The Man Behind the Microchip

by Leslie Berlin

  Graham was probably the first person that Noyce personally fired, a remarkable situation given that Noyce had been a senior manager for more than a decade. Firing Graham must have been excruciating for Noyce, who thought he was smart and good with customers. At Fairchild, those feelings, coupled with Noyce’s own dislike of personal confrontation, would have been reasons enough to keep Graham on board. But Noyce had learned his lesson. He needed the Grove “whip” at Intel. “What that must have taken for Noyce to do,” Grove mused aloud years later. “It must have been like cutting out his liver.”71

  ED GELBACH’S ARRIVAL AT INTEL had an immediate effect on the company’s approach to the processor chip in the Busicom calculator set. Texas Instruments, Gelbach’s previous employer, was trying to build a general-purpose logic chip, and Gelbach, like Noyce and Hoff, saw the promise in the devices. He absolutely thought Intel needed to find a way to get the rights to the microprocessor they were building for Busicom.

  Gelbach’s enthusiasm refocused Noyce’s thoughts on the device. Within a week of Gelbach’s arrival, Noyce began jotting down ideas about microprocessors in the thin, black-covered record book he carried with him nearly all the time. The pages are peppered with ideas that ranged far beyond calculators: “point of sale—inventory control,” “concept for data retrieval,” “electronic checkout,” “Supermarket—3 mkts [markets]—1. hardware, 2. processing of data for mgmt [management], 3. market research data from volatile base.” He imagined customers who might be interested in microprocessors: “NCR, Litton, RCA, Pitney Bowes, AC Nielson, Time Inc.” He could picture the devices used in “traffic control—stop lights” or in “gas pumps.”72

  In August, Noyce requested that Intel’s general counsel investigate “exclusivity” issues related to the microprocessor. This is probably when the process of negotiating the rights to the microprocessor back to Intel began in earnest. Noyce met with Mr. Kojima in Japan on September 21, most likely the date that the negotiations, which were quite casual, were finalized. “The Japanese don’t use lawyers for these kinds of negotiations; the Americans tend to,” Noyce explained later. “So [Busicom] had a guy … who was just writing down the terms of what we agreed on, sitting there talking in the office; very nice simple agreements…. It was just simply writing down, ‘You agree to do this, we agree to do that.’” When the discussions concluded, Intel had secured the right to sell the chip for non-calculator applications. Around this time, concerns about a potential conflict of interest led Noyce to resign from the board of Four-Phase, the company building a computer around a general-purpose logic chip.73

  Intel did not, however, immediately exercise its right to market the microprocessor. Noyce told Hoff, “We have a tiger by the tail, but we’re just not ready to announce the product. We’re not ready to make a decision.” At this point, Noyce and Moore were not concerned about whether a market existed for a microprocessor device, but they were unsure whether this particular microprocessor was, in Noyce’s words, “adequate to serve that market. Should we wait to get something better?” A few board members feared that the microprocessor might push the company into the systems business. Intel, like most other semiconductor companies, sold individual circuits that customers would plug into systems of their own design. But this four-chip microprocessor family Hoff had designed was already a system in itself. Some board members worried that a move to market the microprocessor would put Intel in the position of competing with its own customers, a fear that echoed those faced by Fairchild Semiconductor when it introduced the integrated circuit. Noyce had begun Intel with the expectation that the day would come when semiconductor companies would also build the computer systems that used their circuits—that is, in fact, one reason why Noyce hired Hoff, with his expertise in computer systems—but this was far from a universally accepted opinion.74

  When Noyce finished explaining to Hoff that he was not certain Intel would bring the microprocessor to market any time soon, Hoff was almost beside himself. “Every time you delay [a public announcement of the microprocessor], you are making a decision!” he said. “You’re making a decision not to announce. Someone is going to beat us to it. We’re going to lose this opportunity.”75

  In the end, the few board members’ objections had no hope of succeeding against the force of the support Noyce, Moore, and Gelbach—as well as Arthur Rock—expressed for this new device. (Grove says, “Microprocessors meant nothing to me. I was living and dying on two points of yields in memory.”) It was probably October when Intel decided to market the microprocessor.76

