Fast Truncated SVD of Sparse and Dense Matrices on Graphics Processors
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913260649840640 |
|---|---|
| author | Tomas, Andres E. Quintana-Orti, Enrique S. Anzt, Hartwig |
| author_facet | Tomas, Andres E. Quintana-Orti, Enrique S. Anzt, Hartwig |
| contents | We investigate the solution of low-rank matrix approximation problems using the truncated SVD. For this purpose, we develop and optimize GPU implementations for the randomized SVD and a blocked variant of the Lanczos approach. Our work takes advantage of the fact that the two methods are composed of very similar linear algebra building blocks, which can be assembled using numerical kernels from existing high-performance linear algebra libraries. Furthermore, the experiments with several sparse matrices arising in representative real-world applications and synthetic dense test matrices reveal a performance advantage of the block Lanczos algorithm when targeting the same approximation accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_06218 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Fast Truncated SVD of Sparse and Dense Matrices on Graphics Processors Tomas, Andres E. Quintana-Orti, Enrique S. Anzt, Hartwig Distributed, Parallel, and Cluster Computing G.4.5 We investigate the solution of low-rank matrix approximation problems using the truncated SVD. For this purpose, we develop and optimize GPU implementations for the randomized SVD and a blocked variant of the Lanczos approach. Our work takes advantage of the fact that the two methods are composed of very similar linear algebra building blocks, which can be assembled using numerical kernels from existing high-performance linear algebra libraries. Furthermore, the experiments with several sparse matrices arising in representative real-world applications and synthetic dense test matrices reveal a performance advantage of the block Lanczos algorithm when targeting the same approximation accuracy. |
| title | Fast Truncated SVD of Sparse and Dense Matrices on Graphics Processors |
| topic | Distributed, Parallel, and Cluster Computing G.4.5 |
| url | https://arxiv.org/abs/2403.06218 |