The incident gave the staff a crystal-clear, if unnerving, window into their boss's personality. A couple of months later, Arnold Beckman flew to Palo Alto on a baleful mission. Profits at Beckman Instruments had been shrinking, in part because research and development costs were out of control. When he informed Shockley that his unit would have to refocus its energies on production, Shockley angrily blurted out: "I can take this group and work anywhere else!"
The senior staff in the room were amazed by the scale of Shockley's self-deception; the fact was that not a single one of them would follow him anywhere. It was now clear that the company's future was at stake. After discussing the matter among themselves, Gordon Moore and seven other top scientists decided to relate their misgivings to Beckman.
It fell to Moore to phone Beckman as his colleagues listened in. In an era when most companies were run on strictly hierarchical terms, the entire group understood the grave breach of protocol they were committing by going over Shockley's head. "I can still hear the quaver in Gordon's voice when he asked for Beckman on the phone," Last recalls.
To their relief, Beckman proved a willing listener to their suggestion that he essentially kick Shockley upstairs to continue his theorizing while he brought in an experienced production manager. "We were a long way from a product, and he was very receptive to what we were suggesting," Moore says. Beckman even had a candidate in mind: a manager from one of his other divisions.
But Beckman soon backed off for reasons none of them ever fully understood. It was true that Shockley was furious that the team he had assembled and nurtured could betray him. Some believed that one of Shockley's old colleagues at Bell Labs had reminded Beckman of Shockley's stature as a Nobel laureate and warned that removing him from the management of his own company would be a crushing blow.
Whatever the reason, the eight understood that without Beckman in their corner, their position had become untenable. "We could see that the revolt had failed," says Last. Their only option was to find another company that might hire them as a group. At one point they sat down together with a copy of the Wall Street Journal and scanned its stock listing. "We identified every company we could think of that might conceivably have an interest in semiconductors," says Moore. "There were 35, and they all turned us down."
A contact of theirs at a New York investment firm had suggested an alternative, however: start their own company. Presently the contact turned up a backer. This was the New York industrialist Sherman Fairchild, who thought his high-altitude photography business might be able to exploit the size and efficiency of semiconductor devices. To get Fairchild Semiconductor started he pledged $30,000 in seed capital, in return for an option to buy out the team on a sliding scale, ranging from $2 million after two years to $5 million after five.
The last step was to formally cut the cord with Bill Shockley. On Sept. 18, 1957, Shockley, who had got wind of the planned desertion, started calling the senior staff into his office one by one. "He started with the people he thought were most loyal to him," says Moore, who was one of the first to be summoned.
Moore's relationship with Shockley had remained cordial; as a chemist he had never been treated with quite the arrogance and condescension Shockley reserved for the physicists on the staff. But he was not inclined to sugarcoat the bad news. "I told him, 'There's no use in calling in the rest. They're all going to leave.' "
Shockley seemed to accept the desertion with stoicism. In the ruled notebook that served as his personal journal, he tersely recorded the seismic shock that decimated his company that day with a single handwritten line: Wed 18Sep--Group resigns. But for the rest of his life he would refer to them collectively as "the traitorous eight."
from the moment of the break, Shockley and his former recruits veered off in opposite directions. With Robert Noyce, an organizational and engineering genius, as director of research and Moore as head of production engineering, Fairchild Semiconductor began shipping transistors in 1958, about a year after the break. A few months after that, Noyce achieved the ultimate breakthrough, a way to embed numerous transistors and other devices on a single piece of silicon. This was the integrated circuit, the very first version of what would evolve into the Pentium.
Fairchild went from eight employees to 4,000 in two years. Meanwhile, Shockley Semiconductor dragged. Bill Shockley had finally succeeded in manufacturing his four-layer diode, but the quality was uneven, characteristic of an idea that was too far ahead of its time. There were more staff defections and scant revenue growth until, in 1960, a frustrated Beckman sold the semiconductor unit to another company, which sold it to a third owner in 1965.