The Man Behind the Microchip

by Leslie Berlin

  The four-layer-diode was ingenious in theory; it could, indeed, perform all the functions of a transistor (which amplifies current), a resistor (which restricts the flow of current), and a diode (which allows current to flow in one direction but not in the other). The problem arose in production: the four-layer diode was fabulously difficult to build and proved impossible to manufacture in quantity. Several of the young men at Shockley privately suspected that the diode’s greatest appeal for its inventor was that he had patented it alone.62

  Noyce’s group suggested that the company should focus on the less cutting-edge, but more practical, transistor, or at the very least, that they should perfect transistors before moving on to the trickier diodes. Noyce, Moore, Last, and Hoerni had already overcome some of the major hurdles blocking production of the diffused silicon transistor, and researchers at Bell Labs and elsewhere were making similarly quick progress. The young scientists were convinced that they could not only build the diffused silicon transistors they thought they had been hired to build—they could also sell them. Even without sophisticated research to back them up (there was no marketing or business development group at Shockley), they knew in their guts, as Gordon Moore put it, that “there’d be plenty of market” for diffused silicon transistors. Transistors were increasingly common in hearing aids and radios, and the military, of course, would pay a small fortune for the reliable silicon transistor the group was sure they could build. Beckman’s own data-processing system already used primitive germanium transistors; in fact, for more than six months, Noyce had spent one day each month with the data systems operation, consulting with them on the problems they encountered in their use of the devices.63

  Shockley was unmoved. He set up a team of five to work on the four-layer diode, assigned them to a different building, and put himself in charge of the effort. This sent a clear signal about the company’s future direction: research and development of a device that even Arnold Beckman called “a novel, special-purpose diode.” When other divisions of Beckman Instruments declined the opportunity to use four-layer diodes in their applications, Shockley was not disturbed. He simply extended similar offers to IBM, to little effect. He would pursue his four-layer diode for years. It was his white whale.64

  Meantime, Noyce’s group took matters into their own hands. They decided that when Shockley was not at the lab—he traveled frequently and for weeks at a time, especially after becoming a Nobel laureate—they would work on transistors in the manner they chose. When he got back, they would do as he told them.

  Throughout the first half of 1957, the group focused on building a type of diffused silicon transistor called a “mesa” transistor. Mesa transistors, so named because under magnification they resemble the flat-peaked land masses of the Southwest, represented a solid breakthrough in semiconductor technology because they could be produced by masking, diffusing, and etching on only one side of the silicon. Mesa transistors were built in a three-step process pioneered by Bell Labs. First, dopants were diffused beneath the surface of a slice of silicon. Next, a drop of canuba wax was deposited on top of the wafer. Finally, the entire surface was doused with a strong acid, which etched away the top layer, except where the wax drop protected it. This transistor could be attached to other devices via two wires, which were affixed to the now-flat-topped wax droplet.65

  The scientists organized their work along the lines Shockley had originally specified for them. Noyce set the general direction. His collaborative style provided an important alternative to William Shockley’s fierce competitiveness. The young men worried about impressing Shockley, but they were comfortable making mistakes around Noyce. “Bob you could talk to and not expect to blow up,” explains Vic Jones. “You didn’t get any sturm und drang from Bob.”66

  Last focused on polishing the wafers and put his steady hands to work applying the tiny droplets of wax to the surface of the transistor. Hoerni and Moore took charge of diffusion. In relatively short order, the group found themselves producing rudimentary transistors. Noyce, the other members of the group assumed, had gotten Shockley’s official approval—or at least had not been explicitly forbidden—to undertake the work, but none of them inquired too closely. Certainly they knew that when Noyce filed for a patent for one part of the process to build the transistor, Shockley’s name was listed—before Noyce’s, of course—as a co-inventor.67

  Despite these successes, Shockley’s micromanagement and obsession with the four-layer diode, as well as the start-and-stop, covert nature of the transistor work, combined to make the atmosphere in the lab nearly unbearable for the scientists working with Noyce. Their academic training rewarded open inquiry and placed a high value on the professional opinions of senior people in the field. The sense that William Shockley considered their work if not second-class, then certainly of secondary importance, offended their views of themselves as elite researchers.

