Quantum computing and optimisation

What it is

Quantum computers use quantum bits, or qubits, which can be prepared in combinations of states and correlated through entanglement. Quantum algorithms manipulate probability amplitudes so certain calculations can explore and combine many possible states in ways unavailable to classical bits. This is sometimes called quantum parallelism, but it does not mean testing every answer at once: the algorithm must amplify useful outcomes before measurement. Quantum simulation, factorisation, sampling and some optimisation or linear-algebra problems may be particularly well suited. Gate-based machines run quantum circuits. Quantum annealers encode an optimisation problem as an energy landscape and search for a low-energy state representing a good solution.

What it is not

Quantum computing is not expected to replace mainstream processors for most workloads. It is more likely to operate as a specialist co-processor or accelerator called by conventional software. Classical CPUs and GPUs will still prepare data, divide the problem, orchestrate the quantum hardware and validate the result. A faster quantum stage is therefore not automatically a useful overall advantage if queueing, data preparation, error correction or post-processing removes the gain.