On this page · 3 sections
He was too new for vacation. So he built the microchip.
Jack Kilby was born 103 years ago today, on Nov 8, 1923. He grew up in Great Bend, Kansas, a town named for the spot where the Arkansas River bends, and his father ran a small electric company with customers scattered across the rural western part of the state. When Jack was in high school, a huge ice storm knocked down most of the poles carrying power and phone lines. His father turned to amateur radio operators to reach the customers who had gone dark and silent. The boy watched that and decided radio was the most fascinating thing going.
He also loved Big Band music. Decades later, with a Nobel Prize on the shelf, he admitted there was still a Dallas station that played it, "and with a little luck, I don't have to listen to much else."
He studied electrical engineering at the University of Illinois and graduated in 1947, one year before Bell Labs announced the transistor. That timing stung a little. "It meant that my vacuum-tube classes were about to become obsolete," he wrote. He went to Milwaukee to work at Centralab, a company that made parts for radios, televisions and hearing aids, and took night classes toward a master's degree while he was at it.
By 1958 he wanted to work on making electronics smaller, full time. Texas Instruments in Dallas was the only company that agreed to let him.
The summer the lab emptied out
Kilby started at TI in May 1958 and walked straight into the biggest headache in electronics. Engineers called it the tyranny of numbers. Transistors had replaced bulky vacuum tubes, so designers kept drawing bigger circuits with hundreds or thousands of parts, and every one of those parts still had to be wired to the others by hand. Each soldered joint was one more spot where a machine could fail.
The Army Signal Corps was backing one answer, the Micro-Module, which made every part the same size and shape so they could snap together like building blocks. TI was working on it. Kilby thought it missed the point, since it still left you with a mountain of separate parts.
Then July came, and TI went on what its people called mass vacation. Roughly 90 percent of the workforce cleared out. Kilby had been there about two months and had no vacation time coming, so at 34 he stayed behind in the company's brand-new Semiconductor Building to keep an eye on the place. "I was left with my thoughts and imagination," he wrote later.
He used them. A semiconductor company, he reasoned, could make one thing cheaply, and that was semiconductors. So why build resistors and capacitors out of anything else? On July 24, 1958, he wrote in his lab notebook that resistors, capacitors and transistors, if all made of the same material, could be included in a single chip. He sketched a circuit made entirely from silicon to prove the point.
When his boss, Willis Adcock, got back from vacation, Kilby showed him the sketch. Adcock was enthused, but skeptical, and wanted proof that circuits made only of semiconductors would actually work. So Kilby built one from separate pieces first. He made resistors by cutting little bars of silicon and etching them down to the right values, and he cut capacitors out of power transistor wafers. On Aug 28 the patchwork version ran for Adcock. It proved the materials worked. It still wasn't one piece.
A sliver of germanium on a glass slide
For the real thing, Kilby borrowed what the factory already had: germanium wafers with the transistor layers and contacts already in place. Two technicians, Pat Harbrecht and Tom Yeargan, cut the wafers into tiny bars. Black wax went on by hand to mask the parts that needed protecting before etching. Kilby picked a phase-shift oscillator for the test, a circuit that turns steady power into a repeating wave, because it was the favorite way to show off a new circuit at the time.
On Sept 12, 1958, a small group of TI executives, including Mark Shepherd, gathered at the bench. What they saw did not look like the future. It was a sliver of germanium less than half an inch long, with wires sticking out of it, glued to a glass slide. Kilby hooked it to an oscilloscope and switched it on, and a steady sine wave rolled across the screen. The first unit oscillated at about 1.3 megahertz.
One piece of material was doing the job of a whole circuit. Kilby recorded it on page 20 of his engineering notebook, dated Sept 12, 1958, and signed it "JS Kilby."
TI filed for a patent on Feb 6, 1959. The patent, "Miniaturized Electronic Circuits," was granted in 1964, and it carries one beautifully blunt line of bragging: the method could pack more than thirty million components into a cubic foot, against five hundred thousand for the best anyone had managed before.
