Hubbard model on Semiclassical approximation in combination with an optimizer based on GPU technology (2311.04435v1)

Abstract: We developed a semiclassical approximation method in combination with an adaptive moment estimation optimizer (SCA + ADAM) approach based on the PyTorch plus CUDA library on a the graphics processing unit (GPU). This method was employed to evaluate one-particle properties of the Hubbard model with long-range spatial correlations within an appropriate computing duration. The method was applied to the ionic Hubbard model on a two-dimensional square lattice with long-range spatial correlations. The computation time was evaluated as a function of the lattice size on the central processing unit and GPU. Herein, we also discuss the density of states and antiferromagnetic (AF) order parameter in the Hubbard model without the ionic potential and compare the results with those of the Hartree-Fock approximation. Finally, we present the one-particle properties and order parameter in charge density wave, AF metal and AF insulator of the ionic Hubbard model.