Combinatorial algorithm design on the cell/B.E. processor

David A. Bader, Virat Agarwal, Kamesh Madduri, Fabrizio Petrini

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Combinatorial algorithms play a subtle, yet important, role in traditional scientific computing. Perhaps the most well-known example is the graph partitioning formulation for load-balanced parallelization of scientific simulations. Partitioning algorithms are typically composed of several combinatorial kernels such as graph coloring, matching, sorting, and permutations. Combinatorial algorithms also appear in auxiliary roles for efficient parallelization of linear algebra, computational physics, and numerical optimization computations. In the last decade or so, the paradigm of data-intensive scientific discovery has significantly altered the landscape of computing. Combinatorial approaches are now at the heart of massive data analysis routines, systems biology, and in general, the study of natural phenomena involving networks and complex interactions.

Original languageEnglish (US)
Title of host publicationScientific Computing with Multicore and Accelerators
PublisherCRC Press
Pages195-216
Number of pages22
ISBN (Electronic)9781439825372
ISBN (Print)9781439825365
DOIs
StatePublished - Jan 1 2010

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Mathematics(all)

Fingerprint Dive into the research topics of 'Combinatorial algorithm design on the cell/B.E. processor'. Together they form a unique fingerprint.

Cite this