We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.chem-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Chemical Physics

Title: Structure Optimization with Stochastic Density Functional Theory

Abstract: Linear-scaling techniques for Kohn-Sham density functional theory (KS-DFT) are essential to describe the ground state properties of extended systems. Still, these techniques often rely on the locality of the density matrix or on accurate embedding approaches, limiting their applicability. In contrast, stochastic density functional theory (sDFT) achieves linear- and sub-linear-scaling by statistically sampling the ground state density without relying on embedding or imposing localization. In return, ground state observables, such as the forces on the nuclei, fluctuate in sDFT, making the optimization of the nuclear structure a highly non-trivial problem. In this work, we combine the most recent noise-reduction schemes for sDFT with stochastic optimization algorithms to perform structure optimization within sDFT. We compare the performance of the stochastic gradient descent (sGD) approach and its variations (stochastic gradient descent with momentum (sGDM)) to stochastic optimization techniques that rely on the Hessian, such as the stochastic Broyden-Fletcher-Goldfarb-Shanno (sBFGS) algorithm. We further provide a detailed assessment of the computational efficiency and its dependence on the optimization parameters for each methods for determining the ground state structure of bulk silicon with varying supercell dimensions.
Subjects: Chemical Physics (physics.chem-ph); Materials Science (cond-mat.mtrl-sci); Computational Physics (physics.comp-ph)
Cite as: arXiv:2209.11403 [physics.chem-ph]
  (or arXiv:2209.11403v1 [physics.chem-ph] for this version)

Submission history

From: Ming Chen [view email]
[v1] Fri, 23 Sep 2022 04:23:37 GMT (718kb,D)

Link back to: arXiv, form interface, contact.