The wiring problem
In the 1950s, electronics designers could build complex machines from transistors, resistors and capacitors, but every part had to be connected by hand, usually by soldering. A computer might need many thousands of parts and many thousands of joints, and every joint was a possible failure. Engineers sometimes called this the tyranny of numbers: the more parts a design had, the more unreliable and costly it became to build. The goal was to make the parts and the connections together.
Earlier ideas
The concept of an integrated circuit was in the air. In 1952, a British radar engineer, Geoffrey Dummer, described the idea of electronic parts built into a single solid block, and he later promoted it, but he did not produce a working device. Other groups, including military-funded programs, were also exploring miniaturized circuits. The important point is that when Kilby and Noyce succeeded, they were working in a context where many people expected something like it.
Jack Kilby at Texas Instruments
Jack Kilby joined Texas Instruments in the summer of 1958. Because he was new, he did not have vacation time that colleagues took, and he used the quiet period to work on the problem. His insight was that resistors, capacitors and transistors could all be made from the same semiconductor material, so an entire circuit could be built from a single piece. On September 12, 1958, he demonstrated a simple oscillator circuit on a small slice of germanium. The components were formed in the germanium, but they were connected by fine wires attached by hand, which would not suit mass production.
Robert Noyce at Fairchild
At Fairchild Semiconductor, Robert Noyce approached the problem with a different tool. Jean Hoerni’s planar process, developed in 1959, left a flat surface covered with a protective oxide layer. Noyce proposed in 1959 that metal lines could be deposited on top of the oxide to connect the components on the surface, avoiding separate wires. His design used silicon, and it could be manufactured with the same photographic-style steps used for planar transistors. This approach, with connections made as part of the process, is the foundation of modern chips.
What each contributed
Kilby is generally credited with demonstrating the first working integrated circuit, and Noyce with a design that was practical to manufacture in quantity. They worked separately at different companies, and neither relied on the other’s finished work. Historians therefore describe the invention as simultaneous, with complementary contributions. Choosing one name as the single inventor oversimplifies the record.
Patents and a long dispute
Texas Instruments filed a patent application for Kilby’s work in early 1959, and Fairchild filed for Noyce’s interconnection method later that year. The overlap led to interference proceedings, a legal process to decide priority, and years of court challenges. At one stage, a patent appeals court sided with Noyce’s claims on some points. Eventually the companies settled in 1966 with a cross-licensing agreement under which each recognized the other’s contribution. The legal arguments are technical, and summaries sometimes disagree about who won what. It is accurate to say the dispute was real, long-running and settled by agreement.
Early use
The first integrated circuits were expensive and did little by later standards. Their early customers were the US military and the space program, where size and weight mattered more than cost. Chips were used in guidance computers for missiles and in the Apollo guidance computer, which relied on integrated circuits for its logic. As production increased and prices fell, chips moved into calculators and commercial computers. The shift from a laboratory demonstration to a mass-market product took several years and relied on manufacturing improvements.
Recognition
Kilby received the Nobel Prize in Physics in 2000 for his part in the invention of the integrated circuit. Noyce had died in 1990, and Nobel prizes are not awarded posthumously, so he was not eligible. In his Nobel lecture, Kilby acknowledged Noyce’s role. Noyce later cofounded Intel with Gordon Moore in 1968.
Why it matters
Integrated circuits are the foundation of modern electronics, from the microprocessor in smartphones to the memory in a car. The story is a good example of how a major invention often has more than one parent, and how patents can formalize credit in ways that do not match what historians describe.