15–18 Dec 2026
Duke University
America/New_York timezone
Artificial Intelligence for the Electron Ion Collider (2026)

hls4ml - and open-source tool for deploying machine learning models on FPGAs and ASICs and its application for particle detectors

Not scheduled
20m
LSRC B101 Love Auditorium (Duke University)

LSRC B101 Love Auditorium

Duke University

Durham, NC 27710
AI/ML in Production and Distributed ML AI/ML in Production, Distributed ML

Speaker

Jan-Frederik Schulte (Purdue University)

Description

The tight integration of machine learning (ML) models into detector readout and trigger systems will allow future detectors to move complex reconstruction tasks much closer to the detector. This will enable these detectors to cope with much higher data rates by performing more complex and better targeted event selection or data compression in DAQ and trigger systems. ML algorithms have to be designed from the beginning with the tight latency and computing resource constraints in mind and the integration of ML models into FPGAs or ASICs requires significant technical know-how. We present hls4ml, an open-source tool that translates ML models from standard tools like keras or pytorch into high level synthesis (HLS) code for integration into either FPGA or ASICs. hls4ml supports a large variety of common ML operations, recently adding support for certain transformer architectures. It generates HLS code targeting a variety of vendors, such as Xilinx/AMD, Intel/Altera, or Siemens EDA. Hardware-software co-design is enabled with native solutions, such as pruning techniques that specifically optimize the model for FPGA resource usage, or thanks to the tight integration of hls4ml into a wider eco-system of model pruning and quantization tools. These include established tools such as QKeras or QONNX, but also many advanced open-source tools including PQuant-ML and HGQ2, support for which have recently been added to hsl4ml. hls4ml is already widely used in the HEP community to integrate ML models into existing trigger systems and its capabilities in accelerating ASIC design for detector readout has been demonstrated. In this talk we also demonstrate the applicability of these techniques to the EIC environment using an extension to hls4ml to support a transformer model for track reconstruction in simulated ePIC detector data.

Author

Jan-Frederik Schulte (Purdue University)

Presentation materials

There are no materials yet.