메뉴 건너뛰기




Volumn 139, Issue 13, 2013, Pages

General implementation of the resolution-of-the-identity and Cholesky representations of electron repulsion integrals within coupled-cluster and equation-of-motion methods: Theory and benchmarks

Author keywords

[No Author keywords available]

Indexed keywords

CLUSTER ANALYSIS; COMPUTATION THEORY; ITERATIVE METHODS; NUMERICAL METHODS;

EID: 84903362826     PISSN: 00219606     EISSN: None     Source Type: Journal    
DOI: 10.1063/1.4820484     Document Type: Article
Times cited : (141)

References (72)
  • 3
    • 78649551287 scopus 로고    scopus 로고
    • 10.1080/00268976.2010.531773
    • R. J. Bartlett, Mol. Phys. 108, 2905 (2010). 10.1080/00268976.2010.531773
    • (2010) Mol. Phys. , vol.108 , pp. 2905
    • Bartlett, R.J.1
  • 4
    • 37549014315 scopus 로고    scopus 로고
    • 10.1146/annurev.physchem.59.032607.093602
    • A. I. Krylov, Annu. Rev. Phys. Chem. 59, 433 (2008). 10.1146/annurev. physchem.59.032607.093602
    • (2008) Annu. Rev. Phys. Chem. , vol.59 , pp. 433
    • Krylov, A.I.1
  • 8
    • 36849099978 scopus 로고
    • 10.1063/1.1679012
    • J. L. Whitten, J. Chem. Phys. 58, 4496 (1973). 10.1063/1.1679012
    • (1973) J. Chem. Phys. , vol.58 , pp. 4496
    • Whitten, J.L.1
  • 19
    • 85032356625 scopus 로고    scopus 로고
    • Cholesky decomposition techniques in electronic structure theory
    • in, edited by R. Zaleśny, M. G. Papadopoulos, P. G. Mezey, and J. Leszczynski, Challenges and Advances in Computational Chemistry and Physics (Springer)
    • F. Aquilante, L. Boman, J. Boström, H. Koch, R. Lindh, A. S. de Merás, and T. B. Pedersen, " Cholesky decomposition techniques in electronic structure theory.," in Linear-Scaling Techniques in Computational Chemistry and Physics, edited by, R. Zaleśny, M. G. Papadopoulos, P. G. Mezey, and, J. Leszczynski, Challenges and Advances in Computational Chemistry and Physics (Springer, 2011), pp. 301-343.
    • (2011) Linear-Scaling Techniques in Computational Chemistry and Physics , pp. 301-343
    • Aquilante, F.1    Boman, L.2    Boström, J.3    Koch, H.4    Lindh, R.5    De Merás, A.S.6    Pedersen, T.B.7
  • 30
    • 84903366723 scopus 로고    scopus 로고
    • Turbomole User's Manual (version 6.4), see; accessed 05/25/2013
    • Turbomole User's Manual (version 6.4), 2012, see http://www.turbomole- gmbh.com; accessed 05/25/2013.
    • (2012)
  • 32
    • 0025387691 scopus 로고
    • 10.1016/0010-4655(90)90136-O
    • S. Wilson, Comput. Phys. Commun. 58, 71 (1990). 10.1016/0010-4655(90) 90136-O
    • (1990) Comput. Phys. Commun. , vol.58 , pp. 71
    • Wilson, S.1
  • 37
    • 37649016929 scopus 로고    scopus 로고
    • 10.1002/jcc.20702
    • F. Weigend, J. Comput. Chem. 29, 167 (2008). 10.1002/jcc.20702
    • (2008) J. Comput. Chem. , vol.29 , pp. 167
    • Weigend, F.1
  • 42
    • 84903366724 scopus 로고    scopus 로고
    • Throughout the paper, we adhere to the convention that ijkl denote occupied orbitals, abcd denote virtual orbitals, and pqrs denote orbitals that can be either occupied or virtual
    • Throughout the paper, we adhere to the convention that ijkl denote occupied orbitals, abcd denote virtual orbitals, and pqrs denote orbitals that can be either occupied or virtual.
  • 43
    • 84903366725 scopus 로고    scopus 로고
    • bcd:38V4&esu
    • bcd:38V4&esu.
  • 44
    • 85099494444 scopus 로고    scopus 로고
    • 2 vectors are stored for the DIIS algorithm
    • 2 vectors are stored for the DIIS algorithm.
