WEKO3
アイテム
Hybrid Numerical Solvers for Massively Parallel Eigenvalue Computations and Their Benchmark with Electronic Structure Calculations
https://ipsj.ixsq.nii.ac.jp/records/146084
https://ipsj.ixsq.nii.ac.jp/records/1460846c6ebb5d-1b1f-4e82-a661-cc0e8aedf559
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
IPSJ-TACS0804011.pdf (4.4 MB)
|
Copyright (c) 2015 by the Information Processing Society of Japan
|
|
| オープンアクセス | ||
| Item type | Trans(1) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 公開日 | 2015-11-20 | |||||||||
| タイトル | ||||||||||
| タイトル | Hybrid Numerical Solvers for Massively Parallel Eigenvalue Computations and Their Benchmark with Electronic Structure Calculations | |||||||||
| タイトル | ||||||||||
| 言語 | en | |||||||||
| タイトル | Hybrid Numerical Solvers for Massively Parallel Eigenvalue Computations and Their Benchmark with Electronic Structure Calculations | |||||||||
| 言語 | ||||||||||
| 言語 | eng | |||||||||
| キーワード | ||||||||||
| 主題Scheme | Other | |||||||||
| 主題 | [数値アルゴリズム] massively parallel numerical library, generalized eigenvalue problem, electronic structure calculation, ELPA, EigenExa, the K computer, mini-application | |||||||||
| 資源タイプ | ||||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||||||
| 資源タイプ | journal article | |||||||||
| 著者所属 | ||||||||||
| Tottori University, JST-CREST | ||||||||||
| 著者所属 | ||||||||||
| Tottori University, JST-CREST | ||||||||||
| 著者所属(英) | ||||||||||
| en | ||||||||||
| Tottori University, JST-CREST | ||||||||||
| 著者所属(英) | ||||||||||
| en | ||||||||||
| Tottori University, JST-CREST | ||||||||||
| 著者名 |
Hiroto, Imachi
× Hiroto, Imachi
× Takeo, Hoshi
|
|||||||||
| 著者名(英) |
Hiroto, Imachi
× Hiroto, Imachi
× Takeo, Hoshi
|
|||||||||
| 論文抄録 | ||||||||||
| 内容記述タイプ | Other | |||||||||
| 内容記述 | Optimally hybrid numerical solvers were constructed for massively parallel generalized eigenvalue problems (GEP). The strong scaling benchmark was carried out on the K computer and other supercomputers for electronic structure calculation problems in the matrix sizes of M=104-106 with up to 105 cores. The procedure of GEP is decomposed into the two subprocedures of the reducer to the standard eigenvalue problem (SEP) and the solver of SEP. A hybrid solver is constructed, when a routine is chosen for each subprocedure from the three parallel solver libraries of ScaLAPACK, ELPA and EigenExa. The hybrid solvers with the two newer libraries, ELPA and EigenExa, give better benchmark results than the conventional ScaLAPACK library. The detailed analysis on the results implies that the reducer can be a bottleneck in next-generation (exa-scale) supercomputers, which provides guidance for future research. The code was developed as a middleware and a mini-application and will appear online. \n------------------------------ This is a preprint of an article intended for publication Journal of Information Processing(JIP). This preprint should not be cited. This article should be cited as: Journal of Information Processing Vol.24(2016) No.1(online) ------------------------------ |
|||||||||
| 論文抄録(英) | ||||||||||
| 内容記述タイプ | Other | |||||||||
| 内容記述 | Optimally hybrid numerical solvers were constructed for massively parallel generalized eigenvalue problems (GEP). The strong scaling benchmark was carried out on the K computer and other supercomputers for electronic structure calculation problems in the matrix sizes of M=104-106 with up to 105 cores. The procedure of GEP is decomposed into the two subprocedures of the reducer to the standard eigenvalue problem (SEP) and the solver of SEP. A hybrid solver is constructed, when a routine is chosen for each subprocedure from the three parallel solver libraries of ScaLAPACK, ELPA and EigenExa. The hybrid solvers with the two newer libraries, ELPA and EigenExa, give better benchmark results than the conventional ScaLAPACK library. The detailed analysis on the results implies that the reducer can be a bottleneck in next-generation (exa-scale) supercomputers, which provides guidance for future research. The code was developed as a middleware and a mini-application and will appear online. \n------------------------------ This is a preprint of an article intended for publication Journal of Information Processing(JIP). This preprint should not be cited. This article should be cited as: Journal of Information Processing Vol.24(2016) No.1(online) ------------------------------ |
|||||||||
| 書誌レコードID | ||||||||||
| 収録物識別子タイプ | NCID | |||||||||
| 収録物識別子 | AA11833852 | |||||||||
| 書誌情報 |
情報処理学会論文誌コンピューティングシステム（ACS） 巻 8, 号 4, 発行日 2015-11-20 |
|||||||||
| ISSN | ||||||||||
| 収録物識別子タイプ | ISSN | |||||||||
| 収録物識別子 | 1882-7829 | |||||||||
| 出版者 | ||||||||||
| 言語 | ja | |||||||||
| 出版者 | 情報処理学会 | |||||||||
