How Do Transistors Work?

The transistor is arguably the most important invention of the twentieth century. Without it, there would be no computers, smartphones, or the internet. The device that fits in your pocket contains more transistors than there were vacuum tubes in all the buildings on Earth in the 1950s, and each transistor works a million times faster while using a million times less energy. This exponential shrinking and improvement has transformed every aspect of modern life, from how we communicate to how we work to how we spend our leisure time.

The short answer

A transistor is a semiconductor device that can either amplify an electronic signal or switch it on and off. It has three terminals, and applying a small voltage or current to one terminal controls the current flowing between the other two. Invented at Bell Labs in 1947, transistors replaced bulky and unreliable vacuum tubes and enabled the digital revolution. The two main types are BJTs (bipolar junction transistors) and MOSFETs (metal-oxide-semiconductor field-effect transistors), with MOSFETs now dominating chip manufacturing because they are smaller and more efficient. Modern computer chips contain billions of transistors, and their exponential shrinking has driven the growth of computing power for six decades.

The full picture

The invention that changed everything

The transistor was invented at Bell Telephone Laboratories in New Jersey on December 16, 1947, by John Bardeen, Walter Brattain, and William Shockley, all of whom received the Nobel Prize in Physics in 1956 for their work. Their first transistor was a simple device called a point-contact transistor, which used a strip of gold foil pressed against a germanium crystal with a plastic wedge holding it in place. When they applied a small voltage, the current flowing through the semiconductor increased dramatically.

This invention solved a critical problem. Vacuum tubes, the dominant electronic switches of the era, were large, fragile, expensive, and unreliable. They generated enormous heat and burned out frequently. The transistor performed the same basic function as a vacuum tube (amplifying or switching electronic signals) but was tiny by comparison, ran cool, and lasted for decades. Immediately, designers began replacing vacuum tubes with transistors, creating the first portable radios and the first portable communication equipment.

How transistors actually work

The fundamental principle relies on semiconductors, materials that conduct electricity better than insulators but worse than metals. Pure silicon is a semiconductor, but its electrical properties can be modified by doping it with small amounts of other elements. Adding phosphorus or arsenic adds extra electrons, creating N-type silicon (negative). Adding boron creates P-type silicon (positive), which has electron holes that behave like positive charges.

When you place P-type material next to N-type material, electrons from the N region flood into the P region and recombine with the holes, creating a depletion zone with no free charges. This zone acts as an insulating barrier. But applying voltage in the right direction can narrow or widen the barrier, controlling whether current can flow or not.

The BJT (bipolar junction transistor) stacks two layers of one type (say, N-type) with a thin layer of the other type (P-type) sandwiched between. A small current from the base to the emitter controls a much larger current from the collector to the emitter. Think of it like a water valve: a small force on the handle controls a large water flow.

The MOSFET (metal-oxide-semiconductor field-effect transistor) uses voltage applied to the gate terminal to create an electronic field that opens or closes a channel between the source and drain. The gate is insulated from the channel by a thin layer of silicon dioxide, which is why it uses almost no current to control and is extremely power-efficient. Most computer chips today use MOSFETs for this reason.

The march of Moore’s Law

In 1965, Gordon Moore, co-founder of Intel, observed that the number of components (transistors) on an integrated circuit chip was doubling roughly every two years. This observation, now known as Moore’s Law, has held for over sixty years and has driven the entire technology industry’s expectations and planning.

The first commercially available integrated circuit chip, from 1958, contained four transistors, as noted by Wikipedia. Today’s largest GPU chips contain over one hundred billion. This factor of twenty-five billion in sixty years represents an exponential improvement that has no parallel in any other industry. Each generation of chips has smaller transistors, more of them, lower power consumption, and faster switching speeds.

The current state-of-the-art manufacturing uses three nanometer process technology, where the smallest features are only three billionths of a meter across. At this scale, quantum effects like electron tunneling become significant problems. The industry is actively researching new materials and structures to continue shrinking, though some experts believe we are approaching fundamental physical limits.

What this means in real life

You are reading this article on a device that contains perhaps ten billion transistors (a smartphone) or hundreds of billions (a modern laptop or desktop computer). Each transistor switches on or off billions of times per second, processing the data that displays images, runs applications, and connects you to the internet. Every digital interaction in your life, from unlocking your phone to loading a webpage to playing a video, relies on transistors doing their job flawlessly billions of times per second.

