Advances in Parallel Heat Bath Configuration Interaction

Abstract

Heat-bath configuration interaction (HCI) is a deterministic method that approaches the full CI limit at greatly reduced computational cost. In this work, computational improvements to the HCI algorithm are introduced targeting speed, parallel efficiency, and memory requirements. The new implementation introduces a hash function to distribute determinants and takes advantage of MPI and OpenMP for parallelism allowing for a (22e,168o) active space to be studied, which explicitly includes 2.39 × 10 7 variational determinants and 8.95 × 10 10 perturbative determinants. Benchmarks show up to 86% parallel efficiency of the perturbative step on 32 nodes (4096 cores) and a total efficiency of 74%. The new HCI implementation is benchmarked for accuracy against prior results and applied to study the triplet−quintet gap in the challenging [FeO(NH 3 ) 5 ] 2+ complex.

Exchange Correlation Potentials from Accurate FCI Densities Constructed from Slater Basis Functions

“…All Coulomb integrals will therefore be evaluated using the Resolution of the Identity (RI) approximation. [43][44][45] The orbitals and RDMs for the reference WFs come from the heat-bath configuration interaction (HBCI) procedure [46][47][48][49][50] using a tight threshold (10 −5 Ha) for accuracy of the variational wavefunction. For atoms and small molecules, the RDMs from this approach will be essentially FCI quality.…”

MQM 2022: The 10th Triennial Conference on Molecular Quantum Mechanics

“…16 In their numerical tests using a fragment extracted from a protein, they found that the Holder inequality is preferable for pairs and the Cauchy-Schwarz inequality is preferable for quartets. Heatbath configuration interaction (HCI) methods are the focus of a paper by the Zimmerman group at the University of Michigan, 17 and they report on a series of algorithmic improvements that take advantage of distributed-and shared-memory parallelism to achieve highly scalable performance. Next, Melo, Pederson, and Peralta describe an implementation of the Fermi-Loẅdin orbital self-interaction correction (FLOSIC) method that is suitable for existing quantum chemistry programs.…”

Modular Approach to Selected Configuration Interaction in an Arbitrary Spin Basis: Implementation and Comparison of Approaches

“…Extrapolations to the complete basis set limit have also been undertaken for small molecules and transition metals . The parallelization of the HCI algorithm has also been significantly improved allowing large active spaces ( n = 22, ϕ = 168) to be considered . In our initial implementation of a HCI-selection-like algorithm, we use the standard approach of SCI theories and generate the entire F -space, subjecting them to HCI selection criteria.…”