Efficient broadcasts and simple algorithms for parallel linear algebra computing in clusters
Material type:
ArticleSeries: ^p Datos electrónicos (1 archivo : 246 KB)Subject(s): Online resources: Summary: This paper presents a natural and efficient implementation for the classical broadcast message passing routine which optimizes performance of Ethernet based clusters. A simple algorithm for parallel matrix multiplication is specifically designed to take advantage of both, parallel computing facilities (CPUs) provided by clusters, and optimized performance of broadcast messages on Ethernet based clusters. Also, this simple parallel algorithm proposed for matrix multiplication takes into account the possibly heterogeneous computing hardware and maintains a balanced workload of computers according to their relative computing power. Performance tests are presented on a heterogeneous cluster as well as on a homogeneous cluster, where it is compared with the parallel matrix multiplication provided by the ScaLAPACK library. Another simple parallel algorithm is proposed for LU matrix factorization (a general method to solve dense systems of equations) following the same guidelines used for the parallel matrix multiplication algorithm. Some performance tests are presented over a homogeneous cluster.
| Item type | Home library | Collection | Call number | URL | Status | Date due | Barcode | |
|---|---|---|---|---|---|---|---|---|
Capítulo de libro
|
Biblioteca de la Facultad de Informática | Biblioteca digital | A0148 (Browse shelf(Opens below)) | Link to resource | No corresponde |
Browsing Biblioteca de la Facultad de Informática shelves, Collection: Biblioteca digital Close shelf browser (Hides shelf browser)
Formato de archivo: PDF. -- Este documento es producción intelectual de la Facultad de Informática-UNLP (Colección BIPA / Biblioteca.) -- Disponible también en línea (Cons. 13/03/2009)
This paper presents a natural and efficient implementation for the classical broadcast message passing routine which optimizes performance of Ethernet based clusters. A simple algorithm for parallel matrix multiplication is specifically designed to take advantage of both, parallel computing facilities (CPUs) provided by clusters, and optimized performance of broadcast messages on Ethernet based clusters. Also, this simple parallel algorithm proposed for matrix multiplication takes into account the possibly heterogeneous computing hardware and maintains a balanced workload of computers according to their relative computing power. Performance tests are presented on a heterogeneous cluster as well as on a homogeneous cluster, where it is compared with the parallel matrix multiplication provided by the ScaLAPACK library. Another simple parallel algorithm is proposed for LU matrix factorization (a general method to solve dense systems of equations) following the same guidelines used for the parallel matrix multiplication algorithm. Some performance tests are presented over a homogeneous cluster.
Parallel and Distributed Processing Symposium, 2003.