The industry was not instantly sold. The integrated circuit and its merits "provided much of the entertainment at major technical meetings over the next few years," Kilby wrote, which is a very polite way to say people argued about it a lot. His hand-wired germanium version, with fine gold wires looping between parts, was also hard to mass-produce. In California, Robert Noyce at Fairchild Semiconductor worked out a flat silicon version that factories could print by the batch. The two men later worked together to get chips accepted.
The military came first. TI built the first computer using silicon chips for the Air Force in 1961, and a TI prototype from that era used 587 integrated circuits to replace 8,500 transistors. Then TI's Patrick Haggerty gave Kilby a challenge: build a calculator as powerful as the big desktop machines, but small enough for a coat pocket. Kilby co-invented the handheld calculator, and the chip finally had a product regular people wanted to buy.
The Nobel Prize in Physics came in 2000, 42 years after that summer, "for his part in the invention of the integrated circuit." Kilby had to start his pot of coffee very early the morning the news arrived. Noyce did not live to see it, and Kilby used his Nobel autobiography to make room for him: "If he were still living, I have no doubt we would have shared this prize."
Kilby died in Dallas on June 20, 2005. The chip in the phone you're holding descends from a summer when the new guy couldn't take time off.
He spent the rest of his life watching people photograph that first crude sliver, with its wax marks and wires and glue. He often remarked that if he'd known he'd be showing it for the rest of his life, he would have "prettied it up a little."
Enjoyed this? Follow Tech Cheat Codes on TikTok for quick tech tips: Follow on TikTok. Or open TikTok and search "Tech Cheat Codes".
References
- Jack S. Kilby: Facts — NobelPrize.org, Nobel Prize Outreach, accessed Nov 7, 2026, 5:06 PM ET (born 8 November 1923; died 20 June 2005, Dallas; prize motivation "for his part in the invention of the integrated circuit"; prize share 1/2; affiliation Texas Instruments)
- Jack S. Kilby: Biographical — NobelPrize.org, Jack S. Kilby, from Les Prix Nobel 2000, accessed Nov 7, 2026, 5:05 PM ET (Great Bend upbringing; father's electric company and ice storm; Big Band radio quote; Illinois 1947 and vacuum-tube quote; Milwaukee and Wisconsin night classes; TI 1958 only company allowing full-time miniaturization; handheld calculator; early coffee; Noyce shared-prize quote)
- Jack S. Kilby: Nobel Lecture (PDF) — NobelPrize.org, Jack S. Kilby, Nobel Lecture Dec 8, 2000, accessed Nov 7, 2026, 5:08 PM ET (started TI May 1958 with no vacation; about 90 percent took mass vacation; "left with my thoughts and imagination"; July 24, 1958 notebook entry; Willis Adcock enthused but skeptical; Aug 28 discrete demo; technicians Pat Harbrecht and Tom Yeargan; black wax masking; Sept 12 three oscillators, first at about 1.3 MHz)
- The Chip that Jack Built — Texas Instruments (Internet Archive copy), Texas Instruments, accessed Nov 7, 2026, 5:08 PM ET (tyranny of numbers; Army Signal Corps Micro-Module; new Semiconductor Building; new employee with no vacation stayed to man the shop; Centralab; Sept 12 demo with Mark Shepherd; germanium sliver with wires glued to glass slide; sine wave on oscilloscope; 7/16 by 1/16 inch; "entertainment at major technical meetings"; Air Force computer 1961; Haggerty pocket calculator challenge)
- The Integrated Circuit — Computer History Museum, Revolution, Computer History Museum, accessed Nov 7, 2026, 5:07 PM ET (Sept 12, 1958 germanium circuit; "prettied it up a little"; company-wide summer vacation; gold flying wires impractical for mass production; Type 502 flip-flop 1959; notebook page 20 signed "JS Kilby"; 587 ICs replacing 8,500 transistors)
- US Patent 3,138,743, Miniaturized Electronic Circuits — Google Patents, Jack S. Kilby / Texas Instruments, accessed Nov 7, 2026, 5:09 PM ET (filed Feb 6, 1959; granted June 23, 1964; component density above thirty million per cubic foot vs five hundred thousand)