  Yet the microprocessor was far from the most important item on Noyce’s agenda in the fall of 1971. On October 13, the company went public, offering nearly 300,000 shares at $23.50 apiece. This was not a large slice of the company. Early employees and directors of Intel held a combined 2.2 million shares (after stock splits), for which they had paid an average of $4.04. The offering was a smashing success: oversubscribed and immediately raising nearly $7 million for Intel.77

  Noyce had flown to New York for the IPO, and he spent most of this first day as head of the publicly traded Intel walking the city, stopping every few hours at a pay telephone to call Arthur Rock for the latest word on how Intel was trading.

  Walking with Noyce among the skyscrapers of Manhattan was Barbara Maness, a mask designer in Intel’s MOS group. Maness was 28 and attractive, with long, straw-colored hair and clear blue eyes. She was employee number 43 at Intel. She and Noyce had been engaged in an affair since the summer of 1969, when Noyce, his wife and children in Maine for three months, invited Maness to dinner.78

  As a mask designer, Maness was among the most highly compensated women at Intel. Masks are the detailed blueprints for any type of integrated circuit. In the late 1960s, they regularly consisted of thousands of squares, Xs, dashed lines, dotted lines, fat lines, thin lines—each color-coded and hand-drawn as small as possible, and each symbolizing a transistor or a diode or particular type of connection among the components in a circuit. Engineers had traditionally drawn their own masks, but in 1966, Fairchild had decided to teach a select group of women how to translate the engineers’ plans into mask designs.

  In 1966, Maness worked for Fairchild as a die attacher on the swing shift. (She did not know Noyce at Fairchild.) She was recently divorced from a man who provided no support for her three young children. She lived near the freeway in what she described as a “bad part of San Jose,” in a mobile home that was the only option available to her because most apartment landlords did not want to rent to single mothers. She existed in a state of permanent exhaustion, getting off work at midnight, picking up her children at a sitter’s in the wee hours of the morning, and sleeping only a few hours before the children awoke and needed her. When one of the Fairchild engineers suggested that she try to win a spot in the inaugural mask-design class, she jumped at the chance for a bigger paycheck and more reasonable hours. She took the required tests and was one of six women admitted to the class. She left Fairchild for Intel only because the engineer with whom she worked most closely had accepted a job at Intel and asked her to work for him there.

  Maness’s presence in Noyce’s life was an open secret at Intel, where many people considered Betty’s announcement that she was going to live in Maine for the summers “an open invitation for Bob to find another woman.” Extramarital affairs, while always considered a bit shady, were common enough to have generated their own lexicon. The women were called “girlfriends.” The couples were “dating.” Noyce and Maness regularly spent time with other semiconductor executives and their girlfriends.79

  Barbara Maness was every bit as bright as Betty Noyce but, unlike Betty, she was also young, adventurous, undemanding, and unquestionably devoted. Noyce enjoyed doing with her things he never could have done with his wife. He could talk with her at length about Intel and the technical aspects of its products. She happily joined him to sail in Hawaii, to ski in Aspen, and to watch the Independence Day fireworks from the copilot’s
seat of his plane as they flew down the California coast. Noyce paid for Maness to take flying lessons—“If something happens to me, I want you to know how to land the plane”—which she adored. He kept a collapsible motorcycle in the back of his plane, and he and Maness would often ride to remote locations after they landed. Then they would backpack for hours, sometimes above the snow line. Later they would unwind with cups of “Purple Jesus”—grape Tang and vodka—that Noyce chilled in the snow pack.

  Maness was entirely self-sufficient. While singlehandedly raising three children, she also managed to save enough to buy a parcel of land in the Santa Cruz Mountains, where she intended to build a home. Noyce found her independence attractive. “You’ll do well,” he told her admiringly, “and I’ll just stand by and watch.”

