A. V. Gerbessiotis. Topics in parallel and distributed computing. PhD Thesis (Technical Report TR-03-93), Aiken Computation Laboratory, Harvard University, 1993. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Algorithmic and Practical Considerations for Dense Matrix .. - Alexandros Gerbessiotis Self-citation (Gerbessiotis Computing) (Correct)
....In this section we design and analyze on the BSP model two parallel dense LU decomposition algorithms. The two variants differ in the way data are distributed among the processors. They also complement the BSP algorithm for Gauss Jordan elimination presented in [10] and Cholesky factorization [7, 8]. One algorithm assumes a scattered distribution of the data among the processors to even out processor utilization [6] Let p = p i Delta p j processors are divided into p j groups each group P j consisting of p i processors. Particularly, P j = f(j Gamma 1)p i l j 0 l p i Gamma 1g. ....
....LU decomposition without pivoting, which requires 2n 3 =3 O(n 2 ) arithmetic operations [12] 3. 1 Scattered distribution A BSP algorithm for LU decomposition, based on a scattered distribution of the input matrix is similar to the one presented in [10] for Gauss Jordan elimination and in [7, 8] for dense Cholesky factorization. In each phase a column k of the input matrix is decomposed, and the corresponding elements of L and U are stored in the memory locations previously occupied by the elements of column k of A. Proposition 3 There is a BSP algorithm for the LU decomposition of an ....
A. V. Gerbessiotis. Topics in parallel and distributed computing. PhD Thesis (Technical Report TR-03-93), Aiken Computation Laboratory, Harvard University, 1993.
Algorithmic and Practical Considerations for Dense Matrix.. - Gerbessiotis (1997) (1 citation) Self-citation (Gerbessiotis Computing) (Correct)
....In this section we design and analyze on the BSP model two parallel dense LU decomposition algorithms. The two variants differ in the way data are distributed among the processors. They also complement the BSP algorithm for Gauss Jordan elimination presented in [11] and Cholesky factorization [7, 8]. One algorithm assumes a scattered distribution of the data among the processors to even out processor utilization [6] Let p = p i Delta p j processors are divided into p j groups each group P j consisting of p i processors. Particularly, P j = f(j Gamma 1)p i l j 0 l p i Gamma 1g. ....
....LU decomposition without pivoting, which requires 2n 3 =3 O(n 2 ) arithmetic operations [13] 3. 1 Scattered distribution A BSP algorithm for LU decomposition, based on a scattered distribution of the input matrix is similar to the one presented in [11] for Gauss Jordan elimination and in [7, 8] for dense Cholesky factorization. In each phase a column k of the input matrix is decomposed, and the corresponding elements of L and U are stored in the memory locations previously occupied by the elements of column k of A. Proposition 3 There is a BSP algorithm for the LU decomposition of an n ....
A. V. Gerbessiotis. Topics in parallel and distributed computing. PhD Thesis (Technical Report TR-03-93), Aiken Computation Laboratory, Harvard University, 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