Google celebrated the 84th birthday of Robert Norton "Bob" Noyce with a "doodle" depicting Google's name in the middle of a microchip. Noyce invented the integrated circuit shortly after founding Fairchild Semiconductor in 1957. The rest, as they say, is history.
Some amplifications: the Google "doodle" is the graphic that appears on the search home page; Google often uses it to recognize a national holiday, to honor an important person or just for fun. Jack Kilby, at Texas Instruments, had the idea for the integrated circuit at the very same time: Noyce and Kilby share the credit for this seminal invention.
The son of a congregational clergyman, Bob Noyce was the youngest of three brothers. The eldest, Donald Sterling Noyce, was a distinguished professor of chemistry at UC Berkeley; he lived to 81. The middle brother, Gaylord Brewster Noyce, was a beloved member of the faculty at Yale Divinity School; he passed only two years ago at 83.
Bob Noyce was one of those "scary smart" people. His Grinnell College classmates awarded him the Brown Derby prize, given to the senior "who earned the best grades with the least amount of work." As a physics graduate student at MIT, his quick mind won him the accolade "Rapid Robert" from his fellow graduate students.
Noyce lived his life at the forefront of technology and "on the edge." Whether it was starting companies that changed the course of an industry (Fairchild Semiconductor and then Intel), or his hobbies of hang gliding, flying, scuba diving, Noyce was always at the frontier -- showing the way and living with unashamed gusto.
I spent an hour with Bob Noyce shortly before his untimely death at 62 in 1990. He left his post at Intel to head SEMATECH, a not-for-profit consortium of 14 semiconductor companies with support from the Defense Advanced Research Projects Agency (DARPA). I had just become Chief Technical Officer of NCR, and the head of our Microelectronics Division, Jim van Tassel, persuaded me to join the SEMATECH board. As a new board member, I asked for some one-on-one time with Noyce. It was a chance to get to know an industry icon and a personal hero.
There was nothing flashy about Noyce's office. He brought the same style of free-wheeling debate, open communications and disdain for perks to SEMATECH that he so effectively espoused at Intel.
After a few desultory questions about the structure and operation of SEMATECH, I got to what was really on my mind: what could persuade a visionary, a pioneer, and a free spirit like Noyce to take on leading something as Byzantine as a research consortium, with federal government involvement, just to make it more complex? What on earth?
Noyce smiled in a way that made it clear that I was not the first person to ask. "Let me tell you a story," he began. "Did you know that the US semiconductor industry wiped out the Japanese semiconductor industry, not once, but three times?"
After World War II, Japan lay in ruins -- its industrial infrastructure was largely destroyed by US bombing raids. While the Marshall Plan helped Europe rebuild, Japan was left largely on its own. The Ministry of International Trade and Industry (MITI) oversaw the rebuilding effort.
MITI wanted Japan to have a modern economy. They focused on steel mills and ship building as the obvious first move for an island nation with few natural resources. But MITI appreciated early on that electronics would be important to a modern economy, and encouraged fledgling efforts both at tiny start-ups and established keiretsu (diversified trading companies).
Meanwhile, back in the lab -- to be specific, at Bell Labs -- the United States invented the germanium transistor, and with it, the semiconductor industry. The budding Japanese electronics industry, predicated on then state-of-the-art vacuum tubes, went reeling. Strike one.
So, the Japanese also went into the germanium transistor business, realizing that the small, light and low-power transistor was the future of electronics. Japanese semiconductor companies got going. They laboriously learned how to make somewhat satisfactory germanium transistors.
Meanwhile, back in the US of A, we invented the silicon transistor. Cheaper, better and easier to make, it quickly rendered the germanium transistor obsolete. The US semiconductor industry, in its infancy, went on its merry way -- not noticing that it had just caused the nascent Japanese semiconductor industry to totally collapse. Strike two.
So the Japanese re-started, re-tooled, and re-entered the semiconductor business to make silicon transistors. Just as they were getting somewhere, Noyce and Kilby invented the integrated circuit. Instead of making transistors one at a time, US companies were now laying down four, eight, and 16, and more and more at a time. The Japanese semiconductor industry collapsed -- again -- as an unanticipated side effect of progress in the U.S.
After three strikes, MITI woke up. They realized that it was not the intention of the U.S. to prevent Japan from having an electronics industry of its own. But the cycle of following, copying and then being obsoleted by some new invention would just keep happening. So, MITI got the Japanese companies to work together -- sharing process knowledge, suppliers, equipment, and research while they still competed in the marketplace with end products. They wanted not just to learn what the Americans knew, but to be able to advance the technology by themselves.
By the 1980s, MITI's effort paid off. Japanese semiconductor makers were on a roll, pushing U.S. companies to the brink and dominating product after product - particularly DRAMs, the key form of computer memory. Japanese products had better quality and cost less.
So, Noyce ended the story, "It's our turn to learn." And he couldn't think of a more important challenge for himself.
As I said, Noyce was scary smart. The idea behind the integrated circuit is disarmingly simple. People made transistors by making several hundred individual transistors on a "wafer" of germanium or silicon. Then the wafer would be cut into individual devices with a diamond saw. Then each little "chip" would be put in a metal can -- many, many times the size of the chip. Then the cans would be placed on circuit boards to wire the transistors together to make a useful circuit.
The insight that won Noyce and Kilby the admiration of the ages is simply this: why not wire the transistors together before cutting them into chips rather than after? Literally, a trillion dollar idea.
At SEMATECH, Noyce pursued another scary-smart simple idea. You have to learn how to cooperate in order to be able to compete.