Efficient strategy for parallelisation of multilevel fast multipole algorithm using CUDA | Publicación