Cholesky decomposition of complex two-electron integrals over GIAOs: Efficient MP2 computations for large molecules in strong magnetic fields
Simon Blaschke, Stella Stopkowicz

TL;DR
This paper introduces a Cholesky decomposition approach for complex electron-repulsion integrals in large molecules under strong magnetic fields, significantly improving computational efficiency in quantum chemistry calculations.
Contribution
The authors developed and implemented a Cholesky decomposition method for complex ERIs over GIAOs, enabling efficient MP2 computations for large molecules in magnetic fields with rigorous error control.
Findings
Cholesky decomposition achieves effective compression of complex ERIs.
The method allows MP2 calculations on systems with over 2000 basis functions.
Significant reduction in computational time for large-scale magnetic field calculations.
Abstract
In large-scale quantum-chemical calculations the electron-repulsion integral (ERI) tensor rapidly becomes the bottleneck in terms of memory and disk space. When an external finite magnetic field is employed, this problem becomes even more pronounced because of the reduced permutational symmetry and the need to work with complex integrals and wave-function parameters. One way to alleviate the problem is to employ a Cholesky decomposition (CD) to the complex ERIs over gauge-including atomic orbitals. The CD scheme establishes favourable compression rates by selectively discarding linearly dependent product densities from the chosen basis set while maintaining a rigorous and robust error control. This error control constitutes the main advantage over conceptually similar methods such as density fitting which rely on employing pre-defined auxiliary basis sets. We implemented the use of the…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
