Ultrafast jet classification at the HL-LHC

Three machine learning models are used to perform jet origin classification. These models are optimized for deployment on a field-programmable gate array device. In this context, we demonstrate how latency and resource consumption scale with the input size and choice of algorithm. Moreover, the mode...

Full description

Saved in:
Bibliographic Details
Published inMachine learning: science and technology Vol. 5; no. 3
Main Authors Odagiu, Patrick, Que, Zhiqiang, Duarte, Javier, Haller, Johannes, Kasieczka, Gregor, Lobanov, Artur, Loncar, Vladimir, Luk, Wayne, Ngadiuba, Jennifer, Pierini, Maurizio, Rincke, Philipp, Seksaria, Arpita, Summers, Sioni, Sznajder, Andre, Tapper, Alexander, Årrestad, Thea K.
Format Journal Article
LanguageEnglish
Published United States IOP Publishing 18.07.2024
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Three machine learning models are used to perform jet origin classification. These models are optimized for deployment on a field-programmable gate array device. In this context, we demonstrate how latency and resource consumption scale with the input size and choice of algorithm. Moreover, the models proposed here are designed to work on the type of data and under the foreseen conditions at the CERN large hadron collider during its high-luminosity phase. Through quantization-aware training and efficient synthetization for a specific field programmable gate array, we show that $\mathcal{O}(100)$ ns inference of complex architectures such as Deep Sets and Interaction Networks is feasible at a relatively low computational resource cost.
Bibliography:FERMILAB-PUB-24-0030-CMS-CSAID-PPD; arXiv:2402.01876
USDOE Office of Science (SC), High Energy Physics (HEP)
AC02-07CH11359; SC0021187; FOA-0002705; OAC-2117997; 390833306; PZ00P2 201594; EP/V028251/1; EP/L016796/1; EP/N031768/1; EP/P010040/1; EP/S030069/1
Engineering and Physical Sciences Research Council (EPSRC)
Swiss National Science Foundation (SNSF)
German Research Foundation (DFG)
National Science Foundation (NSF)
ISSN:2632-2153
2632-2153