Implementation of the Multigrid Gaussian-Plane-Wave Algorithm with GPU Acceleration in PySCF
Year: 2026
Type: article
Abstract: We introduce a GPU-accelerated multigrid Gaussian-Plane-Wave density fitting (FFTDF) approach for efficient Fock builds and nuclear gradient evaluations within Kohn-Sham density functional theory, as implemented in the GPU4PySCF module of PySCF. Our CUDA kernels employ a grid-based parallelization strategy for contracting Gaussian basis function pairs and achieve 50% to 80% of the FP64 peak performance on NVIDIA GPUs, with no loss of efficiency for high angular momentum (up to f-shell) functions. Benchmark calculations on molecules and solids with up to 1536 atoms and 20480 basis functions show up to 25 × speedup on an H100 GPU relative to the CPU implementation on a 28-core shared-memory node. For a 256-water cluster, the ground-state energy and nuclear gradients can be computed in ∼30 seconds on a single H100 GPU. This implementation serves as an open-source foundation for many applications, such as ab initio molecular dynamics and high-throughput calculations.
Institutions: California Institute of Technology, ByteDance (China)
Language: English
Subfield: Atomic and Molecular Physics, and Optics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
Awards: G1143651532, G1590684107, G2034695884, G2792428870, Collaborative Research: Frameworks: Building the PySCF Software Infrastructure for Next-Generation Electronic Structure Simulations, G3791614690, CENTER FOR MOLECULAR MAGNETIC QUANTUM MATERIALS, MULTISCALE SIMULATION SOFTWARE FOR ENERGY TRANSDUCTION IN BIOMOLECULAR SYSTEMS: FROM ELECTRONS TO THE MESOSCALE, MULTISCALE SIMULATION SOFTWARE FOR ENERGY TRANSDUCTION IN BIOMOLECULAR SYSTEMS: FROM ELECTRONS TO THE MESOSCALE, G8319700791
OpenAlex ID: W7141574431
