Efficient parallel LU factorization with pivoting on a hypercube multiprocessor
Technical Report
·
OSTI ID:5001510
A message-passing parallel algorithm is developed for forming the LU factors of general non-singular matrices on a hypercube multiprocessor. Partial pivoting is performed to ensure numerical stability, but the scheduling of tasks is such that the pivot search in the parallel algorithm is completely masked. Empirical tests show that the load imbalance produced by random pivoting causes a 5% to 14% increase in execution time. A complementary parallel triangular solution algorithm is also given. Comparisons with the non-pivoting case are used to demonstrate the efficiency of this new algorithm.
- Research Organization:
- Oak Ridge National Lab., TN (USA)
- DOE Contract Number:
- AC05-84OR21400
- OSTI ID:
- 5001510
- Report Number(s):
- ORNL-6211; ON: DE86001461
- Country of Publication:
- United States
- Language:
- English
Similar Records
Matrix factorization on a hypercube multiprocessor
Column LU factorization with pivoting on a hypercube multiprocessor
Parallel Cholesky factorization on a hypercube multiprocessor
Conference
·
Thu Aug 01 00:00:00 EDT 1985
·
OSTI ID:5653196
Column LU factorization with pivoting on a hypercube multiprocessor
Technical Report
·
Thu Oct 31 23:00:00 EST 1985
·
OSTI ID:5001792
Parallel Cholesky factorization on a hypercube multiprocessor
Technical Report
·
Thu Aug 01 00:00:00 EDT 1985
·
OSTI ID:5304680