The continuing improvement in transistor density has made devices dramatically cheaper over time. A smartphone with more computing power than a 1990s supercomputer costs a few hundred dollars. This democratization of technology has transformed education, entertainment, commerce, and communication on a global scale.

Transistors are also increasingly embedded in the physical world. Modern cars contain dozens of microprocessors and perhaps a billion transistors. The Internet of Things extends this further, with transistors embedded in sensors, appliances, and infrastructure, creating a pervasive computational fabric that is still expanding.

Why it matters

The transistor is foundational to modern civilization. Every computation, every digital communication, every piece of technology you interact with depends on transistors working flawlessly. Understanding what they are and how they work helps you appreciate the exponential progress encoded in Moore’s Law, and why future limits on transistor shrinking matter for your device’s capabilities and costs.

The practical reasons this matters are straightforward. Transistor density determines what your devices can do. A smartphone with ten billion transistors can run artificial intelligence applications that would have required a supercomputer twenty years ago. The continuing improvement in transistor technology is why computers get faster, phones get more capable, and technology gets cheaper even as it grows more powerful.

For personal technology decisions, understanding transistor scaling helps. New phones with the newest chip process tend to last longer and perform better. Embedded systems in cars and appliances rely on simpler microcontroller chips that are designed for reliability over performance. The differences matter when choosing products that will be used for years.

For the future, the limits of transistor shrinking may shift industry focus toward specialized processors, new materials, and different computing paradigms. This affects everything from your next phone’s price to which skills are most valuable in the workforce.

Common misconceptions

Transistors are only in computers. While computers contain the most transistors by far, they are in virtually every electronic device. Televisions, radios, phones, cars, washing machines, digital watches, and even simple LED lights all contain transistors. The microcontroller in a modern appliance might have thousands of transistors performing dedicated functions.

Moore’s Law is a physical law. Moore’s Law is an observation and an industry target, not a physical law. Whether it continues depends on economic and engineering factors, not fundamental physics. The industry has maintained it for sixty years through enormous investment, but the pace has slowed recently, and many experts believe it will end within the next decade.

Smaller transistors are always better. Smaller usually means faster and more efficient, but at extremely small scales, quantum effects create new problems like electron leakage and tunneling. The engineering challenge shifts from simply shrinking to finding new materials and structures. There are also fundamental limits to how small a transistor can practically get, set by the size of atoms and the energy needed to control electrons.

Transistors will stop improving. While the pace of improvement has slowed and may end for traditional silicon MOSFETs, the industry is actively researching alternatives like carbon nanotubes, graphene, and quantum computing structures. The end of Moore’s Law for classical computers does not mean the end of progress in computational capability.

Key terms

Semiconductor: A material, most commonly silicon, whose electrical conductivity lies between conductors and insulators. Its conductivity can be precisely controlled by doping with other elements, making it ideal for electronic switches.

BJT (bipolar junction transistor): A transistor type where current flow between emitter and collector is controlled by current at the base. Uses more power than MOSFETs but handles high current better, used in analog and power applications.

MOSFET (metal-oxide-semiconductor field-effect transistor): The dominant transistor type in modern chips. Uses voltage at the gate to control current between source and drain. Extremely power-efficient because the gate is insulated and draws almost no current.

Integrated circuit: A complete electronic circuit containing multiple transistors (today, billions) fabricated on a single chip of semiconductor material. The foundation of all modern electronics.

Moore’s Law: The observation, first made by Gordon Moore in 1965, that the number of transistors on a chip doubles roughly every two years. Has held for over sixty years and driven the technology industry’s planning.

Process node: The manufacturing technology node indicating the smallest feature size on a chip. Current leading-edge is around three nanometers, with one nanometer (next-generation) nodes in development. Smaller is generally better for density and efficiency.

Doping: Adding small amounts of other elements to silicon to make it N-type (extra electrons) or P-type (electron holes), enabling precise control of electrical properties.

Depletion zone: The region at the junction between N-type and P-type semiconductor where electrons and holes have recombined, creating a zone with no free charges that acts as an insulator.

CMOS: Complementary metal-oxide-semiconductor, a logic family using pairs of PMOS and NMOS transistors. Forms the basis of most digital logic because it uses essentially no power when not switching.