| Title:
|
Performance of parallel QR factorization methods on the NVIDIA Grace CPU Superchip (English) |
| Author:
|
Břichňáč, Vít |
| Author:
|
Šístek, Jakub |
| Language:
|
English |
| Journal:
|
Programs and Algorithms of Numerical Mathematics |
| Volume:
|
Proceedings of Seminar. Hejnice, June 23-28, 2024 |
| Issue:
|
2024 |
| Year:
|
|
| Pages:
|
29-40 |
| . |
| Category:
|
math |
| . |
| Summary:
|
This article studies several algorithms for QR factorization based on hierarchical Householder reflectors organized into elimination trees, which are particularly suited for tall-and-skinny matrices and allow parallelization. We examine the effect of various parameters on the performance of the tree-based algorithms. The work is accompanied with a custom implementation that utilizes a task-based runtime system (OpenMP or StarPU). The same algorithm is implemented in the PLASMA library. The performance evaluation is done on the recent NVIDIA Grace CPU Superchip. (English) |
| Keyword:
|
QR factorization |
| Keyword:
|
task-based programming |
| Keyword:
|
NVIDIA Grace CPU |
| MSC:
|
65F05 |
| DOI:
|
10.21136/panm.2024.03 |
| . |
| Date available:
|
2025-06-02T07:38:57Z |
| Last updated:
|
2025-06-09 |
| Stable URL:
|
http://hdl.handle.net/10338.dmlcz/703227 |
| . |
