Data distributions for sparse matrix vector multiplication

Sparse matrix vector multiplication (SpMxV) is often one of the core components of many scientific applications. Many authors have proposed methods for its data distribution in distributed memory multiprocessors. We can classify these into four groups: Scatter, D-Way Strip, Recursive and Miscellaneo...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Parallel computing Ročník 21; číslo 4; s. 583 - 605
Hlavní autoři: Romero, L.F., Zapata, E.L.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Amsterdam Elsevier B.V 01.04.1995
Elsevier
Témata:
Algorithmics. Computability. Computer arithmetics
Applied sciences
Computer science; control theory; systems
Data distribution
Distributed memory multiprocessor
Exact sciences and technology
Memory and file management (including protection and security)
Memory organisation. Data processing
Mesh topology
Parallel algorithm
Performance evaluation
Software
Software engineering
Sparse computation
Sparse matrix
Theoretical computing
Performance evaluation
Parallel algorithm
Data distribution
Sparse matrix
Mesh topology
Distributed memory multiprocessor
Sparse computation
Multiple recursive decomposition
Decomposition method
Sparse matrix vector multiplication
Matrix product
Distributed system
Topology
Generalization
Multiprocessor
Matrix calculus
Recursive method
ISSN:0167-8191, 1872-7336
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Sparse matrix vector multiplication (SpMxV) is often one of the core components of many scientific applications. Many authors have proposed methods for its data distribution in distributed memory multiprocessors. We can classify these into four groups: Scatter, D-Way Strip, Recursive and Miscellaneous. In this work we propose a new method (Multiple Recursive Decomposition (MRD)), which partitions the data using the prime factors of the dimensions of a multiprocessor network with mesh topology. Furthermore, we introduce a new storage scheme, storage-by-row-of-blocks, that significantly increases the efficiency of the Scatter method. We will name Block Row Scatter (BRS) method this new variant. The MRD and BRS methods achieve results that improve those obtained by other analyzed methods, being their implementation easier. In fact, the data distributions resulting from the MRD and BRS methods are a generalization of the Block and Cyclic distributions used in dense matrices.
ISSN:0167-8191
1872-7336
DOI:10.1016/0167-8191(94)00087-Q