GPU-acceleration of tensor renormalization with pytorch using CUDA

DOI10.1016/J.CPC.2023.108941MaRDI QIDQ6086774zbMATH OpenOpenAlexFDO

Authors Raghav G. Jha, Abhishek Samlodia

Publication date 10 November 2023

Published in Computer Physics Communications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2306.00358

zbMATH Keywords

lattice field theoryspeedupspin modelstensor

Mathematics Subject Classification ID

Graphical indices (Wiener index, Zagreb index, Randi? index, etc.) (05C09) Quantum computation (81P68) Spinor and twistor methods applied to problems in quantum theory (81R25) Renormalization group methods applied to problems in quantum field theory (81T17) Quantum field theory on lattices (81T25) Matrix models and tensor models for quantum field theory (81T32)

Abstract: We show that numerical computations based on tensor renormalization group (TRG) methods can be significantly accelerated with PyTorch on graphics processing units (GPUs) by leveraging NVIDIA's Compute Unified Device Architecture (CUDA). We find improvement in the runtime and its scaling with bond dimension for two-dimensional systems. Our results establish that the utilization of GPU resources is essential for future precision computations with TRG.

Recommendations

• GPU accelerated tensor contractions in the plaquette renormalization scheme
• Cost-efficient cutoff method for tensor renormalization group with randomized singular value decomposition
• High performance implementations of the 2D Ising model on GPUs
• GPU acceleration for Hermitian eigensystems
• Transformation of a nucleon-nucleon potential operator into its SU(3) tensor form using GPUs

This page was built for publication: GPU-acceleration of tensor renormalization with pytorch using CUDA

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6086774)