Parallel Geometric Multigrid

Multigrid methods are among the fastest numerical algorithms for the solution of large sparse systems of linear equations. While these algorithms exhibit asymptotically optimal computational complexity, their efficient parallelisation is hampered by the poor computation-to-communication ratio on the coarse grids. Our contribution discusses parallelisation techniques for geometric multigrid methods. It covers both theoretical approaches as well as practical implementation issues that may guide code development.

@incollection{id444,
  author = {H\"ulsemann, F. and Kowarschik, M. and Mohr, M. and R\"ude, U.},
  booktitle = {Numerical Solution of Partial Differential Equations on Parallel Computers},
  doi = {10.1007/3-540-31619-1\_5},
  editor = {Bruaset, A. M. and Tveito, A.},
  isbn = {3-540-29076-1},
  language = {en},
  number = {51},
  pages = {165{\textendash}208},
  publisher = {Springer},
  series = {Lecture Notes in Computational Science and Engineering},
  title = {Parallel Geometric Multigrid},
  year = {2005},
}

%O Book Section
%A Hülsemann, F.
%A Kowarschik, M.
%A Mohr, M.
%A Rüde, U.
%B Numerical Solution of Partial Differential Equations on Parallel Computers
%R 10.1007/3-540-31619-1_5
%E Bruaset, A. M.
%E Tveito, A.
%@ 3-540-29076-1
%G en
%N 51
%P 165–208
%I Springer
%S Lecture Notes in Computational Science and Engineering
%T Parallel Geometric Multigrid
%D 2005