Low Latency Edge Classification GNN for Particle Trajectory Tracking on FPGAs

Shi Yu Huang, Yun Chen Yang, Yu Ru Su, Bo Cheng Lai, Javier Duarte, Scott Hauck, Shih Chieh Hsu, Jin Xuan Hu, Mark S. Neubauer

Research output: Chapter in Book/Report/Conference proceedingConference contribution

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.

Original languageEnglish (US)
Title of host publicationProceedings - 2023 33rd International Conference on Field-Programmable Logic and Applications, FPL 2023
EditorsNele Mentens, Nele Mentens, Leonel Sousa, Pedro Trancoso, Nikela Papadopoulou, Ioannis Sourdis
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages294-298
Number of pages5
ISBN (Electronic)9798350341515
DOIs
StatePublished - 2023
Externally publishedYes
Event33rd International Conference on Field-Programmable Logic and Applications, FPL 2023 - Gothenburg, Sweden
Duration: Sep 4 2023Sep 8 2023

Publication series

NameProceedings - 2023 33rd International Conference on Field-Programmable Logic and Applications, FPL 2023

Conference

Conference33rd International Conference on Field-Programmable Logic and Applications, FPL 2023
Country/TerritorySweden
CityGothenburg
Period9/4/239/8/23

Keywords

  • FPGA
  • Graph Neural Network
  • Particle tracking

ASJC Scopus subject areas

  • Computer Science Applications
  • Software

Fingerprint

Dive into the research topics of 'Low Latency Edge Classification GNN for Particle Trajectory Tracking on FPGAs'. Together they form a unique fingerprint.

Cite this