arXiv preprint ·

Data-Driven Optimization of Seminumerical Exchange for SIMD CPUs

Jesse DiCenso, Henryk Laqua and Martin Head-Gordon

Published
Source
arXiv preprint, posted before peer review
Original
Read it at arXiv
Topics
Chemical PhysicsQuantum Physics

Abstract

We present a seminumerical exchange algorithm (SNK) based on the sn-LinK method [Laqua et al., J. Chem. Theory Comput. 16, 1456 (2020)], optimized for modern multi-core central processing units with single instruction multiple data vector instructions and report its implementation into the Q-Chem quantum chemistry program. Vectorization and memory management strategies for the Boys function and 3-center-1-electron (3c1e) integral kernels are given. Several new considerations are made in the computation of the 3c1e integrals: the Head-Gordon Pople horizontal recurrence relations, same-center optimizations, and multishell optimizations, all implemented symbolically employing the SymPy symbolic algebra library and exported as optimized C++ code leveraging common-sub-expression elimination (CSE) optimization techniques. A general scheme for choosing optimal integral kernels is developed through rigorous timing benchmarks. The resulting implementation is tested against traditional 4c2e exchange for several systems and basis sets, showing 1-2 orders of magnitude speedups for systems of practically relevant sizes (100-500 atoms, 2000-10000 basis functions). For smaller molecules, comparisons are also made against the more efficient occ-RI-K method showing a break-even point around 50-100 atoms beyond which SNK dramatically outperforms due to its superior sub-quadratic (between linear and quadratic) time complexity as well as its effectively negligible memory demand.

The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.

Read it on ReadSearch

Each opens after you create an account, and brings you back to this paper.

Already a reader? Sign in

All new papers in Chemical Physics

Get the few papers worth your time in Chemical Physics every morning.

ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.

Papers belong to their authors and publishers; every page here links to the original. All research topics