메뉴 건너뛰기




Volumn 43, Issue 11, 1991, Pages 8861-8869

Efficient pseudopotentials for plane-wave calculations. II. Operators for fast iterative diagonalization

Author keywords

[No Author keywords available]

Indexed keywords


EID: 33646658147     PISSN: 01631829     EISSN: None     Source Type: Journal    
DOI: 10.1103/PhysRevB.43.8861     Document Type: Article
Times cited : (689)

References (36)
  • 5
    • 84927894234 scopus 로고    scopus 로고
    • Here we indicate by O(N) a quantity that is proportional to N with a coefficient of proportionality of the order of unity. This coefficient would depend on the specific details of the calculation and architecture of the computer. For example, a complex matrix requires twice the storage of a real matrix; a symmetric matrix may be stored in ``packed'' form with a 50% memory storage savings or stored in the ``traditional'' form for efficient vectorization.
  • 18
    • 84927894233 scopus 로고    scopus 로고
    • Following tradition, we will use the same symbol for a function and its Fourier transform. We also use the same symbol for an operator and its representation. Arguments or subscripts and context usually suffice to distinguish between them.
  • 20
    • 84927894232 scopus 로고    scopus 로고
    • Full orthogonalization of the eigensolutions will still require O ( M2N ) operations. Although M2N << M N2, this number still scales with the cube of the number of atoms, and therefore efficient iterative diagonalization algorithms should minimize the number of orthogonalization steps.


* 이 정보는 Elsevier사의 SCOPUS DB에서 KISTI가 분석하여 추출한 것입니다.