arrow
Volume 8, Issue 1
Towards Textbook Efficiency for Parallel Multigrid

Björn Gmeiner, Ulrich Rüde, Holger Stengel, Christian Waluga & Barbara Wohlmuth

Numer. Math. Theor. Meth. Appl., 8 (2015), pp. 22-46.

Published online: 2015-08

Export citation
  • Abstract

In this work, we extend Achi Brandt's notion of textbook multigrid efficiency (TME) to massively parallel algorithms. Using a finite element based geometric multigrid implementation, we recall the classical view on TME with experiments for scalar linear equations with constant and varying coefficients as well as linear systems with saddle-point structure. To extend the idea of TME to the parallel setting, we give a new characterization of a work unit (WU) in an architecture-aware fashion by taking into account performance modeling techniques. We illustrate our newly introduced parallel TME measure by large-scale computations, solving problems with up to 200 billion unknowns on a TOP-10 supercomputer.

  • Keywords

  • AMS Subject Headings

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address
  • BibTex
  • RIS
  • TXT
@Article{NMTMA-8-22, author = {}, title = {Towards Textbook Efficiency for Parallel Multigrid}, journal = {Numerical Mathematics: Theory, Methods and Applications}, year = {2015}, volume = {8}, number = {1}, pages = {22--46}, abstract = {

In this work, we extend Achi Brandt's notion of textbook multigrid efficiency (TME) to massively parallel algorithms. Using a finite element based geometric multigrid implementation, we recall the classical view on TME with experiments for scalar linear equations with constant and varying coefficients as well as linear systems with saddle-point structure. To extend the idea of TME to the parallel setting, we give a new characterization of a work unit (WU) in an architecture-aware fashion by taking into account performance modeling techniques. We illustrate our newly introduced parallel TME measure by large-scale computations, solving problems with up to 200 billion unknowns on a TOP-10 supercomputer.

}, issn = {2079-7338}, doi = {https://doi.org/10.4208/nmtma.2015.w10si}, url = {http://global-sci.org/intro/article_detail/nmtma/12398.html} }
TY - JOUR T1 - Towards Textbook Efficiency for Parallel Multigrid JO - Numerical Mathematics: Theory, Methods and Applications VL - 1 SP - 22 EP - 46 PY - 2015 DA - 2015/08 SN - 8 DO - http://doi.org/10.4208/nmtma.2015.w10si UR - https://global-sci.org/intro/article_detail/nmtma/12398.html KW - AB -

In this work, we extend Achi Brandt's notion of textbook multigrid efficiency (TME) to massively parallel algorithms. Using a finite element based geometric multigrid implementation, we recall the classical view on TME with experiments for scalar linear equations with constant and varying coefficients as well as linear systems with saddle-point structure. To extend the idea of TME to the parallel setting, we give a new characterization of a work unit (WU) in an architecture-aware fashion by taking into account performance modeling techniques. We illustrate our newly introduced parallel TME measure by large-scale computations, solving problems with up to 200 billion unknowns on a TOP-10 supercomputer.

Björn Gmeiner, Ulrich Rüde, Holger Stengel, Christian Waluga & Barbara Wohlmuth. (2020). Towards Textbook Efficiency for Parallel Multigrid. Numerical Mathematics: Theory, Methods and Applications. 8 (1). 22-46. doi:10.4208/nmtma.2015.w10si
Copy to clipboard
The citation has been copied to your clipboard