GPUaccelerated Auxiliaryfield quantum Monte Carlo with multiSlater determinant trial states
Abstract
The accuracy of phaseless auxiliaryfield quantum Monte Carlo (phAFQMC) can be systematically improved with better trial states. Using multiSlater determinant trial states, phAFQMC has the potential to faithfully treat strongly correlated systems, while balancing the static and dynamical correlations on an equal footing. This preprint presents an implementation and application of graphics processing unitaccelerated phAFQMC, for multiSlater determinant trial wavefunctions (GPUaccelerated MSDAFQMC), to enable efficient simulation of largescale, strongly correlated systems. This approach allows for nearlyexact computation of ground state energies in multireference systems. Our GPUaccelerated MSDAFQMC is implemented in the opensource code \texttt{ipie}, a Pythonbased AFQMC package [\textit{J. Chem. Theory Comput.}, 2022, 19(1): 109121]. We benchmark the performance of the GPU code on transitionmetal clusters like [Cu$_2$O$_2$]$^{2+}$ and [Fe$_2$S$_2$(SCH$_3$)]$^{2}$. The GPU code achieves at least sixfold speedup in both cases, comparing the timings of a single A100 GPU to that of a 32CPU node. For [Fe$_2$S$_2$(SCH$_3$)]$^{2}$, we demonstrate that our GPU MSDAFQMC can recover the dynamical correlation necessary for chemical accuracy with an MSD trial, despite the large number of determinants required ($>10^5$). Our work significantly enhances the efficiency of MSDAFQMC calculations for large, strongly correlated molecules by utilizing GPUs, offering a promising path for exploring the electronic structure of transition metal complexes.
 Publication:

arXiv eprints
 Pub Date:
 June 2024
 DOI:
 10.48550/arXiv.2406.08314
 arXiv:
 arXiv:2406.08314
 Bibcode:
 2024arXiv240608314H
 Keywords:

 Physics  Chemical Physics;
 Condensed Matter  Strongly Correlated Electrons;
 Physics  Computational Physics;
 Quantum Physics
 EPrint:
 6 pages, 2 figures