Matrix Product Operator

A matrix product operator is a way of writing a many‑body quantum operator as a series of small building blocks linked together in a chain. Each block is a low‑dimensional tensor that carries one physical index, representing the action on a single site, and two auxiliary indices that connect it to its neighbours. By threading these tensors together, the whole operator — whether it be a Hamiltonian, a density matrix, or any other linear map acting on a lattice of spins or particles — is encoded without ever having to list an astronomically large number of individual matrix elements.

The appeal of this representation lies in its efficiency. Because each tensor only needs to capture the most important correlations across a limited range, the total amount of data grows linearly with the length of the system rather than exponentially. This compactness makes it possible to perform accurate numerical simulations on systems that would otherwise be out of reach, and it also provides insight into how entanglement spreads through a chain.

Matrix product operators appear whenever one studies one‑dimensional quantum many‑body problems, especially in techniques such as the density matrix renormalization group or time‑evolving block decimation. They are used to encode static Hamiltonians, to apply time‑evolution gates, and to represent mixed states of open systems, enabling researchers to explore ground‑state properties, dynamical response, and thermal behaviour of long quantum chains.