2012.62: Performance analysis of asynchronous Jacobi's method implemented in MPI, SHMEM and OpenMP
2012.62: Iain Bethune, J. Mark Bull, Nicholas J. Dingle and Nicholas J. Higham (2012) Performance analysis of asynchronous Jacobi's method implemented in MPI, SHMEM and OpenMP.
Full text available as:
|PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader|
Ever-increasing core counts create the need to develop parallel algorithms that avoid closely- coupled execution across all cores. In this paper we present performance analysis of several parallel asynchronous implementations of Jacobi's method for solving systems of linear equations, using MPI, SHMEM and OpenMP. In particular we have solved systems of over 4 billion unknowns using up to 32,768 processes on a Cray XE6 supercomputer. We show that the precise implementation details of asynchronous algorithms can strongly affect the resulting performance and convergence behaviour of our solvers in unexpected ways.
|Item Type:||MIMS Preprint|
|Subjects:||MSC 2000 > 15 Linear and multilinear algebra; matrix theory|
MSC 2000 > 65 Numerical analysis
MSC 2000 > 68 Computer science
|Deposited By:||Dr Nicholas Dingle|
|Deposited On:||15 June 2012|