(7 of 13)
On his first day of work, Grove knew exactly none of this. He merely wanted to make a good impression. Nervous? You can't imagine. Here he was, trained as a fluid dynamicist and going to work in materials chemistry. (The math, everyone promised him, was pretty much the same.) Someone asked him to study the electrical characteristics of MOS. Grove delivered a sharp, comprehensive report. His bosses were impressed.
Grove and two colleagues he discovered in the company cafeteria--Bruce Deal and Edward Snow--then set out to make silicon usable. After months of work, they discovered that most of the MOS instability was traceable to an impurity--sodium--introduced when the chips were cured. Like a drop of lemon juice added to a cup of milk, sodium soured the precious semiconductors. The discovery solved a fundamental problem in materials science and set the stage for the semiconductor revolution. Grove and his team won one of the industry's most prestigious awards for the work. At home, Eva got a hint that Andy might not be your ordinary Hungarian busboy. It was the kind of scientific triumph Grove craved--proof of the American meritocracy. At Fairchild, however, none of the suits cared.
By 1968, Noyce was fed up with Fairchild. The firm was blowing up: engineers were leaving, top execs didn't understand the semi business, and science was being replaced by politics. Noyce phoned Arthur Rock, now the eminence grise of Silicon Valley investing, and told him that he and Moore wanted to start their own semiconductor company. Fairchild, he said, was finished. Rock (who holds nearly $500 million of Intel stock today) raised the money nearly instantly. Moore told Grove of the plan one day when they were at a conference in Boulder, Colo. The decision to join his bosses was made, Grove says, "almost instantly." Someone suggested the name Integrated Electronics, which was shrunk instantly to Intel.
Intel did not enjoy an uninterrupted march to greatness. The problem wasn't any lack of candlepower--Noyce, Grove and Moore were a dream team. The problem was the business itself. It kept changing. Just as Intel's leaders decided the future was in, say, selling dynamic RAM (a kind of short-term computer memory), messages started trickling back that sales were tanking, customers were evaporating and, ahem, top management had better pick a new strategy. It was a miserable way to run a company: desperately leaping into lifeboats, always at the last possible moment. One night Grove dreamed he was being chased by a pack of wild dogs. "It was a pressure cooker," he says.
But misery loved the company. The years of anguish produced rich rewards made possible by some neck-snapping breakthroughs. The key to the success dated back to an insight Moore had in 1965. Sitting down with a piece of log paper and a ruler, he drew a simple graph. On the vertical axis he tracked the growing complexity of silicon chips, along the bottom he ticked off time, and then he plotted the points out a few years. The resulting line, he saw, showed that chip power doubled roughly every 24 months, even as costs fell by half. The rule (amended to 18 months) became known as Moore's law. Though it frustrates consumers--it's the reason that $2,500 PC you bought will be obsolete in a year--the law has given Intel a road map, allowing the company to shift resources ahead of demand rather than jumping crazily after the fact.