The Man Behind the Microchip

by Leslie Berlin

  Whatever Bob thought of Betty’s childrearing tactics, he respected the division of labor in the household. He intervened with the children only occasionally and then in only the most passive of ways. When their three-year old, who had been sent from the table for unacceptable behavior, kept returning to provoke Betty’s anger, Bob stood up, and without chastising the child or asking Betty to stop yelling, he blocked the door with a chair so the little one could not come in. Then he sat down and resumed eating in silence. Another time, a child had problems keeping her left hand in her lap. Every time her hand approached the table, her mother would slap her wrist. Finally Bob stood up, pulled off his belt, and strapped his daughter’s left hand to her chair. For this the daughter was grateful. She could not move her hand, so she could not be slapped.

  Betty Noyce was in a difficult situation. Wives were seen as an extension of their husbands for most of the 1960s. Indeed, not much had changed since 1953, when William Shockley had dismissed a potential recruit with a jotted notation in his notebook that he “did not want a man whose wife was annoyed about it all.” In 1966, the leading electronics newsletter, which Bob almost certainly read, noted that “American corporations are adopting the practice of taking a careful look at [a man’s wife]” and went on to describe “the ideal executive wife” as “a kind of glorified Girl Scout” who not only performed “the standard wifely duties” but also “provides the emotional and psychological balm to relieve the pressures on her tension-ridden husband.”52

  Betty Noyce did not accede willingly to this role. Even when she tried her best to appear demure and self-effacing, she did not succeed. She told a reporter, for example, that “I fear I neglect my housekeeping as a result of all the books I read.” In other words, she was not a glorified Girl Scout; she was an intellectual. She also refused to think of herself as an object for display and regularly ran her errands with a head full of wet hair and attired in a sartorial style one friend affectionately described as “not bag-lady, but close.”53

  Betty Noyce resented the situation in which she found herself: a Tufts-educated woman with four small children, a husband who was never home, and a life in a place she detested, far from the place she loved. She disliked cooking—“meals should not take longer to cook than to eat,” she often said—and she asked Richard Hodgson, in front of Bob, “Why is it, if we’re so damn rich, that I’m still washing dishes?” Moreover, she considered herself Bob’s intellectual equal and hated appearing as his appendage at social functions, which she avoided as much as she could.54

  She found an intellectual outlet in volunteer work. She helped to establish a library in Los Altos and chaired a fundraiser for the San Francisco educational television station. This work allowed her to exert considerable leverage in support of causes that mattered to her—but on a schedule that enabled her to be home for her children. Betty Noyce also wrote, always using a pseudonym, but she was never published. And she created beautiful needlepoint art and quilts, several of which elaborately chronicled the family’s activities.

  She suspected that her husband was unfaithful. Bob Noyce functioned in a testosterone-drenched world in which all of his equals were men and every woman a subordinate. “The business ran on alcohol and playing around,” recalls Jay Last. Noyce’s trips to Japan included female entertainment. Salesmen openly joked about buxom blonds in noisy conference halls. At the end of an announcement that a pre-production line was changing buildings, a Fairchild newsletter could note that “all the R&D men will miss the smiling faces of the beautiful young girls in the 4200 line.” Noyce was powerful, attractive, unhappy at home, a risk taker who believed in grabbing as much from life as he could, and in regular contact with the “girls” who not only worked on assembly lines and behind reception desks but who also joined the scientists and engineers for drinks after work. He probably did have casual affairs.55

  Noyce knew that he was becoming “a stranger to [his] own family.” He said he envied “people out in the machine shop who could go home at night and sleep with no concern.” He was appalled to discover that the longer he stayed at Fairchild, the less he had any interest in anything except business. “What are you as a person when that happens?” he asked. Then he answered his own question: “You’re nothing.”56

