"I… WANT… TO… GET… RICH"

On a September evening in 1957, eight men sat around a table at the Clift Hotel in San Francisco and signed ten one-dollar bills. The bills had been pulled from the wallet of Alfred Coyle, a partner at the New York investment bank Hayden, Stone and Company, who had flown west with a junior associate named Arthur Rock to talk to a group of disgruntled physicists and engineers. Each of the eight men signed every bill in turn, so that by the end of the ritual each carried the autographs of all the others. There was no other paperwork. The bills were the contract.

The next morning, on September 18, the eight resigned from Shockley Semiconductor Laboratory in Mountain View. Their boss had won a Nobel Prize the year before, for the invention that two of his Bell Labs colleagues had largely figured out without him, and he had spent the months since acting like a man who believed that the prize had ratified his every instinct. He had begun to require staff to take psychological tests and submit to lie-detector exams. He published everyone’s salary. He demanded that researchers rank one another’s competence. When a small cut on a secretary’s hand turned out to be from a thumbtack rather than a saboteur’s malice, he had her tested anyway. Robert Noyce, who at twenty-nine was the unofficial leader of the dissenters, would later describe the year at Shockley’s lab as the most painful of his professional life. William Shockley, when he learned what they had done, called them the traitorous eight, a label they would wear for the next half-century with something close to amusement.

The eight were Noyce, Gordon Moore, Jean Hoerni, Jay Last, Sheldon Roberts, Eugene Kleiner, Victor Grinich, and Julius Blank. Among them they held doctorates in physics, chemistry, metallurgy, and electrical engineering from MIT, Caltech, Stanford, Cambridge, and Geneva. They had come to the Santa Clara Valley because Shockley had come, and Shockley had come because his elderly mother lived in Palo Alto and because Stanford’s provost, Frederick Terman, had been making a sustained pitch to bring high-technology firms to the orchards south of San Francisco. The eight meant to stay. They had no taste for moving back to New Jersey or Massachusetts to work at the laboratories of RCA or Raytheon or General Electric. They wanted to keep doing the silicon work, and to do it on their own terms.

The financing had not been easy. Eugene Kleiner, whose family in Vienna had escaped the Nazis with little more than the clothes they wore, had written a letter to his father’s old broker in New York asking whether Hayden, Stone might be interested in helping a group of researchers start a new company. The letter landed on the desk of Arthur Rock, a thirty-year-old Harvard MBA who had been raised in a candy store in Rochester and who, in the way of postwar New Yorkers, had reasoned his way into finance because finance seemed like the place where consequential things happened. Rock read the letter and was intrigued. He persuaded Coyle to fly out with him. By the time they reached the Clift, Rock had decided that what these men needed was not a job at an existing company but their own company, with their own equity, in their own valley.

Hayden, Stone could not write the check itself. Rock had to find an industrial sponsor willing to provide the working capital and to accept that the eight would run their own shop. He approached, by his later recollection, somewhere around three dozen firms. Most of the executives he spoke with could not understand why eight strangers, none of whom had ever managed a company, none of whom had been the most senior person anywhere they had previously worked, deserved to be set up in business. The pitch sounded, in the corporate vocabulary of 1957, vaguely insulting. Companies hired employees. Companies did not hand them keys.

The man who finally said yes was Sherman Fairchild. Fairchild was the sole heir to one of the largest individual stockholdings in IBM, an inheritance from his father, an early backer of Thomas Watson’s punch-card business. He had spent the inheritance founding more than seventy companies of his own, most of them in aviation and aerial photography, and he ran one of them, Fairchild Camera and Instrument, out of an estate on Long Island. Fairchild was a bachelor with a restless mind, an interest in technology that did not require him to understand the underlying physics, and enough money to indulge a punt on eight men he had never met. He agreed to put up $1.38 million in working capital. In exchange, his company would hold an option to buy out the eight founders for a fixed price of three million dollars if the new firm hit certain milestones. Each of the eight would hold one hundred shares; Hayden, Stone would hold two hundred and twenty-five. The new entity would be called Fairchild Semiconductor.

The eight rented a small building on Charleston Road, on the line between Palo Alto and Mountain View, and started to work. Their first commercial product, a silicon transistor called the 2N697, came out of the lab the next year. It was a so-called mesa transistor, named for the steplike profile of the silicon when seen in cross section, and it was twice as fast as anything its competitors offered. IBM’s Federal Systems Division placed an order for one hundred of them at one hundred and fifty dollars apiece, intended for the bombing computer of the experimental B-70 bomber. The price was roughly thirty times the industry norm for a transistor. Fairchild charged it because IBM would pay it and because the Pentagon, behind IBM, would pay anything for circuitry that could survive launch and reentry and the Mach 3 vibration of an aircraft none of the eight founders would ever see.

