The Transistor and the Semiconductor Revolution

In 1925 and 1926 the physicist Julius Edgar Lilienfeld filed patents describing the field-effect transistor, the operating principle later used in the dominant type of transistor. No working device could be built with the materials and manufacturing of the time. The patents show that the idea preceded the working invention by two decades. Filed in 1925 and 1926 and granted in 1930, the patents described source, gate and drain electrodes, the same geometry used in modern field-effect devices. The idea sat unused for decades while Bell Laboratories sought alternatives to fragile, power-hungry vacuum tubes. When the metal-oxide-semiconductor transistor finally became the industry's workhorse, the patent priority of Lilienfeld's concept was litigated and largely upheld.

Location: New York, United States

On 23 December 1947 John Bardeen and Walter Brattain, working under William Shockley at Bell Labs, demonstrated the point-contact transistor, the first working solid-state amplifier. The device replaced bulky, power-hungry vacuum tubes with a crystal of germanium. It marked the beginning of the semiconductor age. The device was built from a slab of germanium with gold-foil contacts in the experiment recorded in the laboratory's notebooks that December. It amplified a signal with no vacuum, no glass envelope and no warm-up time, drawing a fraction of the power of a tube. The result stayed secret inside the phone company for months while the team worked out the physics and considered how to announce it.

Location: Murray Hill, New Jersey, United States

On 30 June 1948 Bell Labs held a press conference in New York to announce the transistor to the world. The announcement drew restrained coverage at first, but within a few years transistors were replacing vacuum tubes in radios, telephone exchanges and hearing aids. Shockley's junction transistor, developed at Bell Labs in the same period, proved easier to manufacture and became the standard form. The New York press conference framed the transistor as a hearing-aid amplifier, and most newspapers gave it only a few lines. Shockley's junction transistor, developed at Bell Labs in the same period, proved easier to manufacture and became the practical design. Commercial radios such as the Regency TR-1 of 1954 showed the consumer potential that the cautious announcement had undersold.

Location: New York, United States

In May 1954 Texas Instruments, led by Gordon Teal, produced the first commercial silicon transistor. Silicon tolerated heat far better than germanium, making transistors practical for military and industrial use. The switch to silicon set the material foundation for the entire integrated-circuit industry. Gordon Teal, a Bell Labs germanium expert hired by Texas Instruments, grew the single-crystal silicon boules that made the devices possible. Silicon's wider band gap let transistors run hot where germanium failed, opening military, industrial and eventually space applications. The achievement made Texas Instruments a leader in the young semiconductor business and anticipated the industry's decisive shift from germanium to silicon wafers.

Location: Dallas, Texas, United States

In December 1956 John Bardeen, Walter Brattain and William Shockley received the Nobel Prize in Physics for their research on semiconductors and the invention of the transistor. Bardeen remains one of very few people to win two Nobel Prizes. The prize cemented the transistor's status as a foundational discovery of the century. The prize cited research on semiconductors and the discovery of the transistor effect. Bardeen remains the only person to win the Nobel Prize in Physics twice, his second coming for superconductivity in 1972. Shockley used the recognition to leave Bell Labs and found his own laboratory near Palo Alto in 1956, an early seed of what became Silicon Valley.

Location: Stockholm, Sweden

On 12 September 1958 Jack Kilby of Texas Instruments demonstrated the first integrated circuit, a single piece of germanium containing several connected components. Robert Noyce at Fairchild Semiconductor independently devised a silicon version suited to mass production. The integrated circuit ended the era of hand-assembled discrete components. Kilby's prototype, a phase-shift oscillator built in germanium, demonstrated the monolithic circuit at a Texas Instruments internal show. Noyce's silicon planar version, patented in 1959, was the design industry could mass-produce, and the two men's claims ran in parallel for years. Together their inventions ended the era of hand-assembled discrete components and made the microchip possible.

Location: Dallas, Texas, United States

In 1959 Jean Hoerni of Fairchild Semiconductor devised the planar process, which protected transistor junctions with a layer of silicon dioxide and allowed many devices to be made on a single wafer. Fairchild's planar transistors reached the market in 1960. The process became the manufacturing basis of the integrated-circuit industry. By sealing the transistor junctions under a protective layer of silicon dioxide, the planar process allowed many devices and their interconnections to be printed on a single wafer. Fairchild's Robert Noyce used it to design the first practical monolithic integrated circuit. The process survives, in evolved form, as the core of every chip fabrication plant in the world.

