h-analysis and data-parallel physics-informed neural networks

TL;DR

This paper presents a novel data-parallel approach for physics-informed neural networks (PINNs) that enhances scalability and efficiency across multiple GPUs, supported by theoretical convergence bounds and extensive experiments.

Contribution

It introduces a new protocol based on $h$-analysis and data-parallel acceleration with Horovod, providing scale-robust PINNs with theoretical guarantees and practical efficiency.

Findings

Efficient multi-GPU acceleration without training compromise

Theoretical convergence bounds for generalization error

Robustness demonstrated through complex numerical experiments

Abstract

We explore the data-parallel acceleration of physics-informed machine learning (PIML) schemes, with a focus on physics-informed neural networks (PINNs) for multiple graphics processing units (GPUs) architectures. In order to develop scale-robust and high-throughput PIML models for sophisticated applications which may require a large number of training points (e.g., involving complex and high-dimensional domains, non-linear operators or multi-physics), we detail a novel protocol based on $h$-analysis and data-parallel acceleration through the Horovod training framework. The protocol is backed by new convergence bounds for the generalization error and the train-test gap. We show that the acceleration is straightforward to implement, does not compromise training, and proves to be highly efficient and controllable, paving the way towards generic scale-robust PIML. Extensive numerical experiments with increasing complexity illustrate its robustness and consistency, offering a wide range of possibilities for real-world simulations.