Intelligent experiments through real-time AI: Fast Data Processing and Autonomous Detector Control for sPHENIX and future EIC detectors

TL;DR

This work addresses the challenge of processing high-rate data at sPHENIX by developing a real-time, FPGA-embedded AI pipeline that ingests detector hits, constructs tracks, and performs heavy-flavor tagging with Graph Neural Network–based edge classification and an attention-based trigger (BGN-ST). Two deployment paths are explored: a track-based model (BGN-ST) and a hit-based model (GarNet), with both FlowGNN and hls4ml techniques used to generate FPGA-ready implementations on FELIX-712 and Alveo hardware. Key findings show that the BGN-ST approach achieves the highest accuracy for charm and beauty decays, while the GarNet hit-based approach offers low-latency, end-to-end deployment, achieving latencies from hundreds of nanoseconds to tens of microseconds depending on the path, and enabling real-time selection of heavy-flavor events at rates up to several MHz. The work demonstrates that real-time AI–FPGA pipelines can sample the full 90% of delivered luminosity, significantly enhancing heavy-flavor measurements and offering a transferable framework for future EIC DIS-electron tagging and beyond.

Abstract

This R\&D project, initiated by the DOE Nuclear Physics AI-Machine Learning initiative in 2022, leverages AI to address data processing challenges in high-energy nuclear experiments (RHIC, LHC, and future EIC). Our focus is on developing a demonstrator for real-time processing of high-rate data streams from sPHENIX experiment tracking detectors. The limitations of a 15 kHz maximum trigger rate imposed by the calorimeters can be negated by intelligent use of streaming technology in the tracking system. The approach efficiently identifies low momentum rare heavy flavor events in high-rate p+p collisions (3MHz), using Graph Neural Network (GNN) and High Level Synthesis for Machine Learning (hls4ml). Success at sPHENIX promises immediate benefits, minimizing resources and accelerating the heavy-flavor measurements. The approach is transferable to other fields. For the EIC, we develop a DIS-electron tagger using Artificial Intelligence - Machine Learning (AI-ML) algorithms for real-time identification, showcasing the transformative potential of AI and FPGA technologies in high-energy nuclear and particle experiments real-time data processing pipelines.