  One of Noyce’s favorite parts of their illicit romance was the element he called “intrigue.” Always a risk taker, he reveled in bringing his family and Maness into close proximity with each other. He paid for Maness’s children to attend sleep-away camp so that she could live in his house with him for the summers while his family was away. He enjoyed having Maness with him when he talked to Betty on the phone. When his daughter’s dog had puppies, Noyce asked Maness to come to the house to buy one from Betty, posing as a random Intel employee who had read about their sale on a company bulletin board. Maness was “scared to death,” but she did it. It would not be surprising if Maness had designed the mask that the Noyce family followed when they came to Intel to cut rubylith.

  Noyce played the same sorts of games at work. When Intel bought a block of tickets for a play, he obtained a pair and gave one to Maness, whom he dropped off a few blocks from the theater. She sat next to an empty seat until the lights went down, at which point he ducked in and joined her. Immediately before the lights came up again, he left. On a business trip, he made a show of buying a stunning pearl necklace in front of a colleague who would be almost certain to see it around Maness’s neck at the company Christmas party. He had Maness accompany him to a dinner with an important visitor to Intel, and the next morning went out of his way to lead this man past Maness as she worked on her design. This notion of “pulling one over on people” delighted him to no end. For her part, Maness generally kept quiet. “I was very protective of him,” she explains. “I didn’t go bragging about.”80

  Walking through the streets of New York City on the day his company went public, Noyce made sure to take Maness to the Statue of Liberty, the very place he had only dreamed of seeing he was a teenager in Grinnell. Now he was here, an important man with a beauty at his side—and his last phone call to Rock informed him that he was also a millionaire ten times over.81

  ALMOST IMMEDIATELY UPON HIS return to California, Noyce began lessons for certification as a jet pilot. He kept detailed notes in his daily calendar: “Watch—at 5000 flatten to get 500’min—level air plane and check horizon—at 65% power for cruise 5-7 speed altitude and main[tain] for 500 and 1000—hi speed….” He also began shopping for small jets and attending meetings of the Aircraft Owners and Pilots Association. At home, the Noyces’ family life registered few visible changes after the IPO. Noyce worried that it was “tough to teach values” when a family had great sums of money. “It’s much easier to be raised with a father who is poor—[because] you have to hustle—than it is with a father who is wealthy,” he told a friend. He and Betty kept knowledge of the family’s financial situation hidden from the children as much as possible.82

  Almost exactly a month after the IPO, on November 15, 1971, Intel announced its microprocessor with a bold advertisement proclaiming the arrival of “a new era of integrated electronics—a micro-programmable computer on a chip!” More than 5,000 people wrote to the address at the bottom of the advertisement, requesting more information—the most dramatic response to a product announcement Intel had ever experienced. Noyce was attending the Fall Joint Computer Conference in Las Vegas at the time of the announcement, and he almost certainly helped to staff the Intel suite, which was soon overrun with customers wanting to know more about “this computer on a chip thing.”83

  To use the microprocessor, customers had to change their thinking. Before the microprocessor, designers at Intel’s customer firms built their systems by configuring individual integrated circuits, each with a different dedicated function, on a board. Changing the system required changing the physical arrangement of the integrated circuits, or hardware. Intel’s new microprocessor systems required something very different—changes made not by moving physical objects, but by reprogramming the instructions stored in program memory. The microprocessor, in other words, brought software to the semiconductor industry. In doing so, it placed new demands on customers, most of whom were experienced hardware designers unfamiliar with using computer programs to solve their systems problems.84

  Regis McKenna, a marketing and public relations expert hired by Intel late in 1971, summarized the problems facing Intel’s microprocessors thus: “You couldn’t promote the things [if] people didn’t know what they were.” The challenges faced on this front were daunting enough that Gelbach set up an independent microprocessor marketing operation, separate from the rest of the marketing division. The company developed voluminous written technical documentation for the 4004—the owner’s manual alone ran well over 100 pages at a time when other Intel chips were boxed with only a ten-page data sheet. Intel shipped more manuals for the 4004 than actual processors. “There were just lots of people who wanted to read about microprocessors, independent of whether they were buying any,” explained Bill Davidow, whom Gelbach had hired to run microprocessor marketing.85

  Intel developed and marketed hardware tools to facilitate programming the microprocessor. The EPROM, a specialized read-only memory chip, could be programmed electrically and erased with ultraviolet light, which made it possible to reprogram the microprocessor very easily. Development aids, marketed under the name “Intellec,” were hardware simulation environments, simple single-board versions of a computer with all the proper circuitry to emulate the functioning of a central processor. A designer could use this development aid in prototype before committing to a specific microprocessor program. The EPROM and Intellec systems functioned as software but looked like the hardware familiar to designers.

