English

Tensor-structured algorithm for reduced-order scaling large-scale Kohn-Sham density functional theory calculations

Computational Physics 2021-01-12 v2

Abstract

We present a tensor-structured algorithm for efficient large-scale DFT calculations by constructing a Tucker tensor basis that is adapted to the Kohn-Sham Hamiltonian and localized in real-space. The proposed approach uses an additive separable approximation to the Kohn-Sham Hamiltonian and an L1L_1 localization technique to generate the 1-D localized functions that constitute the Tucker tensor basis. Numerical results show that the resulting Tucker tensor basis exhibits exponential convergence in the ground-state energy with increasing Tucker rank. Further, the proposed tensor-structured algorithm demonstrated sub-quadratic scaling with system size for both systems with and without a gap, and involving many thousands of atoms. This reduced-order scaling has also resulted in the proposed approach outperforming plane-wave DFT implementation for systems beyond 2,000 electrons.

Keywords

Cite

@article{arxiv.2011.11917,
  title  = {Tensor-structured algorithm for reduced-order scaling large-scale Kohn-Sham density functional theory calculations},
  author = {Chih-Chuen Lin and Phani Motamarri and Vikram Gavini},
  journal= {arXiv preprint arXiv:2011.11917},
  year   = {2021}
}
R2 v1 2026-06-23T20:28:07.020Z