Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Subspace recursive Fermi-operator expansion strategies for large-scale DFT eigenvalue problems on HPC architectures

Journal Article · · Journal of Chemical Physics
DOI:https://doi.org/10.1063/5.0150287· OSTI ID:2424038

Quantum mechanical calculations for material modeling using Kohn–Sham density functional theory (DFT) involve the solution of a nonlinear eigenvalue problem for N smallest eigenvector-eigenvalue pairs, with N proportional to the number of electrons in the material system. Here, these calculations are computationally demanding and have asymptotic cubic scaling complexity with the number of electrons. Large-scale matrix eigenvalue problems arising from the discretization of the Kohn–Sham DFT equations employing a systematically convergent basis traditionally rely on iterative orthogonal projection methods, which are shown to be computationally efficient and scalable on massively parallel computing architectures. However, as the size of the material system increases, these methods are known to incur dominant computational costs through the Rayleigh–Ritz projection step of the discretized Kohn–Sham Hamiltonian matrix and the subsequent subspace diagonalization of the projected matrix. This work explores the potential of polynomial expansion approaches based on recursive Fermi-operator expansion as an alternative to the subspace diagonalization of the projected Hamiltonian matrix to reduce the computational cost. Subsequently, we perform a detailed comparison of various recursive polynomial expansion approaches to the traditional approach of explicit diagonalization on both multi-node central processing unit and graphics processing unit architectures and assess their relative performance in terms of accuracy, computational efficiency, scaling behavior, and energy efficiency.

Research Organization:
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Oak Ridge Leadership Computing Facility (OLCF)
Sponsoring Organization:
USDOE Office of Science (SC)
Grant/Contract Number:
AC05-00OR22725
OSTI ID:
2424038
Journal Information:
Journal of Chemical Physics, Journal Name: Journal of Chemical Physics Journal Issue: 3 Vol. 159; ISSN 1089-7690; ISSN 0021-9606
Publisher:
American Institute of Physics (AIP)Copyright Statement
Country of Publication:
United States
Language:
English

References (38)

Fast Spectral Projection Algorithms for Density-Matrix Computations journal June 1999
The iterative calculation of a few of the lowest eigenvalues and corresponding eigenvectors of large real-symmetric matrices journal January 1975
Application of complex energy integration to selfconsistent electronic structure calculations journal November 1982
SPARC: Accurate and efficient finite-difference formulation and parallel implementation of Density Functional Theory: Isolated clusters journal March 2017
DFT-FE – A massively parallel adaptive finite-element code for large-scale density functional theory calculations journal January 2020
GPU-acceleration of the ELPA2 distributed eigensolver for dense symmetric and hermitian eigenproblems journal May 2021
DFT-FE 1.0: A massively parallel hybrid CPU-GPU density functional theory code using finite-element discretization journal November 2022
Optimized purification for density matrix calculation journal January 2013
Self-consistent-field calculations using Chebyshev-filtered subspace iteration journal November 2006
Higher-order adaptive finite-element methods for Kohn–Sham density functional theory journal November 2013
Transforms for idempotency purification of density matrices in linear-scaling electronic-structure calculations journal June 2001
Density matrices in O(N) electronic structure calculations: theory and applications journal August 1999
Mixed Precision Fermi-Operator Expansion on Tensor Cores from a Machine Learning Perspective journal April 2021
Quantum-Based Molecular Dynamics Simulations Using Tensor Cores journal October 2021
Quantum Perturbation Theory Using Tensor Cores and a Deep Neural Network journal June 2022
Large Scale Quantum Chemistry with Tensor Processing Units journal December 2022
Linear Scaling Pseudo Fermi-Operator Expansion for Fractional Occupation journal November 2018
Nonmonotonic Recursive Polynomial Expansions for Linear Scaling Calculation of the Density Matrix journal April 2011
Computing the Density Matrix in Electronic Structure Theory on Graphics Processing Units journal October 2012
Linear scaling density matrix search based on sign matrices journal October 2000
Comparison of two genres for linear scaling in density functional theory: Purification and density matrix minimization methods journal February 2005
A note on the Pulay force at finite electronic temperatures journal December 2008
Communication: Generalized canonical purification for density matrix minimization journal March 2016
Flexibilities of wavelets as a computational basis set for large-scale electronic structure calculations journal May 2020
Assessment of density matrix methods for linear scaling electronic structure calculations journal February 2011
The ELPA library: scalable parallel eigenvalue solutions for electronic structure theory and computational science journal May 2014
Self-Consistent Equations Including Exchange and Correlation Effects journal November 1965
Canonical purification of the density matrix in electronic-structure theory journal November 1998
Expansion algorithm for the density matrix journal October 2002
Implicit purification for temperature-dependent density matrices journal December 2003
Continued fraction representation of the Fermi-Dirac function for large-scale electronic structure calculations journal January 2007
Subquadratic-scaling subspace projection method for large-scale Kohn-Sham density functional theory calculations using spectral finite-element discretization journal September 2014
Efficient Linear Scaling Algorithm for Tight-Binding Molecular Dynamics journal July 1994
Some Recent Advances in Density Matrix Theory journal April 1960
Linear scaling electronic structure methods journal July 1999
The kernel polynomial method journal March 2006
Interior Eigenvalues from Density Matrix Expansions in Quantum Mechanical Molecular Dynamics journal January 2014
Toward the Optimal Preconditioned Eigensolver: Locally Optimal Block Preconditioned Conjugate Gradient Method journal January 2001

Similar Records

Accelerating Atomic Orbital-based Electronic Structure Calculation via Pole Expansion plus Selected Inversion
Journal Article · Thu Feb 09 23:00:00 EST 2012 · Journal of Computational Physics · OSTI ID:1076802

Chebyshev-filtered subspace iteration method free of sparse diagonalization for solving the Kohn–Sham equation
Journal Article · Wed Oct 01 00:00:00 EDT 2014 · Journal of Computational Physics · OSTI ID:22382122

TTDFT: A GPU accelerated Tucker tensor DFT code for large-scale Kohn-Sham DFT calculations
Journal Article · Mon Sep 05 00:00:00 EDT 2022 · Computer Physics Communications · OSTI ID:2424043