The Man Behind the Microchip

by Leslie Berlin

  After meeting with one of the company’s founders, Jim Hobart, Noyce agreed join the Coherent board. He also purchased roughly 22,500 shares of stock for about $1.40 per share. “I’m not sure why Noyce wanted to come on board,” admits Hobart. “Maybe he just wanted to try something different [from semiconductors].”

  As an experienced entrepreneur, Noyce served as a stabilizing force at Coherent. “Noyce’s biggest contribution to the company,” Hobart recalls, “was [to say], ‘Let’s not be emotional. Let’s figure this out. Let’s discuss it, debate it, think about it.’” Noyce also pushed Hobart to use stock options as an incentive tool. Hobart had already planned to distribute options to all technical employees, but Noyce promoted their use even more aggressively. He thought that Hobart should use what Noyce called an “evergreen” program—one that he would implement at Intel—in which employees are given a new options grant every year in an effort to keep the staff loyal to the company.

  When Coherent Radiation went public in May, 1970, Noyce made roughly $250,000. He held onto his stock and served on the board for another 13 years.

  Noyce also served on the board of Four-Phase Systems, a Fairchild spinout founded by the leader of Fairchild’s MOS circuit design group, Lee Boysel. Four-Phase aimed to build an MOS-based computer with solid-state memory and a logic chip built from silicon. Intel, too, of course, was building solid-state memory and logic chips. But neither Noyce nor Boysel saw any potential conflict of interest in Noyce’s decision to serve as a director of Four-Phase. Boysel’s company was not going to sell any components at all—only finished computers—and so would not compete with Intel in any market.

  Noyce invested about $50,000 in Four-Phase. His was one of several fairly small investments by individuals; the major backer, Corning Glass (which had also backed Fairchild spinout Signetics) invested $500,000. Noyce most likely did not have high hopes that Four-Phase could accomplish the goals Boysel laid out for the company, but Noyce enjoyed involving himself with operations risky enough to sound almost audacious. He was also fascinated by the Four-Phase computer effort, and it may have been his early exposure to Boysel’s ideas about general-purpose logic chips that made him so receptive to Hoff’s thoughts on the subject.52

  THE BOARD THAT MATTERED MOST to Noyce netted him no income and met in Grinnell, Iowa. Four or six times each year, Noyce traveled to Grinnell College, where he was completing his fourth, and final, year as board chair in 1970. He loved this job. One of his nephews attended the college, and Noyce always stopped by his dormitory to talk to the students and “get a sense of the place.” Noyce led the Grinnell trustees meetings with the same air of apparent nonchalance he brought to all his meetings, seemingly more interested in understanding his colleagues’ ideas than in expressing his own. And yet the questions he asked had a way of bringing people around to his point of view, one of the college presidents recalled. “Somehow they usually reached the conclusions Bob wanted them to reach. You just never had a sense he was leading them there.”53

  In May 1970, Noyce received a call from Grinnell College president Glenn Leggett. At Kent State University in Ohio, members of the National Guard, called in by the governor in response to massive campus protests against the war in Vietnam and the invasion of Cambodia, had killed four unarmed students and wounded nine others. Leggett told Noyce that the Grinnell campus was in an uproar unlike any in its history—and its recent history was the sort that inspired an FBI report stating “this College has a widespread reputation in the Midwest as being of the ultra-liberal type.” In 1968, Grinnell students had protested a Marine recruiter’s visit by transforming the campus into a makeshift military graveyard, complete with hundreds of crosses dotting the lawns. In 1969, a representative from Playboy, on campus in connection with a new policy allowing opposite-sex visits in the dormitories, had been greeted by ten naked students carrying signs reading, “Playboy is a Money Changer in the Temple of the Body”—a protest Noyce attributed to “outside agitators.” Grinnell had also seen near-continual anti-war protests for the past year.54

