skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver

Abstract

We describe a number of recently developed techniques for improving the performance of large-scalenuclear configuration interaction calculations on high performance parallel computers. We show thebenefit of using a preconditioned block iterative method to replace the Lanczos algorithm that hastraditionally been used to perform this type of computation. The rapid convergence of the block iterativemethod is achieved by a proper choice of starting guesses of the eigenvectors and the construction of aneffective preconditioner. These acceleration techniques take advantage of special structure of the nuclearconfiguration interaction problem which we discuss in detail. The use of a block method also allows usto improve the concurrency of the computation, and take advantage of the memory hierarchy of modernmicroprocessors to increase the arithmetic intensity of the computation relative to data movement. Wealso discuss the implementation details that are critical to achieving high performance on massivelyparallel multi-core supercomputers, and demonstrate that the new block iterative solver is two to threetimes faster than the Lanczos based algorithm for problems of moderate sizes on a Cray XC30 system.

Authors:
ORCiD logo [1];  [2];  [1];  [1];  [3];  [3]
  1. Lawrence Berkeley National Laboratory. (LBNL), Berkeley, CA (United States). Computational Research Division
  2. Michigan State University, East Lansing, MI (United States). Department of Computer Science and Engineering
  3. Iowa State University, Ames, IA (United States). Department of Physics and Astronomy
Publication Date:
Research Org.:
Lawrence Berkeley National Laboratory, Berkeley, CA (United States). National Energy Research Scientific Computing Center (NERSC).
Sponsoring Org.:
USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)
OSTI Identifier:
1463273
DOE Contract Number:  
DESC0008485; FG02-87ER40371; AC02-05CH11231
Resource Type:
Journal Article
Journal Name:
Computer Physics Communications
Additional Journal Information:
Journal Volume: 222; Journal Issue: C; Journal ID: ISSN 0010-4655
Country of Publication:
United States
Language:
English

Citation Formats

Shao, Meiyue, Aktulga, H.  Metin, Yang, Chao, Ng, Esmond G., Maris, Pieter, and Vary, James P. Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver. United States: N. p., 2018. Web. doi:10.1016/j.cpc.2017.09.004.
Shao, Meiyue, Aktulga, H.  Metin, Yang, Chao, Ng, Esmond G., Maris, Pieter, & Vary, James P. Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver. United States. doi:10.1016/j.cpc.2017.09.004.
Shao, Meiyue, Aktulga, H.  Metin, Yang, Chao, Ng, Esmond G., Maris, Pieter, and Vary, James P. Mon . "Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver". United States. doi:10.1016/j.cpc.2017.09.004.
@article{osti_1463273,
title = {Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver},
author = {Shao, Meiyue and Aktulga, H.  Metin and Yang, Chao and Ng, Esmond G. and Maris, Pieter and Vary, James P.},
abstractNote = {We describe a number of recently developed techniques for improving the performance of large-scalenuclear configuration interaction calculations on high performance parallel computers. We show thebenefit of using a preconditioned block iterative method to replace the Lanczos algorithm that hastraditionally been used to perform this type of computation. The rapid convergence of the block iterativemethod is achieved by a proper choice of starting guesses of the eigenvectors and the construction of aneffective preconditioner. These acceleration techniques take advantage of special structure of the nuclearconfiguration interaction problem which we discuss in detail. The use of a block method also allows usto improve the concurrency of the computation, and take advantage of the memory hierarchy of modernmicroprocessors to increase the arithmetic intensity of the computation relative to data movement. Wealso discuss the implementation details that are critical to achieving high performance on massivelyparallel multi-core supercomputers, and demonstrate that the new block iterative solver is two to threetimes faster than the Lanczos based algorithm for problems of moderate sizes on a Cray XC30 system.},
doi = {10.1016/j.cpc.2017.09.004},
journal = {Computer Physics Communications},
issn = {0010-4655},
number = C,
volume = 222,
place = {United States},
year = {2018},
month = {1}
}