Designing parallel sparse matrix algorithms beyond data dependence analysis

Algorithms are often parallelized based on data dependence analysis manually or by means of parallel compilers. Some vector/matrix computations such as the matrix-vector products with simple data dependence structures (data parallelism) can be easily parallelized. For problems with more complicated...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings International Conference on Parallel Processing Workshops s. 7 - 13
Hlavní autor: Lin, H.X.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 2001
Témata:
Algorithm design and analysis
Concurrent computing
Costs
Data analysis
Information analysis
Information technology
Parallel algorithms
Parallel processing
Partitioning algorithms
Sparse matrices
ISBN:9780769512600, 0769512607
ISSN:1530-2016
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Algorithms are often parallelized based on data dependence analysis manually or by means of parallel compilers. Some vector/matrix computations such as the matrix-vector products with simple data dependence structures (data parallelism) can be easily parallelized. For problems with more complicated data dependence structures, parallelization is less straightforward. The data dependence graph is a powerful means for designing and analyzing parallel algorithm. However for sparse matrix computations, parallelization based on solely exploiting the existing parallelism in an algorithm does not always give satisfactory results. For example, the conventional Gaussian elimination algorithm for the solution of a tri-diagonal system is inherent sequential, so algorithms specially for parallel computation has to be designed. After briefly reviewing different parallelization approaches, a powerful graph formalism for designing parallel algorithms is introduced. This formalism will be discussed using a tri-diagonal system as an example. Its application to general matrix computations is also discussed and its power in designing parallel algorithms beyond the ability of data dependence analysis is shown.
ISBN:9780769512600
0769512607
ISSN:1530-2016
DOI:10.1109/ICPPW.2001.951838