Accelerated Training of Physics-Informed Neural Networks (PINNs) using Meshless Discretizations

TL;DR

This paper introduces discretely-trained PINNs (DT-PINNs) that replace costly automatic differentiation with meshless RBF-FD discretizations, enabling faster training on irregular domains and improved computational efficiency, especially with double precision.

Contribution

The authors propose a novel DT-PINN method using meshless RBF-FD discretizations to accelerate training and handle irregular geometries, outperforming traditional PINNs in speed and efficiency.

Findings

DT-PINNs achieve 2-4x faster training times than vanilla-PINNs.

Using fp64 precision in DT-PINNs improves training speed without sacrificing accuracy.

RBF-FD discretizations of third-order accuracy are effective for PINN training.

Abstract

We present a new technique for the accelerated training of physics-informed neural networks (PINNs): discretely-trained PINNs (DT-PINNs). The repeated computation of partial derivative terms in the PINN loss functions via automatic differentiation during training is known to be computationally expensive, especially for higher-order derivatives. DT-PINNs are trained by replacing these exact spatial derivatives with high-order accurate numerical discretizations computed using meshless radial basis function-finite differences (RBF-FD) and applied via sparse-matrix vector multiplication. The use of RBF-FD allows for DT-PINNs to be trained even on point cloud samples placed on irregular domain geometries. Additionally, though traditional PINNs (vanilla-PINNs) are typically stored and trained in 32-bit floating-point (fp32) on the GPU, we show that for DT-PINNs, using fp64 on the GPU leads to significantly faster training times than fp32 vanilla-PINNs with comparable accuracy. We demonstrate the efficiency and accuracy of DT-PINNs via a series of experiments. First, we explore the effect of network depth on both numerical and automatic differentiation of a neural network with random weights and show that RBF-FD approximations of third-order accuracy and above are more efficient while being sufficiently accurate. We then compare the DT-PINNs to vanilla-PINNs on both linear and nonlinear Poisson equations and show that DT-PINNs achieve similar losses with 2-4x faster training times on a consumer GPU. Finally, we also demonstrate that similar results can be obtained for the PINN solution to the heat equation (a space-time problem) by discretizing the spatial derivatives using RBF-FD and using automatic differentiation for the temporal derivative. Our results show that fp64 DT-PINNs offer a superior cost-accuracy profile to fp32 vanilla-PINNs.