Accelerating R with high performance linear algebra libraries
Linear algebra routines are basic building blocks for the statistical software. In this paper we analyzed how can we improve R performance for matrix computations. We benchmarked few matrix operations using the standard linear algebra libraries included in the R distribution and high performance lib...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Romanian National Institute of Statistics
2015-09-01
|
| Series: | Revista Română de Statistică |
| Subjects: | |
| Online Access: | http://www.revistadestatistica.ro/wp-content/uploads/2015/07/RRS3_2015_A10.pdf |
| _version_ | 1831571609654132736 |
|---|---|
| author | Bogdan Oancea Tudorel Andrei Raluca Mariana Dragoescu |
| author_facet | Bogdan Oancea Tudorel Andrei Raluca Mariana Dragoescu |
| author_sort | Bogdan Oancea |
| collection | DOAJ |
| description | Linear algebra routines are basic building blocks for the statistical software. In this paper we analyzed how can we improve R performance for matrix computations. We benchmarked few matrix operations using the standard linear algebra libraries included in the R distribution and high performance libraries like OpenBLAS, GotoBLAS and MKL. Our tests showed the best results are obtained with the MKL library, the other two libraries having similar performances, but lower than MKL. |
| first_indexed | 2024-12-17T13:11:51Z |
| format | Article |
| id | doaj.art-da4e0434c2f94dd58d7c06d41f037847 |
| institution | Directory Open Access Journal |
| issn | 1018-046X 1844-7694 |
| language | English |
| last_indexed | 2024-12-17T13:11:51Z |
| publishDate | 2015-09-01 |
| publisher | Romanian National Institute of Statistics |
| record_format | Article |
| series | Revista Română de Statistică |
| spelling | doaj.art-da4e0434c2f94dd58d7c06d41f0378472022-12-21T21:47:05ZengRomanian National Institute of StatisticsRevista Română de Statistică1018-046X1844-76942015-09-01633109117Accelerating R with high performance linear algebra librariesBogdan Oancea0Tudorel Andrei1Raluca Mariana Dragoescu2“Nicolae Titulescu” University of BucharestNational Statistics Institute of RomaniaThe Bucharest University of Economic StudiesLinear algebra routines are basic building blocks for the statistical software. In this paper we analyzed how can we improve R performance for matrix computations. We benchmarked few matrix operations using the standard linear algebra libraries included in the R distribution and high performance libraries like OpenBLAS, GotoBLAS and MKL. Our tests showed the best results are obtained with the MKL library, the other two libraries having similar performances, but lower than MKL.http://www.revistadestatistica.ro/wp-content/uploads/2015/07/RRS3_2015_A10.pdfRlinear algebraBLAShigh performance computing |
| spellingShingle | Bogdan Oancea Tudorel Andrei Raluca Mariana Dragoescu Accelerating R with high performance linear algebra libraries Revista Română de Statistică R linear algebra BLAS high performance computing |
| title | Accelerating R with high performance linear algebra libraries |
| title_full | Accelerating R with high performance linear algebra libraries |
| title_fullStr | Accelerating R with high performance linear algebra libraries |
| title_full_unstemmed | Accelerating R with high performance linear algebra libraries |
| title_short | Accelerating R with high performance linear algebra libraries |
| title_sort | accelerating r with high performance linear algebra libraries |
| topic | R linear algebra BLAS high performance computing |
| url | http://www.revistadestatistica.ro/wp-content/uploads/2015/07/RRS3_2015_A10.pdf |
| work_keys_str_mv | AT bogdanoancea acceleratingrwithhighperformancelinearalgebralibraries AT tudorelandrei acceleratingrwithhighperformancelinearalgebralibraries AT ralucamarianadragoescu acceleratingrwithhighperformancelinearalgebralibraries |