Abstract
The Belle II experiment in Tsukuba, Japan, searches for physics beyond the Standard Model. Electrons and positrons are accelerated in the SuperKEKB collider to collide at the interaction point in the Belle II detector. Since the resulting data volume is too large, a multi-stage trigger system is installed to sort out physically irrelevant events. In order to find decays with displaced vertex, which are candidates for the indirect detection of dark matter, the FPGA-based level 1 trigger has to be upgraded. A convolution neural network (CNN) with parallel convolution presented in this work enables the finding of displaced vertex tracks. To do this, the CNN must process 32,000,000 frames per second in parallel and provide an estimate of the origin of these tracks for each frame. The complete system has been successfully implemented on the FPGA platform (XCVU160) used in the experiment and meets the specified requirements of the trigger system.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Belle II Collaboration. Accessed 03 May 2023. https://www.belle2.org/
Abe, T.: Belle II Collaboration 2010 “Belle II Technical Design Report”. arXiv arXiv:1011.0352
Lai, Y.T., et al.: Development of the Level-1 track trigger with Central Drift Chamber detector in Belle II experiment and its performance in SuperKEKB 2019 Phase 3 operation. J. Instrument. 15(06), C06063 (2020)
Unger, K.L., Bähr, S., Becker, J., Iwasaki, Y., Kim, K., Lai, Y.T.: Realization of a state machine based detection for Track Segments in the trigger system of the Belle II experiment. In: Proceedings of Topical Workshop on Electronics for Particle Physics Proceedings of Science, vol. 370 (2020). https://doi.org/10.22323/1.370.0145
Won, E., Moon, H.: Development of an event time finding algorithm for multi-wire drift chamber-based Level-1 trigger system in the Belle II experiment. J. Korean Phys. Soc. 80, 1–6 (2022)
Won, E., Kim, J.B., Ko, B.R.: Three-dimensional fast tracker for the central drift chamber based level-1 trigger system in the Belle II experiment. J. Korean Phys. Soc. 72(1), 33–37 (2018). https://doi.org/10.3938/jkps.72.33
Baehr, S., et al.: Low latency neural networks using heterogenous resources on FPGA for the Belle II trigger. In: Connecting the Dots and Workshop on Intelligent Trackers (2019)
Iwasaki, Y., Cheon, B., Won, E., Varner, G.: Level 1 trigger system for the Belle II experiment. In: 2010 17th IEEE-NPSS Real Time Conference, pp. 1–9. IEEE (2010)
Köhne, J.K., et al.: “Realization of a second level neural network trigger for the H1 experiment at HERA” nuclear instruments and methods in physics research section A: accelerators. Spectromet. Detect. Assoc. Equip. 389, 128–133 (1997). https://doi.org/10.1016/S0168-9002(97)00062-4
Kiesling, C.M., et al.: The H1 neural network trigger project. In: AIP Conference Proceedings, vol. 583, pp.36–44 (2002). https://doi.org/10.1063/1.1405259
Baehr, S., et al.: A neural network on FPGAs for the z-vertex track trigger in Belle II. J. Instrument. (2017). https://doi.org/10.1088/1748-0221/12/03/C03065
Baehr, S., et al.: Data reduction and readout triggering in particle physics experiments using neural networks on fpgas. In: 2018 IEEE 18th International Conference on Nanotechnology (IEEE-NANO), pp. 1–4 (2018). https://doi.org/10.1109/NANO.2018.8626239
Baehr, S., et al.: Low latency neural networks using heterogenous resources on fpga for the Belle II trigger. arXiv https://doi.org/10.48550/arXiv.1910.13679 (2019)
Unger, K.L., et al.: Operation of the neural z-vertex track trigger for Belle II in 2021-a hardware perspective. In: Journal of Physics: Conference Series (2023). https://doi.org/10.1088/1742-6596/2438/1/012056
Hartmann, F., et al.: The phase-2 upgrade of the CMS level-1 trigger. CERN-LHCC-2020-004, CMS-TDR-021 (2020)
Alimena, J., et al.: Fast convolutional neural networks for identifying long-lived particles in a high-granularity calorimeter. J. Instrument. 15, P12006 (2020). https://doi.org/10.1088/1748-0221/15/12/P12006
Duarte, J., et al.: FPGA-accelerated machine learning inference as a service for particle physics computing. Comput. Softw. Big Sci. 3(1), 1–15 (2019). https://doi.org/10.1007/s41781-019-0027-2
Coelho, C.N., Kuusela, A., Zhuang, H., Aarrestad, T., Loncar, V., Ngadiuba, J., et al.: Automatic deep heterogeneous quantization of deep neural networks for ultra low-area, low-latency inference on the edge at particle colliders. Nat. Mach. Intell. (2020). https://doi.org/10.1038/s42256-021-00356-5
Loncar, V., Pierini, M., Summers, S., Di Guglielmo, G., Duarte, J., Harris, P., et al.: Compressing deep neural networks on FPGAs to binary and ternary precision with hls4ml. Mach. Learn. 2, 015001 (2020 ). https://doi.org/10.1088/2632-2153/
Aarrestad, T., Loncar, V., Ghielmetti, N., Pierini, M., Summers, S., Ngadiuba, J., et al.: Fast convolutional neural networks on FPGAs with hls4ml’. Mach. Learn. Sci. Tech. 2, 045015 (2021). https://doi.org/10.1088/2632-2153/ac0ea1
Deiana, A., et al.: Applications and techniques for fast machine learning in science. Front Big Data Sec. Big Data AI High Energy Phys. 5, 787421 (2022). https://doi.org/10.3389/fdata.2022.787421
Acknowledgments
Funded by the German Federal Ministry of Education and Research under “Verbundprojekt 05H2021 (ErUM-FSP T09) - Belle II: Pixeldetektor, Software und erste Datenanalysen”
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Unger, K. et al. (2023). A Convolution Neural Network Based Displaced Vertex Trigger for the Belle II Experiment. In: Palumbo, F., Keramidas, G., Voros, N., Diniz, P.C. (eds) Applied Reconfigurable Computing. Architectures, Tools, and Applications. ARC 2023. Lecture Notes in Computer Science, vol 14251. Springer, Cham. https://doi.org/10.1007/978-3-031-42921-7_12
Download citation
DOI: https://doi.org/10.1007/978-3-031-42921-7_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-42920-0
Online ISBN: 978-3-031-42921-7
eBook Packages: Computer ScienceComputer Science (R0)