2020
DOI: 10.1016/j.cplett.2019.136983
|View full text |Cite
|
Sign up to set email alerts
|

On preconditioning the self-consistent field iteration in real-space Density Functional Theory

Abstract: We present a real-space formulation for isotropic Fourier-space preconditioners used to accelerate the self-consistent field iteration in Density Functional Theory calculations. Specifically, after approximating the preconditioner in Fourier space using a rational function, we express its realspace application in terms of the solution of sparse Helmholtz-type systems. Using the truncated-Kerker and Resta preconditioners as representative examples, we show that the proposed realspace method is both accurate and… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
24
0

Year Published

2020
2020
2023
2023

Publication Types

Select...
6

Relationship

2
4

Authors

Journals

citations
Cited by 25 publications
(25 citation statements)
references
References 42 publications
0
24
0
Order By: Relevance
“…This is motivated from the observation that the calculation of spectral quadrature weights and nodes -the computationally dominant part of the method -is local to each process, and therefore can be accelerated on Graphics Processing Units (GPUs). The implementation of real-space preconditioning schemes [31,52] can further reduce the number of Self Consistent Field iterations for DFT simulations of large metallic systems. Finally, the incorporation of coarse-graining strategies [42,43] will further reduce the scaling to sublinear with the number of atoms.…”
Section: Discussionmentioning
confidence: 99%
See 2 more Smart Citations
“…This is motivated from the observation that the calculation of spectral quadrature weights and nodes -the computationally dominant part of the method -is local to each process, and therefore can be accelerated on Graphics Processing Units (GPUs). The implementation of real-space preconditioning schemes [31,52] can further reduce the number of Self Consistent Field iterations for DFT simulations of large metallic systems. Finally, the incorporation of coarse-graining strategies [42,43] will further reduce the scaling to sublinear with the number of atoms.…”
Section: Discussionmentioning
confidence: 99%
“…Further, the number of SCF iterations to achieve a fixed target SCF error increases with system size in metallic systems due to charge sloshing [26]. The introduction of real space preconditioning schemes [52,31] is likely to reduce this for large metallic systems.…”
Section: Convergence and Performancementioning
confidence: 99%
See 1 more Smart Citation
“…89 The SCF method is accelerated using the restarted variant 90 of Periodic Pulay mixing 91 with real-space preconditioning. 92 For spin polarized calculations, mixing is performed simultaneously on both components.…”
Section: Real-space Formulation and Implementation A M-sparc Code Basementioning
confidence: 99%
“…For the very first electronic ground state calculation, the superposition of isolated atom electron densities is used as initial guess for the electron density, whereas for every subsequent such calculation, extrapolation based on previous solutions is used [31]. The convergence of the SCF iteration is accelerated using the restarted variant of the Periodic Pulay mixing scheme [32,33], with the option of real-space preconditioning [34]. For spin-polarized calculations, mixing is performed simultaneously for both spin components, i.e., using a vector of twice the original length containing both spin-up and…”
Section: Software Architecturementioning
confidence: 99%