Prior Context
By the late 1950s, the transistor had replaced the vacuum tube in many electronic devices, but engineers building complex circuits, especially for computers and military guidance systems, still faced a practical barrier sometimes called the tyranny of numbers. Advanced circuits could require thousands of individual transistors, resistors, capacitors, and diodes, each manufactured separately and then wired together by hand, one solder joint at a time, and the more components a circuit needed, the more opportunities existed for a faulty connection, a cracked solder joint, or a short circuit to make the whole assembly unreliable.
This made highly complex circuits slow, labor-intensive, and expensive to build, and difficult to make reliable enough for demanding uses such as missile guidance or spacecraft. Researchers at several companies, including Texas Instruments, Fairchild Semiconductor, and Bell Labs, as well as engineers within the U.S. military's electronics programs, recognized that reducing the number of separate parts and manual connections was essential if electronics were going to become more powerful without becoming unmanageably large, complicated, and failure-prone.
Historical Development
In the summer of 1958, Jack Kilby, an engineer recently hired at Texas Instruments who had to work through the company's traditional vacation shutdown because he had not yet accrued leave, used that time to pursue an idea for building an entire circuit from a single piece of semiconductor material. By September 1958, he had built the first working integrated circuit, fabricating a transistor, several resistors, and a capacitor out of a single piece of germanium and connecting them with fine external gold wires bonded by hand, demonstrating that an entire oscillator circuit could be built on one chip rather than assembled from separate parts. Kilby filed a patent application for the device in February 1959.
Working independently at Fairchild Semiconductor in Mountain View, California, Robert Noyce developed a different and ultimately more practical approach around the same time. Building on the planar process that his colleague Jean Hoerni had recently developed for manufacturing transistors, Noyce proposed creating the transistors, resistors, and the metal wiring connecting them directly on a flat silicon surface using photolithography, eliminating the fragile hand-wired connections of Kilby's design entirely. Noyce filed his own patent application in July 1959, and the overlapping claims led to a prolonged legal dispute between Texas Instruments and Fairchild over which company held priority, a dispute that worked its way through the patent office and courts for roughly a decade before being resolved through cross-licensing agreements that allowed both companies to manufacture and sell integrated circuits.
Noyce's silicon-based, planar approach proved far more readily manufacturable than Kilby's germanium device and became the technical basis for the industry that followed. Commercial integrated circuits reached the market in the early 1960s, found early major customers in military missile guidance programs and NASA's Apollo spacecraft and lunar module guidance computers, and by the 1970s had become the standard building block of essentially all electronic devices.
Key Participants and Institutions
Jack Kilby and Robert Noyce are jointly credited as the integrated circuit's inventors, having arrived at the core idea, that multiple electronic components could be built together on a single semiconductor chip, independently and within less than a year of each other in 1958 and 1959. Kilby, trained as an electrical engineer, brought experience in miniaturized circuit design from earlier work on hearing aids, and his germanium device proved the basic concept was workable even though it still relied on manual wiring between components. Noyce, a physicist and one of the founders of Fairchild Semiconductor, combined his own ideas with the planar transistor process developed by his colleague Jean Hoerni to solve the manufacturing problem using printed metal connections, making mass production genuinely practical rather than merely possible in a laboratory.
Kilby later received the 2000 Nobel Prize in Physics for his role in the invention, sharing it with researchers honored for separate work in semiconductor physics; Noyce, who died in 1990, was not eligible for a posthumous award, but he is recognized alongside Kilby in most historical accounts, and he went on to co-found Intel in 1968 with Gordon Moore, extending his influence over the semiconductor industry for decades. Texas Instruments and Fairchild Semiconductor, as institutions, supplied the funding, equipment, and commercial channels that turned both inventions into manufactured products.
