J. Koster. Parallel solution of sparse systems of linear equations on a mesh network of transputers. Institute for Continuing Education, Eindhoven University of Technology, Eindoven, The Netherlands, July 1993. Final Report. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Parallel Pivots LU Algorithm on the Cray T3E - Asenjo, Zapata (1999) (Correct)
....for shared memory multiprocessors exceeds 50 efficiency in 8 processors with great difficulty. Better results are reached for distributed memory machines as presented by Stappen, Bisseling and van der Vorst (1993) 29] for a square Transputer mesh, and by Koster and Bisseling (1994) [24, 25] also for a transputer mesh. In these codes, they do not present a switch to a dense code stage nor parallel solve stage. 6 Conclusions This work presents a complete tool, SpLU, to solve large nonsymmetric linear systems on distributed memory multiprocessors. SpLU code comprises analysefactorize ....
J. Koster. Parallel solution of sparse systems of linear equations on a mesh network of transputers. Institute for Continuing Education, Eindhoven University of Technology, Eindoven, The Netherlands, July 1993. Final Report.
Analyzing Data Structures for Parallel Sparse.. - Tourino, Doallo.. (1996) (Correct)
.... sparse matrices (see Table 3 for a description) The parallel algorithm was coded in C and PVM routines were used for message passing (the Cray T3D specific low latency PVM functions were used as in the MGS) There are some improvements to the linked list implementation developed by Jacko Koster [17], such as the reduction of the communications cost by an implicit pivoting together with sporadic workload re balancing phases. On the other hand, in the parallel version based on packed vectors, the A matrix is distributed following the BCS scheme and the sparse local matrices are stored in the ....
J. Koster (1993), "Parallel Solution of Sparse Systems of Linear Equations on a Mesh Network of Transputers", Tech. Report, Institute for Continuing Education, Eindhoven Univ. of Technology, The Netherlands.
An Improved Algorithm for Parallel Sparse LU Decomposition.. - Koster, Bisseling (1994) (2 citations) Self-citation (Koster) (Correct)
.... multiplier column j in A[I ; I S ] is divided by the corresponding pivot value A jj and the matrix product A[I ; I S ] Delta A[I S ; I ] is subtracted from A[I ; I ] 3 The new algorithm An outline of the new algorithm is given below; a detailed description and a program text can be found in [7]. Algorithm 1 (Parallel LU decomposition) I(s) J(t) fi : 0 i n OE i = sg; fj : 0 j n j = tg; L; U; k : 0; 0; 0; while k n do begin find pivot set S = f(i r ; j r ) k r k mg; I S ; J S : fi r : k r k mg; fj r : k r k mg; I(s) J(t) I(s) n I S ; J(t) n J S ....
J. Koster, Parallel solution of sparse systems of linear equations on a mesh network of transputers, Final Report, Institute for Continuing Education, Eindhoven University of Technology, Eindhoven, The Netherlands, July 1993.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC