Science › Quantum world
Quantum computing
1981· Discovery / experiment
Well supportedStrong evidence or math, but not fully tested.
Computers that use superposition and entanglement to tackle certain problems far faster than any ordinary computer could.
See it on the Science map →How it works
Ordinary bits are 0 or 1. Quantum bits, or qubits, can be blends of both, and entangled qubits represent vast numbers of combinations together. Clever algorithms make wrong answers cancel and right ones reinforce, like waves. This helps only on specific problems: simulating molecules, factoring large numbers, some searches. Qubits are extremely fragile, so errors are the main obstacle. Richard Feynman proposed the idea in 1981.
Why it matters
It could transform chemistry and materials design. A large one could break much of today's internet encryption, which is why governments are moving to new 'post-quantum' codes.
How we know
Working machines from Google, IBM, Quantinuum and others have from dozens to over a thousand qubits. Google showed error correction improving with size in 2024. Clear advantage on commercially useful problems has not yet been shown.
Still a mystery
Whether large, error-corrected machines can be built economically, and how many practical problems truly benefit, are open. Timelines are widely disputed.
Connections (5)
▸Superposition: many states at once is the resource of Quantum computingWell supported
The relationship
A qubit can be a blend of 0 and 1. Algorithms manipulate these blends so that paths leading to wrong answers cancel and paths to right answers reinforce, like interfering waves. This interference across many possibilities, not simply 'trying every answer at once', is the source of a quantum computer's potential speedup.
How the link was found
Richard Feynman (1981) and David Deutsch (1985) proposed quantum computers; Peter Shor's 1994 factoring algorithm showed a dramatic potential use.
Is it proven?
Small quantum processors run algorithms as theory predicts. Large speedups are proven mathematically for some problems but not yet demonstrated on useful real-world tasks.
Open questions and mysteries
Exactly which problems gain big speedups, and whether superpositions can be kept stable at large scale, remain open.
▸Quantum entanglement powers Quantum computingWell supported
The relationship
Entangled qubits cannot be described one at a time; their joint state spans combinations whose number doubles with each added qubit. Algorithms exploit these correlations for problems that would overwhelm ordinary computers. Entanglement is also needed for quantum error correction, which spreads one qubit's information across many.
How the link was found
Feynman (1981) and Deutsch (1985) conceived quantum computing; Peter Shor (1995) and Andrew Steane (1996) showed entanglement-based error correction was possible.
Is it proven?
Computations with too little entanglement can be efficiently simulated on ordinary computers, so for standard quantum circuits it is proven necessary for speedups. Labs routinely entangle dozens to hundreds of qubits.
Open questions and mysteries
Entanglement is necessary but not sufficient; exactly what gives quantum computers their power is still being studied.
▸Richard Feynman proposed Quantum computingWell supported
The relationship
Feynman pointed out that simulating quantum systems on ordinary computers becomes impossibly slow as the system grows, because the possibilities multiply exponentially. His proposed fix: build a computer that itself runs on quantum rules, so it could mimic nature naturally. That idea is the root of quantum computing.
How the link was found
Feynman's 1981 talk at an MIT conference, published in 1982 as 'Simulating Physics with Computers'. Yuri Manin and Paul Benioff had related ideas around 1980.
Is it proven?
Quantum computers now exist in labs. Google reported in 2019 a task beyond ordinary supercomputers and in 2024 showed improving error correction. Practical, useful machines are not yet here.
Open questions and mysteries
Whether large, error-free quantum computers can be built economically is still open.
▸Information theory (Shannon) has a quantum version in Quantum computingWell supported
The relationship
Shannon's bit is either 0 or 1. Its quantum version, the qubit, can be in a superposition of both and entangled with other qubits. Quantum information theory extends Shannon's ideas to qubits: how much they carry, how to correct their errors, and how to transmit them. Quantum computers aim to exploit this to solve certain problems far faster.
How the link was found
Richard Feynman proposed quantum computers in 1981. Peter Shor's 1994 factoring algorithm and 1995 error-correcting code showed real promise; Benjamin Schumacher coined 'qubit' around 1995.
Is it proven?
Quantum error correction and quantum communication have been demonstrated. Large, useful quantum computers are not yet built, so practical advantage is still largely unproven.
Open questions and mysteries
Whether large machines can be made reliable enough, and which practical problems will truly benefit.
▸Quantum computing shows physics can be computed, fueling The simulation hypothesisPhilosophy
The relationship
Quantum computers exist because nature can be imitated by computation: a machine made of quantum parts can model other quantum systems. If every physical process can in principle be simulated, then a whole universe might be too. Simulation fans take this as a reason the idea is not absurd, though it says nothing about whether ours is simulated.
How the link was found
Richard Feynman proposed quantum computers to simulate physics in 1981. David Deutsch argued in 1985 that a universal quantum computer could simulate any physical process.
Is it proven?
Quantum computers have simulated small molecules and materials, and Google's 2019 machine beat classical computers on one narrow task. That supports "physics is computable" in small cases, not that we live in a simulation.
Open questions and mysteries
Simulating a universe in full detail may require a computer at least as large as the universe itself. Whether shortcuts exist is unknown.