  Intel also sponsored a series of technical seminars on the microprocessor as part of the ongoing attempt to “lower the confrontation level with the engineers,” as Davidow puts it. Senior Intel engineers led seminars at a bruising rate in 1971 and 1972. A typical schedule found them one day in Connecticut, the next in New Jersey, the next in Philadelphia, the next in Rochester, and then, after a weekend of rest, in Minneapolis, Chicago, Ann Arbor, Houston, and Dallas. These meetings were as helpful to the presenters, who received pointed commentary on the processors’ architecture and performance, as they were to the audience.86

  Most of the early microcprocessors were used as control devices, not in computers. An applications book that Regis McKenna developed in 1972 to explain “how people use microprocessors” included an airport “marijuana sniffer” to find illlicit drugs in luggage, a device to flush all the commodes in a bank of toilets simultaneously to save water, blood analyzers, traffic light controllers, simple testers, and machine tool controls. When Gelbach presented this list to the board, one director turned to him and asked, “Don’t you have any customers you could be proud of?”87

  NOYCE SPENT MUCH OF THE END OF 1971 trying to placate customers who had bought large numbers of 1103s and were frustrated by the device’s problems. Mike Markkula, who joined Intel’s marketing group in 1969, recalls how Noyce often defused problems: “Bob was just so straightforward and didn’t try to sweep things under the carpet. He’d say, ‘We know what the problem is. We’re fixing it. Here’s when we’ll get it fixed. We’re doing everything humanly possible to meet your requirements—and yeah, we goofed up.’ You know, when you’re that honest and that straightforward, it’s hard [for the customer] to continue to be angry.” M
arkkula saw Noyce use this approach several times with Burroughs, which was using the temperamental 1103s in one of its computers. “Without Bob’s interaction on many occasions, that relationship would have died,” Markkula says. “There was no way that we could keep the whole thing together without his help. [Noyce] would say, ‘Okay, what do you want me to tell these guys?’ And we’d give him all the background, all the scoop, and he would go [talk to them]. And he was really good at it.” Concurs Roger Borovoy, “If you have a disgruntled customer, the best thing you can do is send Noyce. First of all, he’s an instant sponge. Sits down with whomever, the marketing guy or whatever, and instantly absorbs every detail in his head in ten minutes. And then he goes out there and beautifully massages [the details], and he knows everything, and it looks like Noyce—this great god Noyce—is intimately involved in this customer’s problem.”88

  Noyce was proud of his ability to calm hostile customers, but his favorite part of his job was what he called the “missionary work … to spread the word about microprocessors.” He certainly adopted a bit of a religious zeal about the microprocessor-driven world-to-come and was happy to talk about it in the most unlikely places. In 1972, he and Betty surprised his parents with a fiftieth anniversary tour of San Francisco that included the entire extended family. The Noyce clan was riding in a bus Bob and Betty had chartered when Noyce made his way to the front and faced his seated family. “Everybody,” he said, “I want you to see this.” He held up a silicon wafer about three inches in diameter and printed with microprocessors. “This is going to change the world. It’s going to revolutionize your home. In your own house,” he continued, clearly caught up in the moment, “you’ll all have computers. You will have access to all sorts of information. You won’t need money any more. Everything will happen electronically.” He carefully handed the wafer to the person nearest to him and sat back down. At least one person on the bus, the fiancée of Noyce’s nephew, was surprised by her would-be uncle’s showboating and appalled at the arrogance she thought it implied. She remembers thinking, “This guy really thinks he can see the future. This will never happen.”89