https://doi.org/10.1051/epjn/2024034
Regular Article
Status of GPU capabilities within the Shift Monte Carlo radiation transport code*
Oak Ridge National Laboratory, 1 Bethel Valley Road, Oak Ridge, TN, 37830, USA
* e-mail: veb@ornl.gov
Received:
12
June
2024
Received in final form:
18
October
2024
Accepted:
2
December
2024
Published online: 31 January 2025
Shift is a general-purpose Monte Carlo (MC) radiation transport code for fission, fusion, and national security applications. Shift has been adapted to efficiently run on GPUs in order to leverage leadership-class supercomputers. This work presents Shift’s current GPU capabilities. These include core radiation transport capabilities for eigenvalue and fixed-source simulations, and support for non-uniform domain decomposition, Doppler broadening, free-gas elastic scattering, general-purpose geometry, hybrid MC/deterministic transport, and depletion. Transport results demonstrate a 2–5× GPU-to-CPU speedup on a per-node basis for an eigenvalue problem on the Frontier supercomputer and a 28× speedup for a fixed-source problem on the Summit supercomputer.
This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. DOE will provide access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).
© E. Biondo et al., Published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.