Basics Algorithms Hardware Error Correction Cryptography Machine Learning Information Theory Physics & Simulation Applications Industry & Careers

Quantum Information Theory

Entropy, channel capacity, teleportation, complexity classes and the theorems that define what quantum information can and cannot do.

155 articles

The rules of the game

Quantum information theory asks what can be done with quantum information in principle, independent of any particular machine. It is where the field's hard limits live — the theorems that tell you what no amount of engineering will ever achieve, and the ones that tell you what is possible even though it looks impossible.

Two results set the tone. No-cloning: an unknown quantum state cannot be copied. This is why quantum error correction had to be reinvented from scratch, and why quantum key distribution can detect an eavesdropper. Teleportation: a quantum state can be moved from one place to another, using entanglement plus two classical bits — the state is destroyed at the origin, no information travels faster than light, and it is the basis of how a quantum network would actually work.

What the area covers

Entropy and information measures — von Neumann entropy, relative entropy, mutual information, and the divergences used to compare quantum states. These quantify how much information a state holds and how well two states can be told apart.
Channel capacity — how much information survives a noisy quantum channel. Quantum capacities behave strangely: two channels that each carry nothing can carry information when used together, which has no classical analogue.
Entanglement theory — measuring entanglement, distilling it, and treating it as a resource that can be spent.
State discrimination and tomography — what you can learn about an unknown state from measurements, and at what cost.
Complexity classes — BQP, QMA, QCMA and the rest: the formal statements of what quantum computers can and cannot efficiently solve. Note that BQP is not known to contain NP-complete problems, which is the precise version of "quantum computers do not solve everything fast".

Why it matters beyond the theory

The abstractions here are what keep the rest of the field honest. Resource estimates for algorithms, thresholds for error correction and security proofs for cryptography are all written in this language. When a claim sounds too good, it is usually an information-theoretic bound that rules it out.