@inproceedings{9dd2c8f5867b4f17b2d7149896e015df,
title = "Low Latency Edge Classification GNN for Particle Trajectory Tracking on FPGAs",
abstract = "In-time particle trajectory reconstruction in the Large Hadron Collider is challenging due to the high collision rate and numerous particle hits. Using GNN (Graph Neural Network) on FPGA has enabled superior accuracy with flexible trajectory classification. However, existing GNN architectures have inefficient resource usage and insufficient parallelism for edge classification. This paper introduces a resource-efficient GNN architecture on FPGAs for low latency particle tracking. The modular architecture facilitates design scalability to support large graphs. Leveraging the geometric properties of hit detectors further reduces graph complexity and resource usage. Our results on Xilinx UltraScale+ VU9P demonstrate 1625x and 1574x performance improvement over CPU and GPU respectively.",
keywords = "FPGA, Graph Neural Network, Particle tracking",
author = "Huang, {Shi Yu} and Yang, {Yun Chen} and Su, {Yu Ru} and Lai, {Bo Cheng} and Javier Duarte and Scott Hauck and Hsu, {Shih Chieh} and Hu, {Jin Xuan} and Neubauer, {Mark S.}",
note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 33rd International Conference on Field-Programmable Logic and Applications, FPL 2023 ; Conference date: 04-09-2023 Through 08-09-2023",
year = "2023",
doi = "10.1109/FPL60245.2023.00050",
language = "English (US)",
series = "Proceedings - 2023 33rd International Conference on Field-Programmable Logic and Applications, FPL 2023",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "294--298",
editor = "Nele Mentens and Nele Mentens and Leonel Sousa and Pedro Trancoso and Nikela Papadopoulou and Ioannis Sourdis",
booktitle = "Proceedings - 2023 33rd International Conference on Field-Programmable Logic and Applications, FPL 2023",
address = "United States",
}