Location: Mountain View, California, United States

Between 1955 and 1960 Bell Labs researchers, notably Mohamed Atalla and Dawon Kahng, developed the metal-oxide-semiconductor field-effect transistor, first demonstrated around 1960. The MOSFET scaled down more easily than earlier transistors and consumed little power. It became the most widely manufactured device in history and the basic building block of microchips. Atalla and Kahng's device used the silicon-surface oxide layer as a dielectric, controlling current with a gate electrode that drew almost no power. The MOSFET scaled down more easily than any earlier design, making smaller, denser chips a matter of manufacturing rather than new physics. It is routinely cited as the most manufactured device in human history, and it remains the basic building block of every microprocessor.

Location: Murray Hill, New Jersey, United States

On 19 April 1965 Gordon Moore published the observation, later called Moore's Law, that the number of components on a chip was doubling about every year and would continue to do so. Industry planners adopted the doubling as a roadmap and a self-fulfilling target. The projection held for decades and organized investment across the semiconductor supply chain. The article appeared in Electronics magazine, and Moore later revised the doubling period to about every two years in 1975. The projection functioned as an economic coordination device: chip makers that fell behind the curve lost markets, and buyers planned around it. Its slowdown in the 2020s has shifted industry attention to advanced packaging and new architectures as the next sources of performance.

Location: United States

On 15 November 1971 Intel released the 4004, the first commercial single-chip microprocessor, placing an entire central processing unit on a piece of silicon. Originally designed for a Japanese calculator company, it opened the way to programmable electronics. Microprocessors became the core of personal computers, embedded systems and later smartphones. Designed by Ted Hoff, Federico Faggin and Stanley Mazor for the Japanese calculator firm Busicom, the chip's rights were bought back by Intel and marketed as a general-purpose processor. Its 2,300 transistors ran at 740 kilohertz, modest even by the standards that followed within three years. The 4004 created the microprocessor industry and made programmable electronics cheap enough to spread through the economy.

Location: Santa Clara, California, United States

In February 1987 Morris Chang founded the Taiwan Semiconductor Manufacturing Company as a dedicated foundry, manufacturing chips designed by other firms. The model lowered the cost of entry for chip design companies and concentrated manufacturing in Taiwan. Decades later TSMC produced the majority of the world's most advanced chips. The foundry model separated design from fabrication, allowing fabless firms to compete without owning a plant and letting TSMC fill its lines with many customers' designs. Built with strong Taiwanese state support, the company concentrated leading-edge manufacturing on the island over three decades. That concentration became the geopolitical vulnerability the CHIPS and Science Act and European subsidy programmes later tried to counter.

Location: Hsinchu, Taiwan

By 2019 extreme ultraviolet lithography machines built by the Dutch company ASML were in high-volume production, allowing chip features far below the limits of earlier light sources. EUV underpinned the 7-nanometer and smaller chips used in smartphones and data centers. The technology took more than two decades of development to reach the factory floor. EUV machines generate 13.5 nanometre light by vaporizing falling tin droplets with high-power lasers, among the most complex machines ever built and made by only one company, ASML. The technology took more than two decades and billions of euros to move from laboratory to factory floor. Its 2019 rollout enabled the 7-nanometre, 5-nanometre and 3-nanometre chips in modern smartphones and data centres.

Location: Veldhoven, Netherlands

On 9 August 2022 the United States enacted the CHIPS and Science Act, providing about 52 billion US dollars in subsidies and research funding for domestic semiconductor manufacturing. Supporters argued the measure addressed supply-chain risk and strategic dependence on foreign fabrication plants; critics questioned its cost and the use of industrial policy. Comparable subsidy programs followed in Europe, Japan and South Korea. The law funded new fabrication plants, including TSMC and Samsung facilities in Arizona and Texas, and expanded research funding through the National Science Foundation and the Commerce Department. It followed the 2020-2021 chip shortage and tightened export controls on advanced chips and tools to China. Supporters called it a necessary correction to strategic dependence; critics debated the industrial-policy precedent and whether subsidies alone can restore leading-edge manufacturing in the United States.

Location: Washington, D.C., United States