Ultrafast dynamics of finite Hubbard clusters - a stochastic mean-field approach

Finite lattice models are a prototype for strongly correlated quantum systems and capture essential properties of condensed matter systems. With the dramatic progress in ultracold atoms in optical lattices, finite fermionic Hubbard systems have become directly accessible in experiments, including their ultrafast dynamics far from equilibrium. Here, we present a theoretical approach that is able to treat these dynamics in any dimension and fully includes inhomogeneity effects. The method consists in stochastic sampling of mean-field trajectories and is found to be more accurate and efficient than current nonequilibrium Green functions approaches. This is demonstrated for Hubbard clusters with up to 512 particles in one, two and three dimensions.