Then, two years later, Jean Hoerni produced the planar process, the manufacturing innovation that would do more than any single piece of physics to make the integrated circuit a mass-market commodity. Hoerni’s idea was to build the entire transistor flat against the silicon surface, sealing each layer beneath a protective film of silicon dioxide. The mesa devices Fairchild had been shipping were vulnerable to dust and moisture and the slow, mysterious drift of contaminants. The planar device was sealed and reproducible. It could, in principle, be made in great numbers. By the time Noyce, that same year, sketched a way to wire several planar transistors together on a single piece of silicon, the company had the foundations of the integrated circuit, the manufacturing technique to produce it, and a question it had not yet answered. What was it for?

The answer, in the early 1960s, was not obvious. The Air Force and NASA were the most reliable customers, paying the kind of cost-plus margins that had funded Fairchild’s first year. But the military market was small and politically volatile, and Noyce, who had taken over as general manager in 1959, did not want to spend the rest of his career as a vendor to the Department of Defense. He believed the chip belonged in everything: in cash registers, in adding machines, in radios, in consumer goods that did not yet exist because the components had been too expensive to imagine them. The way to get there, he reasoned, was not to wait for demand but to break the price.

In 1961 Fairchild listed its first integrated circuit, a flip-flop called the Micrologic, at one hundred and twenty dollars. Two years later Noyce slashed the price of the equivalent commercial part to under twenty. By 1965 Fairchild was selling integrated circuits for less than the sum of the discrete transistors they replaced, sometimes for less than they cost the company to produce, on the bet that volume orders would arrive and pull the unit cost down behind them. The bet worked. The order from Burroughs in 1966 was for twenty million chips. By 1968, civilian computer makers were buying more chips than the Pentagon. The American consumer electronics industry had a part to design around. The integrated circuit had stopped being a curiosity and become an input. Moore would later argue, in an oral history at the Computer History Museum, that the price cuts had been an innovation as consequential as anything that came out of the lab.

The pricing strategy depended on a culture, and the culture was the second thing Fairchild built. Noyce had grown up in Grinnell, Iowa, a small Congregationalist college town founded by an abolitionist who had outlawed liquor within his city limits, and he carried out of it a midwestern aversion to ostentation that fit oddly with his appetite for risk. The decisions he made about how to run a company were small and consistent. There would be no executive dining room at Fairchild. There would be no reserved parking spaces. The vice presidents would sit in the same cubicles as the engineers, behind the same five-foot partitions. If a manager came in late, he would have to walk in from the back lot like everyone else. Disputes were settled in open meetings. Engineers who came to Noyce for a decision would often be told that the decision was theirs, that they had been hired for their judgment and were expected to use it.

This was not how things were done back East. At RCA in Camden, at General Electric in Schenectady, at IBM in Armonk, the corporate apparatus of mid-century America still ran on the assumption that companies were miniature kingdoms, with kings and lords and offices that grew larger and better paneled as one ascended. Tom Wolfe, when he wrote his profile of Noyce for Esquire in 1983, would describe the eastern executive style as a kind of feudalism: the carved paneling, the limousines and drivers, the dining rooms with sconces and pleated peach-colored shades. Noyce had refused to import any of it. The young engineers Fairchild was hiring out of Berkeley and Stanford had no use for the protocol of an American corporation. They wanted, in Wolfe’s reading of them, to work, and they wanted to be paid in a currency that mattered. Wolfe, who had a sociologist’s ear for the small details that mark a tribe, noted that Noyce was the rare kind of leader for whom these arrangements were not a marketing pose. He had simply concluded that the old structures were inefficient.

The currency that mattered was equity. When Fairchild Camera and Instrument exercised its buyout option in 1959, two years after the founding, the eight men who had signed the dollar bills walked away with about a quarter of a million dollars each in stock, an amount roughly equivalent to two and a half million in present-day terms. None of them was rich by the standards of the East Coast industrial families who had funded them. All of them were rich by the standards of an academic physicist or a junior engineer at RCA. The implication was not lost on the people watching. A small group of researchers had walked out of an established firm, signed ten dollar bills, and within twenty-four months made themselves wealthy by building a better transistor. The story became, in the Bay Area, an existence proof.