  But nothing had prepared the administration for the reaction to Kent State. Students from universities all over Iowa began to congregate at Grinnell, many of them “big city kids,” as Leggett put it. Leggett received threats of “violent action” planned to erupt on campus. The student body sent a resolution to the faculty, which was meeting in special session, declaring the Kent State atrocity an “act of official repression” both “immoral” and “irrational.” The faculty voted to suspend classes for two days, a move that left hundreds of agitated students with time to stage impromptu teach-ins and protests. Adding to the sense of barely contained chaos was the physical condition of the campus: the buildings and grounds employees had been on strike since mid-April, and vandals had overturned furniture in department lounges, defaced the library walls, and dumped garbage on the lawns. As Leggett put it, “The temperature was rising every minute of every day after Kent State.”55

  Leggett wanted to know how Noyce, as chair of the board of trustees, felt about the school shutting down for the safety of the students. Noyce told Leggett, “We [trustees] don’t know anything about what the students will do. That’s your problem. Do what you can about it, and we will support you.”56

  On May 14, Leggett asked Noyce to come to Grinnell immediately. He had decided to cancel not only classes but also the commencement exercises and reunions scheduled for two days hence. Many parents and alumni—and some students, as well—were furious. The people in town, a fair portion of whose income came from serving families and alumni attending these functions, were even angrier.

  For three days, straight through the would-be commencement weekend, Noyce sat in the living room of Grinnell House, the campus home of a dozen Grinnell College presidents and now used primarily for entertaining distinguished guests. Cigarette always in hand, Noyce talked to anyone who came in to register a complaint or get more information. Or, to be more precise, Noyce characteristically listened more than he talked.

  Noyce felt that the events at Kent State were tragically predictable, the near-inevitable outcome of mixing young angry students with young armed troops and high tensions. Privately, and in discussion with other trustees, Noyce worried that Grinnell was pandering to its students, overly sensitive to every individual psyche and interest. The faculty had dropped nearly all graduation requirements as part of what was called “a pioneering adventure into free and open curricular territory.” Student athletes had begun refusing to compete because they “didn’t believe in competition.” Noyce thought this was all absurd. Students had to learn that the world did not always accommodate every whim. He was concerned that Grinnell was failing in its educational responsibility to teach students how to succeed. A few years after Kent State and the events at Grinnell, Gordon Moore said of the team at Intel, “We are really the revolutionaries in the world today—not the kids with the long hair and beards who were wrecking the schools a few years ago.” Noyce undoubtedly would have agreed.57

  Noyce, of course, shared none of these thoughts with the parents, alumni, students, and community members who came to speak to him at Grinnell House. Among this crowd, Noyce was more sponge than speaker—his gaze always focused on the person talking, his head cocked in deliberation and understanding. President Leggett has always felt that Noyce, with his calm demeanor and longstanding affiliations with both the town and the college, helped to defuse a potentially explosive situation.

  BACK AT INTEL, Noyce focused a good deal of his energy on trying to build Intel’s credibility. Most potential customers at this point were large computer companies comfortable working with established suppliers like Texas Instruments, Motorola, or Fairchild. The plethora of little startups nipping at these big players’ heels did not impress executives accustomed to more mature suppliers. Noyce offered Intel’s best tool for standing out from the other upstarts. “It helped that Noyce could come in having run Fairchild, with that understanding and experience,�
�� explains Intel’s marketing vice president. “We were the only [young company] that maybe could count, and Bob [did that].” Customers gave Intel a chance in part because they believed Noyce when he said that this little operation could perform. In some sense, he could play the same role for Intel that Sherman Fairchild had played for Noyce and his seven Fairchild Semiconductor co-founders when Mr. Fairchild vouched for them with IBM. “Nobody else could have done that [made Intel a credible alternative to big suppliers],” the vice president averred. “Andy couldn’t have done that. I couldn’t do it and didn’t do it. Bob did it.”58

