Kilby, Noyce, and the Integrated Circuit
The transistor was The Invention of the Century. In reality it was little more than a miniaturized vacuum tube. A vastly improved vacuum tube, to be sure, but a vacuum tube nonetheless. It could do the same things, only better. It was smaller, less expensive, consumed less power, and in a switching mode could operate much faster. It was a discrete component, and when combined with other discrete components (resistors, capacitors, diodes) it could form a powerful electronic circuit.
But wasn’t it ironic that these components, all formed on substrates of germanium or silicon, were then combined and soldered together on another substrate to form an actual circuit? Might there not be a better way?
Two engineers, Jack Kilby of Texas Instruments and Robert Noyce of Fairchild Semiconductor, thought there might be. Working independently, miles apart, they tackled the problem. And almost simultaneously they solved it. They invented the integrated circuit.
Just what is an integrated circuit? It is an electrical circuit consisting of transistors, diodes, resistors, and capacitors. But these components are integrated during manufacture; they are not placed individually on a slab of germanium or silicon as individual components.
Before the invention of the integrated circuit these many components, each manufactured in separate processes, were placed in bins before a factory worker. One by one the components were taken from the bin, usually with tweezers, and placed in a particular position on a printed wiring card. The result was, of course, a circuit, but it was a circuit laboriously created by hand.
Wouldn’t it be nice, mused Kilby and Noyce, if all this manual labor could be eliminated? And they found a way to do just that.
The technology used is photolithography. A screen is used to cover, or expose, a particular part of the germanium or silicon slab (the substrate). That part may be a small square or a line or any other desired shape.
Next, a chemical-laden vapor strikes the exposed areas, turning these areas into a diode, a part of a transistor, or a resistor. A second screen then exposes, or insulates, different areas of the substrate. More deposition continues and eventually the circuit is complete.
The chip now contains all the components necessary to perform an electronic function. And of course this process is performed on thousands of chips, simultaneously.
The first integrated circuits contained only a few transistors. But technology improved, and soon there were hundreds, thousands, and even millions of circuits on a slab the size of a postage stamp. These chips are now everywhere – in our space vehicles, in toys, in cars, and in our cameras and telephones.
And it all started with Jack Kilby and Bob Noyce.
Jack Kilby was a mid-westerner, through and through. He was born in Missouri, raised in Kansas, and educated in Illinois and Wisconsin.
His first job, after receiving his master's from the University of Wisconsin, was with Globe Union of Milwaukee. In 1958 he took a job with Texas Instruments, and as a newly employed engineer who didn't rate a vacation spent the summer working. He was concerned, at that time, with "the tyranny of numbers." There were always too many of everything, and he came to the conclusion that by manufacturing circuit components en masse on a single piece of semiconductor material the problem might be solved. He tried it; it worked; and on September 12 he presented his finding to management. A patent for a "Solid Circuit made of Germanium" was filed on February 8, 1959.
The invention of the integrated circuit did not go unrecognized; 40 years later he was awarded the Nobel Prize in Physics. (Robert Noyce would most certainly have shared this prize had he been alive. Nobel Prizes cannot be awarded posthumously.) But Kilby was forever modest; when asked what he did after hearing about the award he answered, "I made coffee."
Bob Noyce was a "preacher's kid" from Grinnell, Iowa. As a kid he did the usual nutty things (stole a pig from a farmer for a college luau; allowed himself to be dragged in a large homemade box kite behind a truck) but he survived it all, and graduated from Grinnell College. Next came graduate work at MIT in the field of semiconductor electronics (the transistor was now only a few years old), and then employment at Philco, in Philadelphia.
In 1955 William Shockley, one of the inventors of the transistor, left Bell Labs and established Shockley Labs in Palo Alto, and in 1956 Noyce resigned from Philco with the intent of joining Shockley. His method of doing this typified Noyce's personality. After a couple of telephone conversations with Shockley Noyce put himself and his wife on a night flight to San Francisco. They arrived in Palo Alto at 6:00 a.m. By 12:00 noon Noyce had signed a contract to buy a house. That afternoon he went to Mountain View to see Shockley and ask for a job - and he got it.
But all was not well, and a year or so later eight Shockley employees, including Gordon Moore and Bob Noyce, jumped ship and formed Fairchild Electronics. It was in 1959, while at Fairchild Electronics, that Noyce came up with the idea for an integrated circuit. His device was silicon-based, and he was about six months behind Jack Kilby of TI.
Fairchild Electronics prospered, with Noyce quickly rising in the ranks. By 1968 he held a dozen patents dealing with transistors and integrated circuits. But why, he thought, should he be just another employee of a huge company? Gordon Moore, one of the few Shockley-defectors still at Fairchild, felt the same way.
And in 1968 the two men decided to form their own company: Intel. They quickly took on a third person, Andy Grove, and in subsequent years they built the premier electronics company to a multi-billion dollar organization.
As is usually the case, young and brilliant engineers, a previous discovery, and a germ of an idea had led to an industry that ushered in the Electronics Revolution.
Bob Stoffels is a telecommunications consultant with more than 50 years of experience. He can be reached at 727.867.5378 or at email stoffels@juno.com.