  MATTERS ONLY WORSENED at the office. While Noyce split his attention between Semiconductor and the instrumentation division, new Semiconductor general manager Charlie Sporck reorganized the firm along the product-manager model made famous by Procter and Gamble. He designated several product managers, all of them engineers, to coordinate production of their specific devices. The reorganization did not fully decentralize the Semiconductor operation, since marketing and central production control (which determined the volumes of devices to be manufactured) maintained centralized operations. Nonetheless, it was a significant change from the organization structure under Noyce.57

  The changes came too late. By the end of 1966, Fairchild began to miss its promised deliveries, at times meeting only about one-third of its customer commitments. At the same time, the company failed to market new products developed in R&D because the transfer from development to manufacturing was so inefficient that the devices were never manufactured in volume. Indeed, Semiconductor’s difficulties in bringing its own inventions to market were so renowned that they generated an oft-repeated industry one-liner: “The first parts coming out of Fairchild R&D were probably made in Sunnyvale.” Sunnyvale was home to Signetics.58

  By the end of 1966, Semiconductor’s festering troubles had become apparent even to outsiders. In the fourth quarter, Camera and Instrument’s profits dropped below those of the third quarter. Although the parent company publicly blamed the semiconductor division, Wall Street was not fooled. One vice president of Kidder, Peabody spoke for many in the investment community when he said, “If I could just buy the semiconductor division, I might do it, but I can’t see paying a premium for Fairchild’s management and all the uninteresting stuff you have to take with semiconductors.”59

  At the urging of Camera and Instrument, Semiconductor initiated “FAIRCHILD 71,” a five-year planning tool and the launch pad for “a concentrated program of process cost reduction and mechanization [that] established definitive guidelines for expansion.” A command to reduce costs infuriated the Semiconductor employees. For years, their division had been more profitable than the company as a whole, with other parts of the company losing money and serving as net drains from the Semiconductor bottom line. And even though Noyce now served as a vice president, the Semiconductor division still did not have a formal representative on the parent company’s board. Although another vice president of the corporation served as a director, Noyce did not.60

  In March 1967, over drinks at Chez Yvonne, a disgusted Charlie Sporck told Noyce that he was leaving to take the job of CEO at the moribund National Semiconductor. Leaving with him were several key integrated circuits men. For months Sporck had been complaining about how hard it was to attract new engineers to Fairchild when competing firms could match a Fairchild salary and offer about 1,000 stock options. Sporck further resented that Camera and Instrument was “throwing away in various directions” money made by the sweat of his brow at Semiconductor. Moreover, he recalls watching the rising fortunes of Semiconductor’s spin-offs and wondering, “Why don’t I do that sort of thing?” Why not indeed? Sporck had long thought that he would run a bed-and-breakfast in his native upstate New York after he left Fairchild, but the offer from National Semiconductor was too attractive. Noyce did not even bother trying to convince Sporck to stay at Fairchild. He understood his friend’s frustrations.61

  Sporck’s departure was personally very painful for Noyce, who leaned heavily on his strong second-in-command. “I suppose I essentially cried when he left,” Noyce said. “You know, working with people that you’re fond of, then having them break apart, was I would almost say devastating.” Noyce asked Gordon Moore, the head of R&D, to serve as general manager. Moore had kept the
laboratory functioning smoothly throughout its dramatic growth. He had required every group within R&D to update him on their activities with a brief weekly progress report. The technicians had begun their own series of reports that covered issues ranging from the importance of flushing the pipes to the relative merits of paper towels versus filter paper for drying wafers.62

  Moore declined the general manager’s job. “The rest of the company [other than R&D] was a mess, and I didn’t know what to do about it,” he later explained. Noyce then named Tom Bay, his former marketing lieutenant, as general manager of Semiconductor, a move that did not ease the company’s troubles.63

  A disquieting uncertainty settled over Fairchild Semiconductor in the months after Sporck’s defection. No one knew who was staying and who was secretly hatching plans to go when at the end of March 1967, Noyce and patent counsel Roger Borovoy, along with their wives and Noyce’s assistant Paul Hwoschinsky, traveled to Vienna to negotiate a licensing agreement. When the Noyces and Borovoys realized that the next day was April Fools’, they decided to play a joke on Hwoschinsky. Noyce called Tom Bay in California and asked him to send a telegram. The two couples had carefully worked out the wording:

  Paul. Have just learned that [two senior Semiconductor managers] are leaving for National and have reason to believe that Bob Noyce plans to join them. Delicately probe his intentions and report back. Urgent. Tom.