How It Works
An integrated circuit, often called a chip, combines multiple electronic components, transistors, resistors, capacitors, and the wiring between them, onto a single small piece of semiconductor material, almost always silicon. Rather than manufacturing each component separately and wiring them together by hand, manufacturers use photolithography: a silicon wafer is coated with a light-sensitive chemical, light is projected through a patterned mask to expose precise regions of that coating, and the exposed or unexposed areas are then chemically etched away or treated, building up transistors, insulating layers, and interconnecting metal lines directly in place, one microscopic layer at a time.
Because the components and their connections are formed together in the same sequence of manufacturing steps, a single chip can contain enormous numbers of transistors, numbering in the billions on advanced modern chips, without any hand assembly at all. This batch-fabrication approach also means that many identical chips can be produced simultaneously from one silicon wafer, which is what allows chips to be manufactured in large, consistent, and increasingly inexpensive quantities, a central reason computing power has grown so dramatically, and so cheaply, since the 1960s.
Dates and Milestones
- 1947The invention of the transistor provides the essential building block for later integrated circuits.
- 1958Jack Kilby builds the first working integrated circuit at Texas Instruments.
- 1959Robert Noyce develops a silicon, planar integrated circuit at Fairchild Semiconductor; both Kilby and Noyce file patents.
- early 1960sCommercial integrated circuits reach the market and are adopted by military and space programs, including NASA's Apollo program.
- 1965Gordon Moore articulates the trend of rapidly increasing transistor density on chips, later known as Moore's Law.
- 2000Jack Kilby receives the Nobel Prize in Physics for his role in inventing the integrated circuit.
Impacts
The integrated circuit made possible the sustained growth in computing power often described by Moore's Law, the observation, made by Fairchild and later Intel co-founder Gordon Moore in 1965, that the number of transistors that could be fit on a chip roughly doubled every couple of years. This steady increase in component density, combined with falling manufacturing costs, drove computers down in size and price from room-filling mainframes to desktop minicomputers, then to personal computers in the late 1970s and 1980s, and eventually to the processors inside smartphones, enabling computing to spread from specialized institutions into ordinary homes, schools, offices, and pockets worldwide.
Integrated circuits also became essential to space exploration and defense electronics: NASA's Apollo Guidance Computer, which helped navigate the Apollo missions to the Moon, was among the earliest major applications of mass-produced chips, and military guidance systems were early adopters as well. Beyond computing and aerospace, integrated circuits transformed telecommunications switching equipment, enabled the pocket calculator and digital watch industries of the 1970s, and underlie countless other fields of applied electronics, from medical imaging devices to industrial automation to modern automobiles. The chip is widely regarded as the central enabling technology of the digital age, underlying essentially every modern computer, communications device, and embedded electronic system in use today.
Limitations
Early integrated circuits were limited by the manufacturing precision available at the time, which restricted how many components could reliably be placed on a single chip and kept yields, the percentage of chips on a wafer that worked correctly, relatively low and costly. The germanium material Kilby originally used was soon abandoned industry-wide in favor of silicon, which offered better electrical and thermal properties and became the near-universal standard for chip manufacturing.
The legal dispute between Texas Instruments and Fairchild over which company's patent took precedence was never cleanly resolved in court; it ended instead in negotiated cross-licensing agreements, meaning that, unlike many inventions with a single clear patent holder, the integrated circuit has no single, legally settled sole inventor, a fact that continues to complicate popular accounts of its origin.
Common Misconceptions
The integrated circuit is sometimes attributed to a single inventor, usually Kilby, because he built the first working device and later received a Nobel Prize for it, but this overstates his individual role and understates Noyce's contribution. Kilby's 1958 chip proved the concept was possible, but its hand-wired external connections made it impractical to manufacture in volume; Noyce's silicon, planar design, developed independently and patented separately only months later, is what actually became the technical basis for the integrated circuits manufactured ever since.
The two inventions are best understood as parallel, independent breakthroughs rather than one building directly on the other, and most historians of technology credit both men, and the companies and colleagues who supported their work, as co-inventors of the integrated circuit.
