Enabling Low-Latency Machine learning on Radiation-Hard FPGAs with hls4ml
Massachusetts Institute of Technology (MIT)(麻省理工学院);Department of Experimental Physics, European Organization for Nuclear Research (CERN)(欧洲核子研究中心(CERN)实验物理系);University of Milano-Bicocca (Italy)(米兰-布雷拉大学)
Katya Govorkova, Julian Garcia Pardinas, Vladimir Loncar, Victoria Nguyen, Sebastian Schmitt, Marco Pizzichemi, Loris Martinazzoli, Eluned Anne Smith
发表信息:
Machine Learning: Science and Technology 7 (2026) 045024
This paper presents an end-to-end demonstration of a viable, ultra-fast, radiation-hard machine learning (ML) application on FPGAs, which could be used in future high-energy physics experiments. We present a three-fold contribution, with the PicoCal calorimeter, planned for the LHCb Upgrade II experiment, used as a test case. First, we develop a lightweight autoencoder to compress a 32-sample timing readout, representative of that of the PicoCal, into a two-dimensional latent space. Second, we introduce a systematic, hardware-aware quantization strategy and show that the model can be reduced to 10-bit weights with minimal performance loss. Third, as a barrier to the adoption of on-detector ML is the lack of support for radiation-hard FPGAs in the High-Energy Physics community's standard ML synthesis tool, hls4ml, we develop a new backend for this library. This new back-end enables the automatic translation of ML models into High-Level Synthesis (HLS) projects for the Microchip PolarFire family of FPGAs, one of the few commercially available and radiation hard FPGAs. We present the synthesis of the autoencoder on a target PolarFire FPGA, which indicates that a latency of 25 ns can be achieved. We show that the resources utilized are low enough that the model can be placed within the inherently protected logic of the FPGA. Our extension to hls4ml is a significant contribution, paving the way for broader adoption of ML on FPGAs in high-radiation environments.
Low Latency Edge Classification GNN for Particle Trajectory Tracking on FPGAs
National Yang Ming Chiao Tung University(国立阳明交通大学);University of California San Diego(加利福尼亚大学圣迭戈分校);University of Washington(华盛顿大学);University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)
Towards Tensor Network Models for Low-Latency Jet Tagging on FPGAs
Dipartimento di Fisica e Astronomia "Galileo Galilei", University of Padua(伽利略·伽利莱物理与天文学系,帕多瓦大学);PlanQC GmbH(PlanQC公司);Institute for Complex Quantum Systems, Ulm University(复杂量子系统研究所,乌尔姆大学);European Organization for Nuclear Research (CERN)(欧洲核子研究中心);Padua Quantum Technologies Research Center, University of Padua(帕多瓦量子技术研究中心,帕多瓦大学)
Ultra-low latency quantum-inspired machine learning predictors implemented on FPGA
Istituto Nazionale di Fisica Nucleare (INFN)(国家核物理研究所);Dipartimendo di Fisica e Astronomia (DFA)(物理与天文学系);University of Padua(帕多瓦大学);Dipartimendo di Ingegneria dell’Informazione (DEI)(信息工程系);Dipartimendo di Ingegneria Industriale (DII)(工业工程系);Padua Quantum Technology Research Center(帕多瓦量子技术研究中心);Tensor AI Solutions GmbH(张量人工智能解决方案有限公司)
Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml
University of Washington(华盛顿大学);Massachusetts Institute of Technology(麻省理工学院);SLAC National Accelerator Laboratory(SLAC国家加速器实验室);European Organization for Nuclear Research (CERN)(欧洲核子研究组织(CERN))
SparsePixels: Efficient Convolution for Sparse Data on FPGAs
University of Pennsylvania, USA(美国宾夕法尼亚大学);Institute of Physics Belgrade, Serbia(塞尔维亚贝尔格莱德物理研究所);Massachusetts Institute of Technology, USA(美国麻省理工学院);Institute for Artificial Intelligence and Fundamental Interactions, USA(美国人工智能与基本相互作用研究所)
Fast inference of deep neural networks in FPGAs for particle physics
Fermi National Accelerator Laboratory(费米国家加速器实验室);Massachusetts Institute of Technology(麻省理工学院);CERN(欧洲核子研究中心);University of Illinois at Chicago(伊利诺伊大学芝加哥分校)
Fast inference of Boosted Decision Trees in FPGAs for particle physics
European Organization for Nuclear Research (CERN)(欧洲核子研究组织(CERN));Columbia University(哥伦比亚大学);University of California San Diego(加利福尼亚大学圣迭戈分校);Massachusetts Institute of Technology(麻省理工学院);Rhodes College(罗德学院);Fermi National Accelerator Laboratory(费米国家加速器实验室);Institute of Physics Belgrade(贝尔格莱德物理研究所);University of Illinois at Chicago(伊利诺伊大学芝加哥分校)
Low Latency Neural Networks using Heterogenous Resources on FPGA for the Belle II Trigger
Belle II触发系统的主要组成部分之一是神经网络触发器。它的任务是估计在实验漂移室中观察到的z-顶点粒子径迹。触发器在FPGA上实现,以确保操作期间的灵活性并利用其IO功能。同时,实现必须在几百纳秒内估计顶点以满足实验的要求。在第一次碰撞期间,该触发器的第一个版本是可操作的。虽然它能够估计顶点,但它在可能的吞吐量和时序收敛方面存在一些缺点。这些都是这项工作的重点,它修改了原来的设计,允许两个网络并行运行,减少路由拥塞。我们进行了重调度的乘法和累加,这是在这样的网络中的基本操作。原设计试图尽可能多地并行化,而重调度则试图通过重用处理模块来减少并行数据传输的数量。这样,DSP的资源消耗减少了40%。为了通过并行操作额外的网络来进一步提高吞吐量,我们研究了用于乘法和累加操作的SRAM-LUT和DSP的平衡使用。利用找到的平衡比率,触发器能够在所需的延迟内在目标FPGA上并行操作两个神经网络。
Fast convolutional neural networks on FPGAs with hls4ml
European Organization for Nuclear Research (CERN)(欧洲核子研究组织);Institute of Physics Belgrade(贝尔格莱德物理研究所);Politecnico di Milano(米兰理工大学);California Institute of Technology(加州理工学院);Chalmers University of Technology(查尔姆斯理工大学);Zenseact(Zenseact公司);ICEPP, University of Tokyo(东京大学国际基本粒子物理研究所);Columbia University(哥伦比亚大学);University of California San Diego(加州大学圣地亚哥分校);Massachusetts Institute of Technology(麻省理工学院);Fermi National Accelerator Laboratory(费米国家加速器实验室);Purdue University(普渡大学);HawkEye360(鹰眼360公司);University of Illinois at Chicago(伊利诺伊大学芝加哥分校);Rhodes College(罗德学院)