  Bay agreed to send the telegram immediately, and the group, eagerly anticipating the next morning’s foolery, went to sleep.

  When they assembled for breakfast, Hwoschinsky looked miserable. It was clear he had been up all night. Recalls Brenda Borovoy, “Our scheme had worked perfectly, and Betty, my husband and I glanced at each other conspiratorially, waiting for the fun to begin. No sooner had Paul cleared his throat, however, than Bob blurted out, ‘It’s a joke.’ We felt cheated. How could he? But Bob Noyce, in the face of someone so troubled, could only be Bob—a nice, honest, and fundamentally good human being.”64

  The merriment in Europe was a welcome break for Noyce. “I just felt that things were falling apart,” Noyce would later say of the months following Sporck’s departure. Noyce knew that there was no way he could keep high-caliber employees at Semiconductor much longer: the offers from outside were simply “too enticing” and the situation at Semiconductor too dismal.65

  Noyce was a highly creative man who by 1967 found himself functioning in an almost purely reactive mode: fighting lawsuits, assimilating acquisitions he did not want in the first place, and wrestling with Camera and Instrument management over stock options. His notes from meetings in Syosset deal almost exclusively with administrative details that Noyce would have found mind-numbing: accounts receivable, advertising budgets, organization charts, breakdown of overhead costs, and ongoing personnel issues. At one point, he wrote to himself, “Try to get East Coast out!”—apparently, managers in Syosset were trying to dictate precisely how many people should be assigned to each group within Semiconductor.66

  Noyce had no respect for senior Camera and Instrument management other than Hodgson. Noyce often told the story of waking up in Syosset to a snowstorm so severe that he could not find a cab to take him to his meeting at Camera and Instrument headquarters. He walked to the building, muttering and cursing to himself the whole way, only to discover that no one else was coming. He had managed to make it from the other side of the continent, but the men who ran Camera and Instrument could not be troubled to attempt the trip of several miles from their homes. Noyce had no use for these people.

  By the late 1960s, Noyce was not only alienated from his bosses, he was also out of touch with the innovative, technical side of Semiconductor in which he had taken such pride a decade before. Although he continued to receive copies of all the reports from R&D, this was simply acquiring scientific knowledge, not contributing to it. And while he brought his lab notebook with him to the general manager’s office, he did not write an entry in it for nearly three years—and then he wrote only at very scattered intervals. He missed doing science, and even after Semiconductor had grown to tens of thousands of employees, even after it had been years since he sat at a lab bench, Noyce would tell his family that he was “going to the lab” when he left for work each day. “After growing up with a baby, you don’t like to abandon it,” he ruefully told a reporter about the Fairchild lab. “You’d rather keep in touch, to stay aware of what’s going on, but there is less and less time to do this and more and more time that must be spent worrying about other kinds of problems—people, organization, production, marketing, and all the rest of what makes an industry rather than what makes a science.” It was no doubt around this time that Noyce, who was nearing 40, began to consider leaving Semiconductor.67

  While Noyce plotted privately, he urged Camera and Instrument to take steps to soften the blow of Sporck’s departure for the rest of the employees—some of whom he knew were wavering—at Semiconductor. Camera and Instrument finally took decisive action. Noyce was given a seat on the board, and shortly thereafter, the directors pushed through a vast democratization of the stock-option plan, authorizing 300,000 additional shares for options. The stock-option committee, which had met only sporadically in the past, began meeting monthly to distribute options. Almost 100 employees, many of them middle managers at Semiconductor, received new option grants. About half of these employees had never before held options. For the other half, the new grant more than doubled their total holdings.68

