How the Transistor Remade the Modern World

In December 1947, inside a cramped laboratory in Murray Hill, New Jersey, a small assembly of germanium, gold foil, and a spring-loaded wire amplified an electrical signal for the first time without using a vacuum tube. The device was ugly, unreliable, and about the size of a fingertip. It also made possible almost every electronic object that would define the following century.

That object was the transistor, and its story rarely gets the attention it deserves. Most people can name the light bulb, the airplane, or the internet as turning points in modern history. Few can explain what a transistor actually does, even though hundreds of billions of them now operate inside a single smartphone.

The central question is not simply when the transistor was invented. It is why replacing one type of electrical switch with another set off a chain of events that produced computers, the internet, and the digital economy. Understanding that chain requires starting with the technology the transistor was built to replace.

The Problem the Vacuum Tube Could Not Solve

Before 1947, electronic amplification and switching depended almost entirely on the vacuum tube, a glass bulb containing electrodes that controlled the flow of electrons through a vacuum. Vacuum tubes had made radio, radar, and early computers possible, but they carried serious drawbacks.

Tubes were fragile, generated significant heat, and consumed considerable power. They also failed often. The ENIAC, one of the first electronic general-purpose computers, used roughly 17,000 vacuum tubes and required constant maintenance because tubes burned out with regularity. Engineers at Bell Telephone Laboratories, then part of AT&T, were especially concerned with a narrower version of this problem: the telephone network depended on thousands of tube-based amplifiers to boost signals across long distances, and each one represented a point where the system could break down.

Bell Labs assigned a research team to look for a solid-state alternative, a device built from a solid material rather than a vacuum, that could switch and amplify electrical signals more reliably. The team included physicist William Shockley, along with theoretical physicist John Bardeen and experimental physicist Walter Brattain. Their work centered on semiconductors, materials like silicon and germanium whose ability to conduct electricity can be precisely controlled.

Three Physicists and a Fragile Breakthrough

Shockley’s original theoretical design for a semiconductor amplifier did not work as predicted. Frustrated by repeated failures, Bardeen and Brattain shifted their approach and began experimenting directly with the surface properties of germanium crystals. On December 16, 1947, they built a working device: two closely spaced gold contacts pressed against a germanium crystal, mounted on a plastic wedge. When they applied a small current to one contact, it controlled a much larger current flowing through the other. That was the first working point-contact transistor.

Bell Labs kept the discovery confidential for several months while patent applications were filed, and the invention was formally announced to the public in June 1948. The name itself was a deliberate blend of “transfer” and “resistor,” chosen internally to describe how the device transferred a signal across a resistance.

The point-contact transistor was a genuine breakthrough, but it was also unstable and difficult to manufacture consistently. Shockley, dissatisfied with a design he had not personally engineered, went on to develop a more robust alternative: the junction transistor, patented in 1948 and demonstrated in the early 1950s. Built from layered semiconductor material rather than fragile point contacts, it proved far easier to produce reliably and became the foundation for commercial transistor manufacturing.

In 1956, Bardeen, Brattain, and Shockley shared the Nobel Prize in Physics for the invention. The three men, however, did not continue working together. Their professional relationship had grown strained well before the award, a tension that would soon reshape the American technology industry in an unexpected way.

The Split That Created Silicon Valley

Shockley left Bell Labs in 1955 and founded Shockley Semiconductor Laboratory in Mountain View, California, recruiting a group of talented young scientists and engineers. His management style, however, was widely regarded by his own employees as controlling and difficult to work under. In 1957, eight of his researchers, later nicknamed the “traitorous eight,” resigned together and founded a new company, Fairchild Semiconductor.

Fairchild would become one of the most consequential companies in the history of computing, not primarily because of Shockley’s original transistor design, but because of what its engineers built on top of it. Robert Noyce, one of the eight founders, helped develop the planar manufacturing process, a technique that allowed transistors to be built directly onto the flat surface of a silicon wafer using photographic masking rather than assembled by hand.

The planar process solved a problem that had limited the transistor’s usefulness: manufacturing consistency at scale. It also created an unexpected opportunity. If transistors could be printed onto silicon, so could the wires connecting them.

One Chip, Many Transistors

In 1958, engineer Jack Kilby at Texas Instruments built the first working integrated circuit, connecting multiple components on a single piece of germanium using fine wires. Independently, Noyce at Fairchild developed a more practical integrated circuit in 1959 using the planar process, printing both the transistors and their interconnections directly onto silicon. Noyce’s approach proved far more suitable for mass production, and it became the model the semiconductor industry adopted.

The integrated circuit changed the basic unit of electronic design. Instead of wiring together individual transistors, resistors, and capacitors by hand, engineers could now etch an entire circuit onto a single chip. This made electronics smaller, cheaper, and dramatically more reliable, since there were fewer physical connections that could fail.

In 1965, Gordon Moore, another of Fairchild’s founders and later a co-founder of Intel, observed in a widely cited article that the number of components engineers could fit on an integrated circuit had roughly doubled every year since 1958, and he predicted the trend would continue for at least a decade. The pattern, later revised to a doubling roughly every two years, became known as Moore’s Law. It was not a law of physics. It was an industry-wide expectation that shaped investment, competition, and research priorities for decades, encouraging manufacturers to treat continuous miniaturization as an achievable goal rather than an occasional accident.

