The paper describes DME, a distributed matrix environment for creating and manipulating sparse matrices on distributed-memory multiprocessors. The goals of this environment are: to define a distributed sparse matrix representation scheme that can serve as a platform for building a range of important matrix operations; to provide a clean, consistent way for parallel applications to build distributed matrices; and to provide a tool-kit of utility routines for common distributed matrix operations. The paper describes the design philosophy behind DME, the DME distributed sparse matrix representation scheme, the mechanisms for creating and modifying these matrices, and the set of matrix utilities DME provides.<<ETX>>
[1]
Jack J. Dongarra,et al.
A set of level 3 basic linear algebra subprograms
,
1990,
TOMS.
[2]
Horst D. Simon,et al.
Partitioning of unstructured problems for parallel processing
,
1991
.
[3]
John G. Lewis,et al.
Sparse matrix test problems
,
1982,
SGNM.
[4]
JAMES DEMMEL,et al.
LAPACK: A portable linear algebra library for high-performance computers
,
1990,
Proceedings SUPERCOMPUTING '90.
[5]
Jack Dongarra,et al.
ScaLAPACK: a scalable linear algebra library for distributed memory concurrent computers
,
1992,
[Proceedings 1992] The Fourth Symposium on the Frontiers of Massively Parallel Computation.