Linear algebra algorithms in a heterogeneous cluster of personal computers
Cluster computing is presently a major research area,mostly for high performance computing. The work hereinpresented refers to the application of cluster computing in asmall scale where a virtual machine is composed by a smallnumber of off-the-shelf personal computers connected by alow cost network....
| Autor principal: | |
|---|---|
| Outros Autores: | , |
| Formato: | book |
| Idioma: | eng |
| Publicado em: |
2000
|
| Texto completo: | https://repositorio-aberto.up.pt/handle/10216/423 |
| País: | Portugal |
| Oai: | oai:repositorio-aberto.up.pt:10216/423 |
| Resumo: | Cluster computing is presently a major research area,mostly for high performance computing. The work hereinpresented refers to the application of cluster computing in asmall scale where a virtual machine is composed by a smallnumber of off-the-shelf personal computers connected by alow cost network. A methodology to determine the optimalnumber of processors to be used in a computation ispresented as well as the speedup results obtained for thematrix-matrix multiplication and for the symmetric QR algorithm for eigenvector computation which are significantbuilding blocks for applications in the target image processing and analysis domain. The load balancing strategy is also addressed. |
|---|