Recent studies have shown promising results for track finding in dense environments using Graph Neural Network (GNN)-based algorithms. However, GNN-based track finding is computationally slow on CPUs, necessitating the use of coprocessors to accelerate the inference time. Additionally, the large input graph size demands a large device memory for efficient computation, a requirement not met by all computing facilities used for particle physics experiments, particularly those lacking advanced GPUs. Furthermore, deploying the GNN-based track-finding algorithm in a production environment requires the installation of all dependent software packages, exclusively utilized by this algorithm. These computing challenges must be addressed for the successful implementation of GNN-based track-finding algorithm into production settings. In response, we introduce a ``GNN-based tracking as a service'' approach, incorporating a custom backend within the NVIDIA Triton inference server to facilitate GNN-based tracking. This paper presents the performance of this approach using the Perlmutter supercomputer at NERSC.
@article{arxiv.2402.09633,
title = {Graph Neural Network-based Tracking as a Service},
author = {Haoran Zhao and Andrew Naylor and Shih-Chieh Hsu and Paolo Calafiura and Steven Farrell and Yongbing Feng and Philip Coleman Harris and Elham E Khoda and William Patrick Mccormack and Dylan Sheldon Rankin and Xiangyang Ju},
journal= {arXiv preprint arXiv:2402.09633},
year = {2024}
}
Comments
7 pages, 4 figures, Proceeding of Connected the Dots Workshop (CTD 2023)