FPGA-accelerated SpeckleNN with SNL for real-time X-ray single-particle imaging

Today's article comes from the Frontiers in High Performance Computing journal. The authors are Dave et al., from the SLAC National Accelerator Laboratory at Stanford University. In this paper they show us how to port a large power-hungry ML model to a Field Programmable Gate Array (FPGA).

DOI: 10.3389/fhpcp.2025.1520151