What is quantum computing?

Bits and qubits

Every app on your phone runs on bits. A bit is a tiny switch that is either off (0) or on (1), and everything a normal computer does, from playing video to checking your spelling, is a long sequence of operations on bits.

A quantum computer uses qubits (quantum bits) instead. When you measure a qubit you still get 0 or 1. Before you measure it, though, a qubit can be in a superposition: a combination of 0 and 1, described by two numbers called amplitudes that set how likely each result is. What a qubit is has the details.

The three ingredients

Quantum computers get their power from three effects working together.

• Superposition. A group of qubits can be in a combination of many possible bit patterns at once. Ten qubits are described by 1,024 amplitudes, and 300 qubits by more amplitudes than there are atoms in the observable universe.
• Entanglement. Qubits can be linked so that the group behaves as one system, with results that are correlated in ways no ordinary system can copy. Entanglement is what lets a quantum computer build up that huge state space.
• Interference. Amplitudes behave like waves: they can add up or cancel out. A quantum algorithm is a carefully designed sequence of operations, called quantum gates, that makes the amplitudes of wrong answers cancel and the amplitude of the right answer grow.

Interference is the part most explanations leave out. A quantum computer does not simply "try every answer at once" and hand you the best one. When you measure, you get a single result. The skill is in arranging things so that the single result is very likely to be the one you want.

What quantum computers are good at

Quantum computers are not faster versions of your laptop. They are specialised machines that should beat ordinary computers on particular kinds of problems:

• Simulating molecules and materials. Nature is quantum, so a quantum computer is a natural tool for modelling chemical reactions, new drugs, batteries and catalysts. This is the idea Richard Feynman proposed in 1981.
• Factoring large numbers. Shor's algorithm could factor numbers far faster than any known classical method, which is why quantum computers threaten some of today's encryption.
• Searching unsorted data. Grover's algorithm finds an item among N possibilities in about √N steps instead of N, a smaller but very general speedup.
• Some optimisation problems. Scheduling, logistics and finance are active research areas, but clear advantages have not been shown yet.

For email, games, spreadsheets and almost everything else, ordinary computers will stay faster, cheaper and simpler.

Why they are not everywhere yet

Qubits are extremely fragile. Heat, vibration and stray electromagnetic fields disturb them, an effect called decoherence, and every operation adds a small chance of error. Many machines have to be cooled to a fraction of a degree above absolute zero.

Today's quantum computers have at most a few thousand physical qubits, and they are still too noisy to run long algorithms reliably. The way forward is quantum error correction, which spreads each reliable "logical" qubit across many physical ones.

A very short history

• 1981: Richard Feynman suggests that simulating quantum physics needs a computer that is itself quantum.
• 1985: David Deutsch describes a universal quantum computer.
• 1994: Peter Shor publishes his factoring algorithm, and the field takes off.
• 1996: Lov Grover publishes his search algorithm.
• 2016: IBM puts a small quantum computer on the cloud for anyone to use.
• 2019: Google reports a calculation its Sycamore chip did far faster than its estimate for a supercomputer, on a test problem with no practical use.
• 2024: Google's Willow chip shows error correction getting better as it is made bigger, a key step towards reliable machines.

Where to start

You do not need a physics degree to understand the core ideas. Start with what a qubit is, then superposition and entanglement. Once those make sense, gates, circuits and algorithms follow naturally.