Data Traffic Reduction Schemes for Cholesky Factorization on Asynchronous Multiprocessor Systems
Abstract
Communication requirements of Cholesky factorization of defense and sparse symmetric, positive definite matrices are analyzed. The communication requirement is characterized by the data traffic generated on multiprocessor systems with local and shared memory. Lower bound proofs are given. Partitioning schemes that are variations of block assignment scheme are described and it is shown that the data traffic generated by these schemes are asymptotically optimal. Itis shown that the block based partitioning schemes allow a better utilization of the data accessed from shared memory and thus reduce the data traffic than those based on column-wise wrap around assignment scheme. Keywords: Data processing; Processors.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1989
- Accession Number
- ADA211655
Entities
People
- Merrell L. Patrick
- Vijay K. Naik