Proc. Int. Conf. on High Performance Computing (HiPC), New Delhi 1995, pp. 734-739.
Ömer Egecioglu and Ashok Srinivasan
Givens and Householder Reductions for Linear Least Squares on a Cluster of
Workstations
Abstract.
We report on the properties of implementations of fast-Givens rotation
and Householder reflector based parallel algorithms for the solution of linear
least squares problems on a cluster of workstations. It is shown that
the Givens rotations enable communication hiding and take greater
advantage of parallelism than Householder
reflectors, provided the matrices are sufficiently large.
omer@cs.ucsb.edu