The Transistor, Explained

How did a simple on-off switch become the building block of all modern computers? A primer on the history and function of the transistor.

Tech 101

  • May 31, 2024

  • Download this Tech 101

  • Follow Intel Newsroom on social:

    Twitter logo
    YouTube Icon



Transistors are microscopic switches that make computer chips work.

That’s right, switches.

Modern chips are essentially massive collections of teensy on-off transistors.

You’d be forgiven to suspect something more sophisticated than a switch, but there are good reasons that the transistor is the foundation of the ever-more-powerful computer – and considered one of the most important inventions in history.

Manufacturing 101
Illustration of a transistor under a magnifying glass

The Transistor’s Pre-History

The pioneers of 20th century computing explored a range of devices and designs to perform calculations and store data, but eventually converged on a Goldilocks (just right) combination of the binary system and the vacuum tube. Binary allows information to be encoded as strings of ones and zeros — on or off — and the vacuum tube served as the first reliable electronic switch. That one or zero is known as a binary digit — a bit.

Using bits and binary arithmetic kept the hardware simple and reliable while allowing computer builders and programmers to take on more and more sophisticated problems. (Bits would also prove useful in the future transportation of data, further cementing this Goldilocks solution.)

The Vacuum Tube

When the heater is powered, heat causes the electrons to flow from the anode to the cathode.

Diagram of a vacuum tube connected to a binary system

Glass tube




The first programmable, electronic, general-purpose digital computer was the ENIAC, completed in 1945 and composed of 18,000 vacuum tubes. But tubes use a fair amount of power, get hot and break often. Tubes on the ENIAC failed daily, and the system was down about half the time. Ask any modern data center provider — 50% uptime is untenable, like a car that gets you to your destination only half of the time.

It was clear we could do better.

From Discrete Transistors to Silicon Chips

Years of research led to the transistor, first demonstrated at Bell Labs in 1947, and improved dramatically over the next decade. By the mid-1960s, another Goldilocks combination evolved:

  • The planar manufacturing process, which allowed for efficient construction of circuits on silicon, a natural semiconductor.
  • The metal oxide semiconductor field-effect transistor (MOSFET), built into silicon using controlled oxidation (in other words, by adding materials to specific areas).

Projected light



Silicon wafer

Diagram of the planar manufacturing process. Photolithography is a critical step in planar manufacturing. Patterns exposed on the surface of the wafer are used to repeatedly add, remove, or implant materials to build rows of transistors.

Now you had a device that was very power-efficient and easy to make, without the drawbacks of vacuum tubes.

As Intel co-founder Gordon Moore famously observed in 1965, transistors were also getting exponentially cheaper — transistors were shrinking, and performance was rising. His observation of this repeating cycle of cheaper transistors was later called Moore’s Law. Human beings have since made more transistors than any other manufactured device ever.

How the Modern Transistor Works

Each MOSFET transistor is made of a source and a drain where the current flows in and out (that is, when it’s flipped “on”). The area between the source and drain is called the “channel,” controlled by a “gate” on top. When the gate is off, current can’t flow from the source to the drain.

The gate is turned on by applying a voltage to the gate to attract the appropriate charge in the channel. This completes the connection between source and drain, allowing current to flow.

The electric fields drive the current flow in the transistors — hence “field effect” in the MOSFET name.

We can create transistors with either a negative charge or positive charge — nMOS or pMOS. Together they make complementary, or cMOS, and we use them to create circuits that perform computations, from simple math to modeling the origins of the universe.


Diagram of a nMOS transistor, showing the inputs, outputs, and channel



How Transistors Team Up to Get Work Done

Before you can simulate the Big Bang, however, you need to be able to do things like add and subtract. The first step in exploiting transistors is to combine them in specific ways to create logic gates, which are used to perform basic logical functions. These logic gates are then combined to create more complex circuits and perform more complex operations.

The hierarchy from single transistor to a processor with several billion of them wired together goes like this:

  1. Gates
  2. The three elementary logic gates, the most basic units of logic that transistors can form, include AND, OR and NOT gates. Basic gates can be combined to form more complex gates.

  1. Circuit
  2. By combining gates in various ways, we create complex circuits that can perform arithmetic and logical operations. For example, an adder circuit can be created using a combination of gates using less than 30 transistors.

  1. Processor
  2. A processor is made up of millions or billions of transistors arranged in complex circuits. These circuits include components like arithmetic logic units (ALUs), registers that store data and control units that manage the operation of the processor.

Diagram showing the way multiple transistors can be connected to eachother via circuits, which themselves are connected to form a processor

As transistors continue to become more plentiful and better-performing, we keep finding more clever ways to put them to work. On and off, on and off, a billion times per second. How does Intel make transistors? That’s a topic for another Tech 101.

Want to go deeper on this topic? The Intel Technology channel on YouTube hosts two outstanding videos you’ll enjoy: The evolution of transistor innovation with Marisa Ahmad and a master class on transistor technology with Intel Fellow Paul Packan.