Monte Carlo Particle Transport on GPU-based Systems

PI Name John Tramm
PI Institution Argonne National Laboratory
Collaborating ANL Division Computational Science (CPS)
Project Description

The goal of this project is to complete a CUDA-based port of the Monte Carlo particle transport code OpenMC and to develop and implement optimizations for it so as to maximize performance on NVIDIA GPUs. Many of these optimizations may be highly useful to other GPU-based projects that OpenMC is involved in (namely, the “Coupled Monte Carlo Neutronics and Fluid Flow Simulation of Small Modular Reactors” project at JLSE). Any novel optimizations that are developed in this effort would also be of great interest to the field so would be likely to be shared via publications.

Testbed

gpu_a100, gpu_p100, gpu_v100_smx2, gpu_v100_smx2_debug