  Arnold Beckman was frustrated for other reasons. Beckman Instruments was facing a difficult year. Net income was dropping precipitously—the company would earn scarcely $200,000 in 1957 and would lose money in 1958—due to substantial losses on government contracts, an industry-wide recession, and what Beckman obliquely referred to as “certain inadequacies in internal organization and controls.”68

  In May 1957, Beckman convened a meeting of senior research and development managers from every division of his company. Shockley and Horsley attended on behalf of Shockley Semiconductor. R&D costs were spiraling out of control, Beckman told them, and were projected to reach nearly 14 percent of sales by 1958 if they continued unchecked. The meeting seemed designed specifically to rein in Shockley. Every division but his offered presentations on the screening methods, development schedules, and procedures they used to evaluate proposed projects and bring them to completion. Shockley had no such formal methods; all such decisions rested on his personal whim. Beckman proposed a list of eight “dangers to be guarded against in development work.” Shockley had fallen victim to every danger, but two in particular seemed targeted at his efforts: “6. Using boys for men’s jobs” and “8. Failure to regularly assay progress of development programs objectively. It is important to recognize dead horses and bury them.”69

  Beckman decided it was time to pay a visit to Mountain View and speak directly with Shockley’s research staff. He laid out his concerns and suggested several cost-saving measures, all of which were familiar to Shockley from the earlier meeting but nonetheless enraged him. Shockley jumped to his feet and announced that if Beckman did not like the way he was running his business, he would pick up his team and find another backer.70

  The outburst astounded the scientists on Shockley’s staff. Already Vic Jones, frustrated with “spending so much unproductive time on putting out fires and trying to keep Shockley from doing awful things,” had left, with Shockley’s blessing, to teach at Harvard. Several other employees were considering departures, and no one had any illusions about what it was like to work for Shockley. He would never be able to inspire even his senior staff—much less the entire company—to follow him to another venture. “The situation was such that some drastic action was called for,” Noyce wrote his parents a few days after the confrontation, “or we could all pack our bags and leave.” Shortly after this meeting, Noyce, Kleiner, Hoerni, Grinich, Roberts, Moore, and Last went to lunch at the Black Forest. When one of them began running through the often-repeated litany of complaints about Shockley, Grinich, the son of a lumberman and never one to put things delicately, began to shout. “Look, goddammit! We either have to do something about this or stop talking about it!”71

  Beckman needed to know that Shockley’s threats were hollow, the group decided. They had always admired Beckman, a scientists’ scientist who had made it big, and he seemed genuinely interested in their opinions. Gordon Moore was chosen to make the phone call. The others clustered nearby as he dialed and asked, his voice quaking with anxiety, for an appointment to meet with Beckman in private.72

  NOYCE W
ROTE to his parents shortly after the call was made. Beckman “had gone far enough into Shockley’s background to be fully aware of the possibility of this sort of turn,” he explained. Beckman flew in from Southern California and met with the young scientists in a private room at a fine restaurant. The group told Beckman that Shockley was not a leader, but a “disruptive force” in the efforts to build silicon transistors. He was technically terrific, but his outbursts and unpredictability were destroying the group once so eager to work with him. “Beckman assured us that he would support the staff rather than Shockley in a showdown, if indeed some arrangement could be found which seemed to have a reasonable chance of success,” Noyce told his parents. The next week was occupied with what Noyce called “secret evening meetings of the staff, discussing the problem and thrashing out what we thought was the best solution.” At the end of May, they again met secretly with Beckman to suggest that he “try to get Shockley to accept an academic position, removing him from contact with the laboratory on a day-to-day basis, but consulting with us if he so desired.” Sympathetic to their complaints, concerned about the welfare of his company, and committed to finding a solution, Beckman took Shockley to dinner tell him about the discontent at the lab.73

