Two-dimensional block partitionings for the parallel sparse Cholesky factorization
From MaRDI portal
The planar mappings of matrix data in sparse Cholesky factorization for distributed-memory computer systems of the MIMD-type are discussed. Several mapping strategies are analyzed and tested, e.g. fan-in, fan-out, grid mapping with row balancing, proportional and subforest-to-subcube mapping. New heuristics enable to improve existing block column mapping algorithms. The algorithms developed are tested and compared on IBM SP1 and Crat T3D systems.
Recommendations
Cited in
(7)- The impact of high-performance computing in the solution of linear systems: Trends and problems
- Analysis of a sparse hypermatrix Cholesky with fixed-sized blocking
- Block Sparse Cholesky Algorithms on Advanced Uniprocessor Computers
- scientific article; zbMATH DE number 554769 (Why is no real title available?)
- An Efficient Block-Oriented Approach to Parallel Sparse Cholesky Factorization
- scientific article; zbMATH DE number 720649 (Why is no real title available?)
- scientific article; zbMATH DE number 833696 (Why is no real title available?)
This page was built for publication: Two-dimensional block partitionings for the parallel sparse Cholesky factorization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1385158)