1. Introduction: Why MPI for Tensor Contractions? In MPS (Matrix Product State) algorithms like DMRG and CheMPS, the computational bottleneck is tensor contraction — multiplying tensors with 3-5 ...