  • 46
    • 84903366727 scopus 로고    scopus 로고
    • The three-tensor contraction is defined as T, where i, j, k, a, b, c designate collectively respective inner and outer indices in tensors A, B, C. To compute such a contraction, first an intermediate needs to be formed by multiplying two of the arguments, followed by a contraction with the third argument to form the result. If the size of the intermediate exceeds memory limits, it is desirable to proceed in batches by forming the intermediate by parts and contracting such that each batch is completed in core memory. This approach allows us to minimize I/O by automatically tuning the batch size to the memory limit and retain the total computational cost of the contraction
    • The three-tensor contraction is defined as T, where i, j, k, a, b, c designate collectively respective inner and outer indices in tensors A, B, C. To compute such a contraction, first an intermediate needs to be formed by multiplying two of the arguments, followed by a contraction with the third argument to form the result. If the size of the intermediate exceeds memory limits, it is desirable to proceed in batches by forming the intermediate by parts and contracting such that each batch is completed in core memory. This approach allows us to minimize I/O by automatically tuning the batch size to the memory limit and retain the total computational cost of the contraction.
  • 47
    • 85099494510 scopus 로고    scopus 로고
    • 2M
    • 2M.
  • 49
    • 33644973569 scopus 로고    scopus 로고
    • 10.1021/ar0402006
    • A. I. Krylov, Acc. Chem. Res. 39, 83 (2006). 10.1021/ar0402006
    • (2006) Acc. Chem. Res. , vol.39 , pp. 83
    • Krylov, A.I.1
  • 55
    • 84903366729 scopus 로고    scopus 로고
    • During the final revision stage, a small algorithmic improvement has been implemented that resulted in ∼5% speed-uin CD/RI CC/EOM calculations. Thus, the reported timings are roughly 5% slower than the current code
    • During the final revision stage, a small algorithmic improvement has been implemented that resulted in ∼5% speed-up in CD/RI CC/EOM calculations. Thus, the reported timings are roughly 5% slower than the current code.
  • 59
    • 84903366730 scopus 로고    scopus 로고
    • Nwchem: High-Performance Computational Chemistry Software, see (last accessed March 13)
    • Nwchem: High-Performance Computational Chemistry Software, see http://www.nwchem-sw.org/index.php/Benchmarks (last accessed March 13, 2013).
    • (2013)
  • 61
    • 84903366731 scopus 로고    scopus 로고
    • Q-Chem's keywords controlling the CC and EOM convergence: CC-T-CONV = 4, CC-E-CONV = 6, EOM-DAVIDSON-CONV = 5, EOM-DAVIDSON-THRESH = 5
    • Q-Chem's keywords controlling the CC and EOM convergence: CC-T-CONV = 4, CC-E-CONV = 6, EOM-DAVIDSON-CONV = 5, EOM-DAVIDSON-THRESH = 5.
  • 63
    • 84903366722 scopus 로고    scopus 로고
    • See supplementary material at E-JCPSA6-139-029335 for additional details on timings, Cartesian geometries, and relevant energies
    • See supplementary material at http://dx.doi.org/10.1063/1.4820484 E-JCPSA6-139-029335 for additional details on timings, Cartesian geometries, and relevant energies.
  • 67
    • 33746614482 scopus 로고
    • 10.1063/1.456153
    • T. H. Dunning, J. Chem. Phys. 90, 1007 (1989). 10.1063/1.456153
    • (1989) J. Chem. Phys. , vol.90 , pp. 1007
    • Dunning, T.H.1


* 이 정보는 Elsevier사의 SCOPUS DB에서 KISTI가 분석하여 추출한 것입니다.