Entanglement-informed distributed wavefunction approach to scalable quantum many-body systems

Abstract

We show that the entanglement structure of quantum many-body states defines a natural and optimal distributed representation for their simulation. An arbitrary entanglement cut induces a bipartite decomposition of the wavefunction, mapping its distribution onto that of the entanglement spectrum. In this representation the Hamiltonian application, the core of Krylov-subspace methods, reduces to local contractions and communication-optimal operations. Using benchmarks from different methods and models, we demonstrate near-linear scaling for sufficiently large systems and identify entanglement spectrum fragmentation as a key factor controlling computational cost. This establishes entanglement as an organizing principle and unified, method-independent, route for scaling up quantum many-body simulations.