Small Enough to Carry

While engineers were solving problems of manufacturing and scale, the transistor was also changing what ordinary consumers could own. In 1954, the Regency TR-1 became the first commercially available transistor radio, using four transistors licensed from Texas Instruments. It was expensive for its time and modest in sound quality, but it demonstrated that a radio no longer needed the bulk or fragility of vacuum tubes.

The more lasting commercial breakthrough came from Tokyo Tsushin Kogyo, a small Japanese electronics company that would soon rename itself Sony. In 1955, the company released Japan’s first transistor radio, and in 1957 introduced the TR-63, compact enough to fit in a shirt pocket. Sony’s transistor radios sold well internationally and helped establish the company’s global reputation for consumer electronics, years before it became known for products such as the Walkman.

The shift from vacuum tubes to transistors also transformed computing directly. Early transistorized computers of the late 1950s were smaller, cheaper to operate, and far less prone to failure than their tube-based predecessors, a change that made computing practical for a wider range of businesses and institutions rather than only government and military laboratories.

The Switch That Uses Almost No Power

A second major transistor design proved just as important as the original junction transistor, even though it is far less famous. In 1959, Bell Labs researchers Mohamed Atalla and Dawon Kahng developed the metal-oxide-semiconductor field-effect transistor, commonly known as the MOSFET. Unlike earlier transistor designs, the MOSFET could be manufactured at extremely small sizes and consumed very little power while switching, a property that would prove essential once engineers began packing millions, and eventually billions, of transistors onto a single chip.

In 1963, Fairchild engineer Frank Wanlass described a circuit design called complementary MOS, or CMOS, which paired two types of MOSFETs so that current flowed only briefly during switching rather than continuously. CMOS circuits ran cooler and used far less energy than earlier alternatives. By the 1980s, CMOS had become the dominant technology for digital chips, and it remains the basis for virtually all modern processors and memory chips.

This combination, the MOSFET transistor built at increasingly small scales using CMOS design, is what ultimately allowed a single chip to contain the roughly 100 billion transistors found in a modern smartphone processor, switching billions of times per second while generating a manageable amount of heat.

What Popular Memory Gets Wrong

The transistor is often described as the invention of a single genius, usually Shockley, whose name became the most publicly recognized of the three Nobel laureates. The historical record complicates that story. The working transistor was built by Bardeen and Brattain using an approach that departed from Shockley’s own theoretical design, and the commercially practical version of the technology owed as much to the manufacturing innovations of Noyce, Kilby, Wanlass, and their colleagues as it did to the original 1947 breakthrough.

A related misconception treats Silicon Valley’s rise as an inevitable outcome of the transistor’s invention. In fact, it depended on a specific and somewhat accidental sequence of events: a difficult manager, a mass resignation, a manufacturing innovation developed at the resulting company, and years of subsequent engineering work to make miniaturized circuits commercially viable. Remove any one of those steps, and the industry’s geographic and institutional shape might have looked considerably different.

Why the Switch Still Matters

The transistor’s deepest significance lies less in any single device it enabled and more in what it made possible to build in combination. Vacuum tubes could amplify and switch signals, but their size, heat, and failure rates placed a practical ceiling on how much computation a machine could perform. The transistor removed that ceiling gradually, and then almost entirely, as manufacturing techniques advanced from individual components to integrated circuits containing billions of switches.

That progression enabled the personal computer, the internet’s supporting infrastructure, mobile phones, and the broader shift toward a digital economy, developments that unfolded over decades rather than as an immediate consequence of the 1947 discovery. It is worth remembering that the connection between a laboratory prototype and a global economy was neither immediate nor guaranteed. It required manufacturing innovation, competitive pressure, and continuous incremental engineering sustained across generations of researchers.

Physical limits are now narrowing the transistor’s decades-long trend of miniaturization. As transistors shrink toward the scale of a few nanometers, quantum effects such as electron tunneling make it increasingly difficult to prevent current from leaking where it should not flow, and manufacturers have increasingly turned to three-dimensional transistor structures and new materials to keep progress moving. Whether or not the historical pace of miniaturization continues, the underlying device it depends on remains, in essentially the same functional role Bardeen and Brattain demonstrated on a workbench in 1947: a small, precisely controlled switch that decides whether current flows.

Frequently Asked Questions

Who invented the transistor?

The first working transistor was built in December 1947 by John Bardeen and Walter Brattain at Bell Labs, with William Shockley as the head of the research group. Shockley later developed a separate and more manufacturable design, the junction transistor. All three shared the 1956 Nobel Prize in Physics for the invention.

What is the difference between a transistor and an integrated circuit?

A transistor is a single semiconductor switch or amplifier. An integrated circuit combines many transistors, along with other components, onto a single piece of semiconductor material, allowing far more complex circuits to be built in a much smaller space.

Why did transistors replace vacuum tubes?

Transistors were smaller, used far less power, generated much less heat, and were considerably more reliable than vacuum tubes, which made electronic devices cheaper to build, more compact, and less prone to failure.