  In any case, it was all too late to stanch what had become a hemorrhaging of employees from Semiconductor. Six months after Sporck’s departure, some 35 people had left to join him at National. Employees began fleeing Semiconductor from almost every possible exit door. “Suddenly,” wrote Business Week, “every semiconductor company in the Bay area was able to hire Fairchild professional people.”69

  Semiconductor’s fortunes plummeted quickly. In late March 1967, the company announced that several production difficulties had been overcome, but the announcement went largely unnoticed in a tide of bad news. An unanticipated drop in consumer demand throughout the industry meant many customers no longer needed the devices they had ordered. In October, Semiconductor, for only the third time in its ten year history, reported monthly losses.70

  Due in large measure to Semiconductor’s slipping performance, Camera and Instrument’s earnings for the third quarter of 1967 were a paltry $137,000—down a staggering 95.5 percent from the preceding year’s third-quarter profit of $3 million, and a worse performance than an already leery Wall Street had expected. With the company barely breaking even, the stock price slid to 52 from 92 at the beginning of the year.71

  In the wake of the appalling third-quarter earnings report, the board of Camera and Instrument ousted the company’s high-living CEO, John Carter, and asked Richard Hodgson, the man who first lured the Shockley defectors to Fairchild, to add CEO to his title of president. In 1981, a reporter claimed that Noyce had forced the change of leadership and that in 1967, shortly after Carter’s departure, Noyce had proudly told the journalist, “When you set out to kill the king, you’d better kill him dead.” Noyce’s purported intervention seems plausible. With its Semiconductor golden goose so clearly in trouble, and Noyce now a member of the Camera and Instrument board, it is likely that the board would follow his suggestions about how to remedy the crisis. And Noyce undoubtedly would have suggested that Hodgson replace Carter.72

  Whatever its genesis, Hodgson’s elevation to CEO did not generate immediate effects, of course. By year’s end, Camera and Instrument, admitting that the semiconductor division accounted for well over half the company’s sales, reported a $7.7 million loss. Some $4 million of this were write-offs, but the remaining $3.5 million trail of red ink compared to a profit the previous year of more than $12 million. The company described the results as “a deliberate attempt to group all losses and take the beating at one time.”73

  Just weeks into the new year, it became apparent to Noyce that his
fellow Camera and Instrument board members were losing patience with Hodgson’s leadership. If Hodgson were to be ousted, no one Noyce respected would remain in the executive suite. Noyce needed to consider his options. He decided to pay a visit to Gordon Moore.

  “I’m thinking of leaving Fairchild,” Noyce told Moore in the offhanded tone he so often used to announce big news. Perhaps, Noyce said, he would try to start a company to build computer memory chips out of integrated circuits. What did Moore think? Noyce would love to have him as his co-founder of a new company. If he left, would Moore join him?74

  Noyce must have known the idea of building semiconductor-based memory devices would appeal to Moore. Just a few months before, in fact, Moore had told Noyce that he thought semiconductor memories were “one of the first ideas I’ve seen in a long time that you could probably start a company on.” The computer market was growing dramatically. In the past two years, the number of minicomputers had increased fivefold. Noyce and Moore knew this market well; some estimates contend that by 1968, Fairchild held 80 percent of the computer market for integrated circuits. Meanwhile, the semiconductor industry’s progress in squeezing ever more devices onto a circuit meant that soon it would be possible to build an integrated circuit complex enough to be practical for computer memories. Fairchild itself had brought to market two-bit flip-flop chips that could perform memory-like functions. IBM, by far the world’s dominant computer manufacturer, announced that it had begun researching semiconductor memories with the intent to use them as the primary memory in future IBM computers.75

  It seemed to Moore that Noyce’s proposal that they start a company together in January, 1968, was another one of his boss’s why-not mind flashes—a big idea that Noyce, as was so often his wont, needed Moore to help determine whether or not to take further. Moore told Noyce that while he still believed in the future of computer memories, he liked his job in the Fairchild lab. He was insulated from the Syosset politics, and he ran the finest commercial research operation in the industry. Moore said that he was not ready to leave Fairchild. Noyce made no effort to change Moore’s mind.