Speaker
Zepeng Li
Description
cgra4ml is a highly flexible, high-performance accelerator system that helps researchers build, train, and implement machine learning models on Field Programmable Gate Arrays (FPGAs). It extends the capabilities of HLS4ML by allowing off-chip data storage and supporting a broader range of neural network architectures, including models like ResNet and PointNet. Using this new framework, we developed a new pipeline to reconstruct each event’s position and energy by implementing a machine learning model, PointNet, on an FPGA.
This marks one of the first instances of applying hardware-AI co-design in the context of neutrino-less double beta decay experiments.
