A General Sparse Tensor Framework for Electronic Structure Theory
Linearscaling algorithms must be developed in order to extend the domain of applicability of electronic structure theory to molecules of any desired size. But, the increasing complexity of modern linearscaling methods makes code development and maintenance a significant challenge. A major contributor to this difficulty is the lack of robust software abstractions for handling blocksparse tensor operations. We therefore report the development of a highly efficient symbolic blocksparse tensor library in order to provide access to highlevel software constructs to treat such problems. Our implementation supports arbitrary multidimensional sparsity in all input and output tensors. We then avoid cumbersome machinegenerated code by implementing all functionality as a highlevel symbolic C++ language library and demonstrate that our implementation attains very high performance for linearscaling sparse tensor contractions.
 Authors:

^{[1]};
^{[2]};
^{[3]};
^{[1]}
 Univ. of California, Berkeley, CA (United States). Kenneth S. Pitzer Center for Theoretical Chemistry; Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). Chemical Sciences Division
 QChem, Inc., Pleasanton, CA (United States)
 Univ. of Southern California, Los Angeles, CA (United States). Dept. of Chemistry
 Publication Date:
 Grant/Contract Number:
 AC0205CH11231
 Type:
 Accepted Manuscript
 Journal Name:
 Journal of Chemical Theory and Computation
 Additional Journal Information:
 Journal Volume: 13; Journal Issue: 3; Journal ID: ISSN 15499618
 Publisher:
 American Chemical Society
 Research Org:
 Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
 Sponsoring Org:
 USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC22); USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC21)
 Country of Publication:
 United States
 Language:
 English
 Subject:
 37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY
 OSTI Identifier:
 1379778
Manzer, Samuel, Epifanovsky, Evgeny, Krylov, Anna I., and HeadGordon, Martin. A General Sparse Tensor Framework for Electronic Structure Theory. United States: N. p.,
Web. doi:10.1021/acs.jctc.6b00853.
Manzer, Samuel, Epifanovsky, Evgeny, Krylov, Anna I., & HeadGordon, Martin. A General Sparse Tensor Framework for Electronic Structure Theory. United States. doi:10.1021/acs.jctc.6b00853.
Manzer, Samuel, Epifanovsky, Evgeny, Krylov, Anna I., and HeadGordon, Martin. 2017.
"A General Sparse Tensor Framework for Electronic Structure Theory". United States.
doi:10.1021/acs.jctc.6b00853. https://www.osti.gov/servlets/purl/1379778.
@article{osti_1379778,
title = {A General Sparse Tensor Framework for Electronic Structure Theory},
author = {Manzer, Samuel and Epifanovsky, Evgeny and Krylov, Anna I. and HeadGordon, Martin},
abstractNote = {Linearscaling algorithms must be developed in order to extend the domain of applicability of electronic structure theory to molecules of any desired size. But, the increasing complexity of modern linearscaling methods makes code development and maintenance a significant challenge. A major contributor to this difficulty is the lack of robust software abstractions for handling blocksparse tensor operations. We therefore report the development of a highly efficient symbolic blocksparse tensor library in order to provide access to highlevel software constructs to treat such problems. Our implementation supports arbitrary multidimensional sparsity in all input and output tensors. We then avoid cumbersome machinegenerated code by implementing all functionality as a highlevel symbolic C++ language library and demonstrate that our implementation attains very high performance for linearscaling sparse tensor contractions.},
doi = {10.1021/acs.jctc.6b00853},
journal = {Journal of Chemical Theory and Computation},
number = 3,
volume = 13,
place = {United States},
year = {2017},
month = {1}
}