GPU Accelerated Minimal Auxiliary Basis Approach TDDFT for Large Organic Molecules

TL;DR

This paper presents a GPU-accelerated TDDFT implementation with minimal auxiliary basis, enabling efficient excited-state calculations for large organic molecules on high-performance GPUs.

Contribution

The authors develop and demonstrate a GPU-accelerated TDDFT method with minimal auxiliary basis, achieving practical large-system excited-state calculations with high accuracy and efficiency.

Findings

Achieves about 0.03--0.05 eV excitation energy error on benchmark set.

Calculates 15 low-lying excited states for systems with 300-3000 atoms in minutes to hours.

Enables practical excited-state calculations for large biomolecular systems.

Abstract

We introduce a GPU-accelerated implementation of time-dependent density functional theory with the minimal auxiliary basis approach (TDDFT-risp) in GPU4PySCF, together with large system demonstrations carried out using the Tamm--Dancoff approximation (TDA-risp). The method combines GPU-accelerated three-center integral evaluation, tensor contractions, exchange-space truncation, omission of hydrogen atoms from the auxiliary basis, and a host memory assisted Davidson solver. On the EXTEST42 benchmark set, a conservative 40 eV exchange cutoff yields excitation-energy errors relative to standard TDA of about 0.03--0.05 eV for low-lying states. For systems of 300 to 3000 atoms, we demonstrate that TDA-risp calculations of 15 low-lying excited states with $\omega$B97XD/def2-SVP complete on a single A100 GPU with wall times ranging from minutes to hours. These results position GPU-TDDFT-risp as a practical route toward excited-state calculations for large organic and biomolecular systems with thousands of atoms.