Optimization of sparse matrix–vector multiplication using reordering techniques on GPUs | Publicación