  In early February 1971, Noyce spent two weeks in Japan encouraging Japanese firms to design the 1103 into their products. The island nation would come to account for about 15 percent of Intel’s 1972 sales. During his visit to Japan, Noyce spent a day at Busicom, the calculator manufacturer for whom Hoff had developed the microprocessor. Noyce must have been feeling good about the project. Frederico Faggin had proven an inspired hire. Only nine months after he joined Intel, Faggin produced working samples of the entire calculator set—a remarkable accomplishment. Busicom president Kojima, too, must have been pleased with the progress on the chip set. But he had a new issue on his mind. The calculator market was proving more competitive than expected. Even though Busicom was only beginning to receive shipments from Intel, Kojima wanted Intel to reduce their prices.59

  When Noyce returned to California, he asked Hoff what Intel’s top priorities should be if a chance arose to re-negotiate the Busicom contract. Hoff was insistent: “If you can’t get any other concession, just get the right to sell to other people.” Frederico Faggin recalls making the same request to Noyce. Busicom cared about the microprocessor only as a calculator, but Noyce, Moore, and the technical men behind the chip saw that it could just as easily be programmed to do other things. A programmable, general-purpose logic device, Noyce thought, could be the “standard 2-by-4 or 6-penny nail” in the electronics industry. As things stood now, engineers at computer companies designed every plank and nail equivalent used in their systems.60

  Shortly after Busicom chose Hoff’s architecture for their calculator, Noyce had quietly begun his own private version of market research on “a general purpose way of programming logic.” (The term microprocessor had not yet come into use.) When he visited customers who were requesting custom circuits for certain simple logic functions, Noyce made sure at some point to ask, in an offhanded way, why the customer did not just buy a computer and program it to do the task. The answer was always the same: I could do that, but it’s too expensive. This research further reinforced Noyce’s own hunch that the microprocessor could be used in dozens of potential applications—not only in computers, but also in areas largely untouched by microelectronics, such as cars and home appliances—if the price was right.61

  Hoff believes that Noyce was almost alone among senior management in his excitement about the microprocessor in 1969 and 1970. Moore insists that although “Hoff thinks it was Noyce who kept that from being killed, [there was] no way that project was going to be killed! The microprocessor was an example of exactly the kind of product that we were looking for for the next generation.” Andy Grove wanted the microprocessor to go away. Bob Graham felt the company already had enough to do just selling memory chips. How could they sell microprocessors? And why would they want to, since Graham estimated Intel could expect to sell, at best, 2,000 units per year? It would just be a big distraction with little potential for income.62

  Perhaps this opposition explains why Noyce did not act immediately on the Busicom request to renegotiate prices. He certainly had other things on his mind. In June, the company was scheduled to move into a new building on a 26-acre abandoned pear orchard in Santa Clara, about ten miles south of the original Mountain View headquarters. The building had been under construction for more than a year and represented the first step in a $15–$20 million office and fab complex that Intel expected would eventually occupy nearly 400,000 square feet. The shareholders meeting was scheduled for April, and before that, Noyce planned to attend an electronics research meeting sponsored by the IEEE in New York and then leave for a packed ten-day trip to Europe. In Paris, he planned to speak on “microcircuits” to about 100 people, roughly one-third of whom were potential customers, one-third competitors, and one-third potential investors. In Brussels, he would meet with the head of Intel’s first European sales office. Noyce and Bob Graham, who joined him for part of the trip, also met with potential customers and distributors to promote the 1103 and Intel’s other products.63

  Hoff, who accompanied Noyce on this trip, treasures one memory from it. “When I came down for breakfast one morning, I met [the Intel Europe sales manager]. He looked like he had been run over by a train. It seems that Bob had wanted to stay up talking and drinking, and the salesman made the mistake of trying to keep up with him. Just as I was wondering what condition Bob was going to be in, he came downstairs as fresh and chipper as anything. And the salesmen are the ones who are supposed to be the ones who can do all the drinking and such!”64

  As soon as Noyce returned from Europe, Moore left town for a week. These sorts of oscillating schedules were not at all unusual in the early years at Intel. As long as the other was home to “mind the store,” both Noyce and Moore were comfortable going away.