  The news devastated Shockley. He knew that Last was unhappy and, of course, that Jones had left, but overall, he thought things were progressing just fine. He saw himself as his employees’ benevolent-but-firm leader, teaching them the basics of his field, correcting their mistakes, paying them well, and offering them the opportunity to coauthor papers with a Nobel Prize winner. If he had been a bit harsh, it was because he was the intellectual equivalent of a boot camp drill sergeant, harassing his recruits for their own good. After his dinner with Beckman, Shockley went straight to bed.74

  The next morning he called Bob Noyce into his office. Shockley’s notes from this meeting appear in a spiral-bound notebook dated 23 May–10 June 1957, on the front of which he has scrawled, as if to remind himself. “Try to work it out for the benefit of everyone,” “Like you did at meeting, listening.”

  The lab, Noyce told Shockley, was a family, and at this point, it was having “family troubles.” Noyce was “very factual,” noted Shockley, and the young man stressed that he was not disputing any facts with his boss, but rather was bringing to him a “different viewpoint.” Noyce was not particularly ready to walk out the door, but he was worried that almost everyone else in the organization was profoundly discontented. Shockley consistently angered and abused his staff, Noyce told him. “If a man has not spurs,” Shockley wrote to himself poetically, “criticism really hurts.” Noyce reminded Shockley that the team was “doing [its] best with [its] ability,” and assured him that they were not thirsting for blood. “We are not out to get Bill—hope you can see,” was Noyce’s message, as interpreted by Shockley.75

  The staff was “a little disappointed on the technical situation,” Noyce mentioned diplomatically, presumably in reference to the four-layer diode, and he stressed that the decision to go to Beckman had been an act of courage tinged with desperation. “Felt they couldn’t talk to W=S [Shockley’s shorthand reference to himself].” “A few had tried. Would be fired,” Shockley wrote, adding, “It took a lot of doing to go behind my back.”

  Finally Noyce said that his colleagues believed there was “no feeling of stability in the lab.” He spoke of “whims” and “individual reactions” and said something that made Shockley understand “instability due to Bill [himself].” And then, just as it seemed Shockley might begin to see “W=S” as the source of his problems, he added two additional points to the ones made by Noyce. “12. Rex Sittner is sales mgr of Motorola. 13. Always thought it a mistake not to offer Rex [a job].” The problem, in other words, was not Shockley. The problem was the people working for him.

  Shockley, Beckman, and the scientific staff met in various combinations in the final weeks of May 1957. At one point, Shockley brought with him a man whom he did not identify; Noyce was convinced this was a psychiatrist Shockley hired to analyze the staff from afar. Again and again, Noyce was the intermediary. Shockley thought Noyce was on his side. The scientists, with the possible exception of Last, thought he was on theirs. Through Noyce, the dissidents suggested to Shockley that he forego management for an advisory role. Shockley detested this idea. “Non directing position is not effective,” he wrote in outlining his thoughts on potential solutions. “If this is the intention, let’s work out terminations starting now.” He promised that he would provide his team “increased security and work satisfaction,” and sought to forestall any mass resignations by suggesting that no one be allowed to go off the Beckman payroll for six months.76

  He contemplated a new organizational structure, with Noyce not only heading R&D, but also managing the lab. He sketched out a different plan, in which work was administered through a team of four, including Noyce, with Noyce also serving as an “independent authority.” Ultimately, Shockley decided not to grant Noyce independent authority—or even the job as manager. He believed, as he told Arnold Beckman, that Noyce was good technically, but not an “aggressive leader.” He lacked what Shockley called “push,” by which he probably meant that Noyce did not drive his employees as hard as Shockley thought he should.77

  Shockley’s assessment of Noyce’s managerial shortcomings would ultimately prove to hold some truth, but at the time, his decision to bring in an unnamed “mature and experienced manager for non-technical, non-policy decisions” seemed conclusive proof that he was unwilling to give his employees, including Noyce, any power whatsoever. Some have suggested that if Shockley had named Noyce manager, the company might have never faced the problems it did, but that seems unlikely. Whatever Noyce’s pull with Shockley, it almost certainly could not have swayed him from the four-layer diode project—and this project, as much as interpersonal conflict, fed the young scientists’ dissatisfaction.

