Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml
| Autor: | Elham E Khoda, Dylan Rankin, Rafael Teixeira de Lima, Philip Harris, Scott Hauck, Shih-Chieh Hsu, Michael Kagan, Vladimir Loncar, Chaitanya Paikara, Richa Rao, Sioni Summers, Caterina Vernieri, Aaron Wang |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Physics - Instrumentation and Detectors hep-ex cs.LG FOS: Physical sciences Machine Learning (stat.ML) Instrumentation and Detectors (physics.ins-det) stat.ML Machine Learning (cs.LG) High Energy Physics - Experiment Computing and Computers Human-Computer Interaction Computer Science::Hardware Architecture High Energy Physics - Experiment (hep-ex) Artificial Intelligence Statistics - Machine Learning Detectors and Experimental Techniques Mathematical Physics and Mathematics physics.ins-det Software Particle Physics - Experiment |
| Popis: | Recurrent neural networks have been shown to be effective architectures for many tasks in high energy physics, and thus have been widely adopted. Their use in low-latency environments has, however, been limited as a result of the difficulties of implementing recurrent architectures on field-programmable gate arrays (FPGAs). In this paper we present an implementation of two types of recurrent neural network layers -- long short-term memory and gated recurrent unit -- within the hls4ml framework. We demonstrate that our implementation is capable of producing effective designs for both small and large models, and can be customized to meet specific design requirements for inference latencies and FPGA resources. We show the performance and synthesized designs for multiple neural networks, many of which are trained specifically for jet identification tasks at the CERN Large Hadron Collider. 12 pages, 6 figures, 5 tables |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|