Tensor-Structured Techniques for Large-Scale Electronic Structure Calculations
Abstract
This report summarizes the research objectives achieved in this project during the period 02-01-2017 to 07-31-2020. Computational techniques based on low-rank tensor decomposition and tensor representation have been developed that enable large-scale, reduced-order scaling electronic structure calculations based on Kohn-Sham density functional theory. In particular, the various components of the developed techniques include (i) develop an additive separable approximation to the Kohn-Sham Hamiltonian and use this to generate a reduced-order, and yet a systematically convergent tensor-structured basis adapted to the Kohn-Sham Hamiltonian to represent the electronic structure; (ii) develop localization techniques to transform the basis into a localized basis while preserving the tensor structure and accuracy of representation of the electronic structure; (iii)develop efficient algorithms to project the Kohn-Sham problem into the localized tensor-structured basis and solve the resulting discrete Kohn-sham eigenvalue problem.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 17, 2021
- Accession Number
- AD1146059
Entities
People
- Vikram Gavini
Organizations
- Board of Regents of the University of Michigan