  While Beckman hunted for his mature manager, he and Shockley arranged a managing committee made up of Noyce, Smoot Horsley (then heading the four-layer diode project), and two others. Shockley could overrule any decisions made by the committee, but “decisions and proposals [would be] a matter of record.” It was a tenuous compromise; enough to keep the company from falling apart, but not enough to push it past the crisis.78

  The interim management committee lasted less than a month, at which point Beckman abruptly reversed his previous support of shared management and declared that Shockley was in charge. The reasons behind this reversal are unclear, but it seems that someone at Bell Labs or at Stanford told Beckman that undercutting Shockley’s authority would irredeemably damage the Nobel laureate’s reputation. Did Beckman really believe that the scientific efforts of a group of young unknowns were more valuable than the contributions of William Shockley?79

  Beckman tapped Maurice Hanafin, who had proved an able administrator in Beckman Instrument’s Spinco centrifuge division, to serve as a buffer between Shockley and his employees. Noyce, given charge of R&D with seven senior staff, reported directly to Hanafin. “He is a very good man … and the chances of things cooling down are much improved by his presence,” Noyce told his parents. Horsley, who led the four-layer diode project with five senior staff, and Knapic, who oversaw engineering and production with three senior staff, also reported to Hanafin.80

  At this point, Noyce declared himself “more confident of the eventual success of the venture than I have been since I arrived”—a confidence that inspired him to buy the biggest, most modern refrigerator he could find. It was a gift of sorts for Betty, who, in the last weeks of her third pregnancy in as many years, admitted “the larger box ought to simplify housekeeping and childcare immensely.”81

  Noyce’s optimism was not shared by the group of disaffected scientists. At a staff meeting Hanafin convened shortly after his arrival, he was pelted with exceedingly direct questions about “future technical decision making and where it would reside.” When, no matter how they put the question, Hanafin re
sponded that Shockley was the ultimate decision maker, the atmosphere in the room turned ugly.82

  Most of the group that had met with Beckman felt that they were right back where they started. Shockley was in charge; production was given short shrift (research and development work received four times the senior staff and nearly twice the funding of engineering and production); and the four-layer diode project was alive and well. Shockley had succeeded in making the leadership of the lab into yet another competition with only one possible winner, and Beckman had cast his lot with Shockley. The young scientists had, as Moore put it, “grossly overestimated our power.” They felt they had no choice but to leave.83

  Julius Blank, Victor Grinich, Jean Hoerni, Eugene Kleiner, Jay Last, Gordon Moore, and Sheldon Roberts (all but two of whom reported to Noyce under the new management structure) were absolutely determined to go. Last, Hoerni, and Roberts were the ringleaders. Last already had another job offer, but he told the others that he would rather continue working with them. By mid-June, the group of seven had resolved to resign en masse.84

  They wanted Noyce to join them, but he was not particularly interested. Unlike the group of seven, Noyce had a managerial title and employees who reported to him. He was Shockley’s favorite, and he thought that Hanafin might solve some of the problems the others perceived as intractable. Moreover, Noyce had joined Shockley with the expectation that he would spend his career there, and he felt he had a moral and professional obligation to make the company successful. “He was the son of a minister and he was not any purer than the rest of us,” one Shockley dissident recalled, “but sometimes he worried, what would his father think, or what would God think, of what he was doing. Was it disloyal or not?” Noyce also had a brand-new baby at home, a little girl named Polly who had been born during the rounds of meetings among Shockley, Beckman, and the unhappy scientists. It was not the right time to contemplate a move.85