Sherman Fairchild’s New York leadership did not, however, allow Noyce to extend the same arrangement to the engineers he was hiring. Stock options at Fairchild Camera were reserved for executives, by the convention of the era. The young silicon engineers at Mountain View watched the founders’ money compound while their own salaries crept up by single-digit percentages. They began to leave. Jay Last and Jean Hoerni left in 1961 to start Amelco. A man named Charlie Sporck, who had been brought in from a General Electric tube plant in Hudson Falls to run Fairchild’s manufacturing line, left in 1967 to take over a small money-losing firm called National Semiconductor, taking several lieutenants with him. Spence Lindgren and Bob Widlar went to National. Jerry Sanders, a flamboyant marketing executive whose suits were tailored, would leave in 1969 with seven others to found Advanced Micro Devices. Don Valentine left to start Sequoia Capital. Eugene Kleiner left to start what would become Kleiner Perkins. By the end of the decade, the family tree of Fairchild Semiconductor had begun to bear fruit.

Noyce and Moore were the last to go. By 1968 Fairchild Camera’s New York management had grown sufficiently disconnected from its semiconductor division that Noyce, when he was passed over for the corporate presidency in favor of an outsider named Lester Hogan, decided he was finished. Moore went with him. They called Arthur Rock, who by then had moved from Hayden, Stone to a partnership of his own in San Francisco, and asked him to put together a financing for a new company. Rock raised two and a half million dollars from his contacts in a single afternoon, allegedly with handshake commitments and a one-page document. The company they founded was called Intel, and it would inherit not only Noyce and Moore but the whole portable culture of Mountain View: the open offices, the absence of perks, the egalitarian rituals, the stock options for everyone.

This was the Bay Area’s contribution to American capitalism, and it had taken a decade to fully assemble. The pieces were there in 1957, in the dollar bills at the Clift, but they only resolved into a model after the buyout, after the price cuts, after the spinouts. The model had four parts: a small founding team, willing to leave a stable employer over questions of culture or upside; a financier willing to take a personal stake in unproven scientists; an equity structure that distributed wealth broadly enough to align the interests of the engineers with the company; and a flat, fast operating culture that minimized the friction between an idea and a product. None of the four parts had been invented from scratch in the Santa Clara Valley. Sherman Fairchild was a New York industrial financier; Hayden, Stone was a Wall Street firm; the planar process emerged from materials science problems that ran back through a dozen prior labs. What was new was the combination, and the willingness of a generation of young American technologists to organize their working lives around it.

The historian Christophe Lecuyer, who studied the early valley firms in detail, drew a sharp distinction between the previous generation of Bay Area electronics entrepreneurs, who had built companies like Varian Associates and Hewlett-Packard as quasi-civic institutions, and the Fairchild cohort, who, he wrote, treated a firm as a vector for financial gain and market penetration. The shift was not cynical. The Fairchild engineers saw no contradiction between getting rich and changing the world. They believed that price cuts would deliver the integrated circuit to American industry, that the integrated circuit would deliver computing to American consumers, and that they, the founders and the early employees, would become wealthy in proportion to the amount of useful technology they put on the market. The argument was clean enough that no one needed to articulate it. By the late 1960s the engineers leaving Fairchild for the next startup did not need to explain why they were going. The reason had become so embedded in the local culture that it functioned as a kind of common sense.

Tom Wolfe, looking back on this from 1983, would write that the young engineers of the valley were not the casual, sensual bohemians the rest of the country imagined when it heard the word California. They were the opposite: relentless, monogamous about their work, often unhappy in their marriages because the work consumed everything else. They went to a bar in Mountain View called the Wagon Wheel after their shifts and traded gossip about which company was hiring, which fab line was struggling, who had quit and who was about to. Wolfe noticed that the marriages decayed at a rate comparable to the divorce rate in the NASA towns of Florida. He noticed, too, that none of the engineers seemed to mind. They were getting somewhere.

What they were getting somewhere toward, as the 1960s ended, was a market that had not existed when the eight signed their dollar bills. The Pentagon was no longer their only customer. The price of a logic gate had collapsed by more than an order of magnitude. The chips were going into adding machines and traffic lights and radios and, soon enough, the first programmable computers small enough to sit on a desk. Noyce, in his fortieth year, had become the public face of an industry on the verge of reorganizing itself around a new product, the microprocessor, which a young Intel team would design at the start of the next decade. The company Noyce and Moore had founded in 1968 would carry forward not only the engineering but the entire compensation structure Fairchild had pioneered, and the spinouts that would follow Intel, and the spinouts of the spinouts, would carry it again. What had been signed at the Clift was not a partnership agreement but a way of working that an American technology industry would, for the next half century, refuse to abandon.