Improving the performance of parallel triangularization of a sparse matrix using a reconfigurable multicomputer | Publicación