Monolithic integrated circuit

The integrated circuit emerged twice within six months, invented independently by two engineers who had never met, working at companies nine hundred miles apart. The convergent emergence of Jack Kilby's demonstration at Texas Instruments in September 1958 and Robert Noyce's conception at Fairchild Semiconductor in early 1959 proves that the adjacent possible had aligned completely. The invention was inevitable; only the credit was contested.

Kilby was a new hire at Texas Instruments in Dallas during the summer of 1958. While his colleagues vacationed, he worked alone on a problem that had frustrated the electronics industry: the "tyranny of numbers." As circuits grew more complex, the hand-wired connections between components became the limiting factor. Kilby's insight was that transistors, resistors, and capacitors could all be made from the same semiconductor material and connected on a single chip.

On September 12, 1958, Kilby demonstrated a working integrated circuit—a phase-shift oscillator built on a slab of germanium about the size of a pencil eraser. The components weren't truly integrated; they were connected by tiny gold wires bonded by hand. But the principle was proven: multiple electronic components could coexist on a single piece of semiconductor.

Six months later, Robert Noyce at Fairchild Semiconductor had a different insight. Jean Hoerni had just invented the planar process, which left a protective oxide layer on top of transistors. Noyce realized that aluminum metal lines could be deposited directly on this oxide layer, interconnecting components without external wires. His planar integrated circuit was truly monolithic—a single stone—with everything fabricated in place.

The patent dispute lasted a decade. Texas Instruments filed first, but Noyce's architecture was more practical. The courts eventually split the credit: Kilby received recognition for the concept; Noyce received the patent for the manufacturable implementation. The two companies cross-licensed their technologies in 1966, and modern integrated circuits are based on Noyce's monolithic design.

The cascade from integrated circuits reshaped computing economics. Before the IC, each transistor had to be manufactured, tested, and wired individually. Complexity had a cost that grew faster than linearly. With integrated circuits, complexity became essentially free—adding more transistors to a chip cost almost nothing once the design was complete. Gordon Moore, Noyce's colleague at Fairchild, observed that transistor density doubled every two years, a pattern that held for half a century.

Kilby received the 2000 Nobel Prize in Physics for his contribution to the integrated circuit. Noyce, who had died in 1990, could not share the prize. Both men are remembered as co-inventors, their convergent insight the foundation of the digital age.