Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
133 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Low Latency Edge Classification GNN for Particle Trajectory Tracking on FPGAs (2306.11330v2)

Published 20 Jun 2023 in cs.AR, cs.LG, and hep-ex

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.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (28)
  1. A. Ryd and L. Skinnari, “Tracking triggers for the HL-LHC,” arXiv:2010.13557, 2020.
  2. P. Billoir, “Progressive track recognition with a kalman-like fitting procedure,” Comput. Phys. Commun., vol. 57, no. 1-3, pp. 390–394, 1989.
  3. P. Billoir and S. Qian, “Simultaneous pattern recognition and track fitting by the kalman filtering method,” Nucl. Instrum. Methods Phys. Res. A, vol. 294, no. 1-2, pp. 219–228, 1990.
  4. R. Mankel, “A concurrent track evolution algorithm for pattern recognition in the HERA-B main tracking system,” Nucl. Instrum. Methods Phys. Res. A, vol. 395, no. 2, pp. 169–184, 1997.
  5. R. Frühwirth, “Application of kalman filtering to track and vertex fitting,” Nucl. Instrum. Methods Phys. Res. A, vol. 262, no. 2-3, pp. 444–450, 1987.
  6. CERN, “CERN yellow reports: Monographs, vol 4 (2017): High-luminosity large hadron collider (HL-LHC) technical design report v. 0.1,” Jan. 2017.
  7. Z. Tao, “Level-1 track finding with an all-FPGA system at CMS for the HL-LHC,” arXiv:1901.03745, 2019.
  8. J. Shlomi, P. Battaglia, and J.-R. Vlimant, “Graph neural networks in particle physics,” Machine Learning: Science and Technology, vol. 2, no. 2, p. 021001, 2020.
  9. X. Ju, S. Farrell, P. Calafiura, D. Murnane, Prabhat, L. Gray, T. Klijnsma, K. Pedro, G. Cerati, J. Kowalkowski, G. Perdue, P. Spentzouris, N. Tran, J.-R. Vlimant, A. Zlokapa, J. Pata, M. Spiropulu, S. An, A. Aurisano, J. Hewes, A. Tsaris, K. Terao, and T. Usher, “Graph neural networks for particle reconstruction in high energy physics detectors,” arXiv:2003.11603, 2020.
  10. J. Duarte and J.-R. Vlimant, “Graph neural networks for particle tracking and reconstruction,” in Artificial Intelligence for High Energy Physics.   WORLD SCIENTIFIC, 2022, pp. 387–436.
  11. J. Pata, J. Duarte, J.-R. Vlimant, M. Pierini, and M. Spiropulu, “MLPF: efficient machine-learned particle-flow reconstruction using graph neural networks,” Eur. Phys. J. C Part. Fields, vol. 81, no. 5, pp. 1–14, 2021.
  12. A. M. Sirunyan, C. collaboration et al., “Particle-flow reconstruction and global event description with the cms detector,” JINST, vol. 12, no. 10, p. 10003, 2017.
  13. G. DeZoort, S. Thais, J. Duarte, V. Razavimaleki, M. Atkinson, I. Ojalvo, M. Neubauer, and P. Elmer, “Charged particle tracking via edge-classifying interaction networks,” Comput. Softw. Big Sci., vol. 5, no. 1, pp. 1–13, 2021.
  14. M. Yan, L. Deng, X. Hu, L. Liang, Y. Feng, X. Ye, Z. Zhang, D. Fan, and Y. Xie, “HyGCN: A GCN accelerator with hybrid architecture,” in 2020 IEEE International Symposium on High Performance Computer Architecture (HPCA).   IEEE, 2020, pp. 15–29.
  15. S. Liang, Y. Wang, C. Liu, L. He, H. Li, D. Xu, and X. Li, “EnGN: A high-throughput and energy-efficient accelerator for large graph neural networks,” IEEE Trans. Comput., vol. 70, no. 9, pp. 1511–1525, 2021.
  16. K. Kiningham, P. Levis, and C. Re, “GRIP: A graph neural network accelerator architecture,” IEEE Trans. Comput., pp. 1–12, 2022.
  17. T. Geng, A. Li, R. Shi, C. Wu, T. Wang, Y. Li, P. Haghi, A. Tumeo, S. Che, S. Reinhardt, and M. C. Herbordt, “AWB-GCN: A graph convolutional network accelerator with runtime workload rebalancing,” in 2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO).   IEEE, 2020, pp. 922–936.
  18. J. Li, A. Louri, A. Karanth, and R. Bunescu, “GCNAX: A flexible and energy-efficient accelerator for graph convolutional neural networks,” in 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA).   IEEE, 2021, pp. 775–788.
  19. X. Chen, Y. Wang, X. Xie, X. Hu, A. Basak, L. Liang, M. Yan, L. Deng, Y. Ding, Z. Du, and Y. Xie, “Rubik: A hierarchical architecture for efficient graph neural network training,” IEEE Trans. Comput.-aided Des. Integr. Circuits Syst., vol. 41, no. 4, pp. 936–949, 2022.
  20. T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” arXiv preprint arXiv:1609.02907, 2016.
  21. W. Hamilton, Z. Ying, and J. Leskovec, “Inductive representation learning on large graphs,” Advances in neural information processing systems, vol. 30, 2017.
  22. FastML Team, “fastmachinelearning/hls4ml,” 2021. [Online]. Available: https://github.com/fastmachinelearning/hls4ml
  23. J. Duarte et al., “Fast inference of deep neural networks in FPGAs for particle physics,” JINST, vol. 13, no. 07, p. P07027, 2018.
  24. A. Zabi, J. W. Berryhill, E. Perez, and A. D. Tapper, “The Phase-2 Upgrade of the CMS Level-1 Trigger,” 2020.
  25. A. Elabd, V. Razavimaleki, S.-Y. Huang, J. Duarte, M. Atkinson, G. DeZoort, P. Elmer, S. Hauck, J.-X. Hu, S.-C. Hsu et al., “Graph neural networks for charged particle tracking on fpgas,” Frontiers in big Data, vol. 5, 2022.
  26. P. W. Battaglia, R. Pascanu, M. Lai, D. Rezende, and K. Kavukcuoglu, “Interaction networks for learning about objects, relations and physics,” vol. 29, 2016.
  27. A. Heintz, V. Razavimaleki, J. Duarte, G. DeZoort, I. Ojalvo, S. Thais, M. Atkinson, M. Neubauer, L. Gray, S. Jindariani, N. Tran, P. Harris, D. Rankin, T. Aarrestad, V. Loncar, M. Pierini, S. Summers, J. Ngadiuba, M. Liu, E. Kreinar, and Z. Wu, “Accelerated charged particle tracking with graph neural networks on FPGAs,” arXiv:2012.01563, 2020.
  28. S. Amrouche, L. Basara, P. Calafiura, V. Estrade, S. Farrell, D. R. Ferreira, L. Finnie, N. Finnie, C. Germain, V. V. Gligorov, T. Golling, S. Gorbunov, H. Gray, I. Guyon, M. Hushchyn, V. Innocente, M. Kiehn, E. Moyse, J.-F. Puget, Y. Reina, D. Rousseau, A. Salzburger, A. Ustyuzhanin, J.-R. Vlimant, J. S. Wind, T. Xylouris, and Y. Yilmaz, “The tracking machine learning challenge: Accuracy phase,” in The NeurIPS ’18 Competition.   Cham: Springer International Publishing, 2020, pp. 231–264.
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now