two- and three-bedroom bungalows that sold for roughly $9,000—monthly payments of only $60, advertisements touted—and included a guaranteed “six or more bearing-fruit trees” to each 6,000 square-foot lot. Do-it-yourself shops and hardware stores sprung up around town, including at the Stanford Shopping Center, an upscale mall that opened in 1956, the same year as the town’s first golf course. To accommodate the area’s families, Palo Alto built 15 public schools between 1948 and 1956.9
Several successful technical companies had gotten their start in Palo Alto, and Shockley undoubtedly would have welcomed the opportunity to associate himself with this auspicious entrepreneurial history. A hotbed of radio and microwave research in the first decades of the century, the town was home to Federal Telegraph’s manufacturing plant and 626-foot transmitting tower until 1932, when the Depression-plagued company consolidated its operations in New Jersey. Electronics giant Hewlett-Packard was founded in a Palo Alto garage in 1939. In 1948, two scientists affiliated with Stanford’s Physics Department started Varian Associates to commercialize innovations that they believed would prove useful in radar technology. By 1950, traffic from the industrial districts a few miles south of town was so heavy that locals were lobbying to expand the narrow Bayshore Freeway (nicknamed “Bloody Bayshore”) to four lanes.
A concerted effort by Stanford University accelerated Palo Alto’s trend towards electronics industrialization in the 1950s. Frederick Terman, professor of engineering (later dean and provost) at Stanford, wanted the university to serve as the center of what he called a “community of technical scholars”—a web of academic and industrial researchers that would work together to advance “sophisticated technologies.” Terman envisioned a symbiotic relationship in which technically oriented companies would support advanced research at Stanford while at the same time benefiting from a supply of well-educated graduates and professors interested in consulting work. He encouraged students and faculty—most notably William Hewlett, David Packard, and the Varian brothers—to start their own companies, rather than travel back East to work for established ones. He pioneered an innovative “Honors Cooperative Program” that allowed employees of local electronics firms to work part-time towards advanced degrees. Shortly after endowment-poor Stanford opened a research park in 1953 in an effort to raise funds, Terman began wooing “smokeless industries” whose work would be relevant to the university’s academic programs. He even alerted highly placed contacts at the Defense Department to the existence of local firms that might be logical contracting choices.10
When Terman learned of Shockley’s hopes to locate in Palo Alto—the two men belonged to several of the same professional societies—he wrote to him with assurances that the university “would heartily welcome this activity in the Stanford area, and I believe that its location here would be mutually advantageous.” Terman had been contemplating how to increase Stanford’s participation in the semiconductor field for several years. He told Shockley that Stanford had recently hired its first faculty member who specialized in semiconductors, and that other prominent Stanford professors were already incorporating transistors into their coursework. He suggested that Shockley might find the Honors Cooperative Program a useful recruiting tool and spoke glowingly of the 200 potential employees enrolled in graduate electrical engineering courses at Stanford. He concluded, “It is an exciting business to observe the University and the technical community grow cooperatively to the benefit of both. We hope that you will see your way clear to participate in it.”11
After sending his letter, Terman contacted the local Chamber of Commerce on Shockley’s behalf (though without disclosing Shockley’s identity) and secured a list of appropriate industrial sites and knowledgeable real estate agents that he forwarded to Shockley. He offered occasional use of “the facilities of the University” and planned to name Shockley to the unpaid but prestigious post of university lecturer. Even for Terman, the courtship was unusually intense. He was excited by the knowledge that “Shockley-Beckman [are] playing for big stakes.”12
Terman’s overtures and the proximity to Stanford may well have served to make Arnold Beckman more comfortable with a Palo Alto site for the semiconductor operation. It is unlikely, however, that such attractions held much sway with Shockley, whose location plans were driven almost entirely by personal concerns. Having Stanford and Frederick Terman nearby was a nice bonus—and Shockley certainly mentioned it in recruiting workers to his company—but he wanted to come to Northern California regardless.
Shockley had devised a unique plan to staff his company based on research he had conducted during his spare time while serving in the navy. Detailed analysis of scientific papers written by researchers at some of the nation’s most prestigious labs—including the nuclear weapo
ns facility at Los Alamos—had convinced Shockley that every person had a “mental temperature” that could be objectively determined through a series of tests and evaluations of written work. The brighter the mind, the higher the temperature. Theorizing that “a mere doubling of mental temperature may jump a man’s scientific creativity a hundredfold, like a heat-triggered chemical reaction,” Shockley was determined to staff his company with the hottest minds in the world.13
After his attempts to recruit Bell Lab scientists failed—his reputation preceded him—Shockley spent the end of 1955 phoning colleagues at other leading research labs and technical firms around the country, asking their assistance in finding the top young men with expertise for the various positions he sought to fill. He wanted to reproduce the division of labor at Bell Labs. To do so, he needed to hire several experimental and theoretical physicists, as well as chemists, metallurgists, and electrical and mechanical engineers.
First to join Shockley were Smoot Horsley, a clean-cut, mild-mannered, 40-year-old physicist from Motorola with extensive semiconductor experience; and Dean Knapic, an engineer who had worked at Western Electric and whom Shockley hired to head up production. Beckman Instruments proudly included a photo of the three men in its 1956 annual report, along with a comment from Arnold Beckman that could have been penned by Shockley himself: “In research there is no substitute for superiority.”14
In October 1955, Shockley flew to Pittsburgh to hunt for hot minds at a semiconductor symposium sponsored by the Electrochemical Society. The symposium had been a last-minute addition to the regular meeting agenda, tacked on by the conference organizers. After receiving so many papers on the burgeoning new semiconductor field, they decided to sponsor a series of ten-minute presentations on the topic over the course of a single day.15