TT-PINN: A Tensor-Compressed Neural PDE Solver for Edge Computing
Machine Learning
2023-02-28 v1 Hardware Architecture
Distributed, Parallel, and Cluster Computing
Numerical Analysis
Numerical Analysis
Abstract
Physics-informed neural networks (PINNs) have been increasingly employed due to their capability of modeling complex physics systems. To achieve better expressiveness, increasingly larger network sizes are required in many problems. This has caused challenges when we need to train PINNs on edge devices with limited memory, computing and energy resources. To enable training PINNs on edge devices, this paper proposes an end-to-end compressed PINN based on Tensor-Train decomposition. In solving a Helmholtz equation, our proposed model significantly outperforms the original PINNs with few parameters and achieves satisfactory prediction with up to 15 overall parameter reduction.
Cite
@article{arxiv.2207.01751,
title = {TT-PINN: A Tensor-Compressed Neural PDE Solver for Edge Computing},
author = {Ziyue Liu and Xinling Yu and Zheng Zhang},
journal= {arXiv preprint arXiv:2207.01751},
year = {2023}
}