  Throughout the spring and summer of 1971, Noyce was also preparing for Intel’s initial public offering of stock, slated for the fall. The 1103 was beginning to look like a breakaway success. It had met and then passed Noyce’s penny-a-bit target and every major computer mainframe manufacturer other than IBM (which used proprietary devices) had committed to the product or had the 1103 in prototype. This news helped position Intel as a promising upstart bursting with potential—investors’ favorite type of company. Noyce and Rock also wanted to take Intel public because they felt that it was time to give the employees, the earliest of whom had seen the value of their stock options nearly quintuple, a chance to sell some of the stock on the public market. Plus, Intel could use the cash. The move to the new building generated losses of more than $400,000 for July and August, and while September was looking more promising, Noyce warned employees that “one robin doesn’t make a spring, and the competitive situation remains severe.”65

  In preparation for the IPO, Noyce consulted with attorneys and bankers, reviewed drafts of the prospectus, met with auditors, signed the certificates necessary for the offering, wrote explanatory letters to employees and current investors, invited employees to buy stock in the offering as “friends of the company,” and met dozens of times with Rock and with executives from C. E. Unterberg, Towbin, the investment bank that was underwriting the offering. By August, most of Noyce’s days included at least one IPO-related appointment.66

  SEC rules for public offerings required Intel to cancel the stock-purchase plan set up at the company’s establishment. Noyce dreamed of replacing this plan with options packages that would be distributed to every employee, “including janitors.” He worried, though, if people with limited educations could understand what a stock option was and how volatile the markets could be. He and Moore finally decided that once a plan could be developed that met SEC guidelines for publicly held companies, Intel should re-institute a stock-purchase plan, rather than options, for nonprofessional employees. Under this plan, which was implemented in 1972, every employee would be allowed to take up to 10 percent of base pay in Intel stock, which could be bought at 15 percent below market rates. The stock purchase plan met Noyce and Moore’s goals of giving employees a stake in the company without requiring the sophisticated financial knowledge associated with stock options.67

  Adding to Noyce’s workload were the tensions between Graham and Grove, which reached a breaking point in the summer of 1971. In May, Grove had marched into Moore’s office and told him that he thought he might need to leave Intel because, as he later put it, “it was far too painful for me to continue to do what I needed to do at work and fight Bob
Graham.” Faced with this news, Moore had begun to fiddle with a paper clip on his desk—a sure sign of stress to those who know him. Bob Graham was a close personal friend of Moore’s. They often fished together.68

  Grove left Moore’s office worried that his boss was not going to do anything, but Moore almost immediately walked over to Noyce’s office and shut the door. Shortly thereafter, Noyce began interviewing candidates for the marketing job. Although Noyce and Grove had very different approaches to management at Intel, Noyce had grown to like Grove—and more importantly, to appreciate his value to the company. Noyce saw what had happened at Fairchild when his detail-man Sporck left. He could not let that happen again.

  By June, Noyce had found a marketing candidate who impressed him and who he thought could work with Grove: Ed Gelbach, a luxuriously mustachioed, straight-talking, technically trained, number crunching marketing expert from Texas Instruments. Just to be safe, Noyce and Moore insisted that Grove interview Gelbach before they offered him a job. Grove came away so impressed by Gelbach that he not only approved his hiring but also began growing his own mustache. Gelbach immediately demonstrated his negotiation skills by securing a very rich stock-options package for himself.69

  With Gelbach on deck and Moore away on a fishing trip—a planned absence that “just made it a lot easier for me,” Moore says—Noyce walked into Bob Graham’s office in early July and sat down. “Shit,” Noyce said, cutting straight to the point. “We have an incompatible situation here, and you’re going to have to go.” When Graham called home to tell his wife that he had lost his job, all she said was, “Thank God.” The battles with Grove had been destroying them all.70