How high-performance computing gave me a new appreciation for nuclear data
Abstract: In this talk, I will discuss the need for high-fidelity nuclear reactor simulations in nuclear reactor simulations and the performance of such simulation tools on modern computing hardware. Additionally, this talk will illustrate some of the main bottlenecks in achieving better performance on these modern systems and that the answer is not always better software practices. New nuclear data models will be presented that can improve performance while also improving the fidelity of the simulation.
Figure: (left) Geometrical representation of full core pressurized water reactor BEAVRS benchmark; (right) power distribution within the core obtained with OpenMC.
Bio: Benoit Forget is the Department Head and the KEPCO Professor of the Nuclear Science and Engineering Department at the Massachusetts Institute of Technology. Upon joining MIT in 2008, he founded the Computational Reactor Physics Group (CRPG) which has produced the open source codes OpenMC and OpenMOC designed for high-fidelity, efficiency and parallel performance on leadership class computing. He has taught reactor physics, reactor engineering and radiation transport courses at both the undergraduate and graduate level and co-organized summer short courses on reactor safety and reactor engineering for young professionals and nuclear executives. Prior to joining MIT, Prof. Forget worked at the Idaho National Laboratory as a Reactor engineer where he worked on methods development and fuel cycle analysis.
Prof. Forget graduated from the Georgia Institute of Technology in 2006 with a PhD in Nuclear Engineering, and from École Polytechnique de Montréal in 2003 with a M.Eng degree in Energy Engineering and a B.Eng in Chemical Engineering. He has published more than 100 papers in leading peer-reviewed nuclear engineering journals and conferences. He has been a member of the American Nuclear Society since 2003 and served as Chair of the Reactor Physics Division in 2011-2012. He was also the recipient of the ANS John Landis Young Member Engineering Achievement award in 2013.
