Robust optimization of the ICF design on El Capitan - the world's fastest supercomputer

During the summer of 2025, the computing power of El Capitan, the world's largest supercomputer, will be devoted to find an optimal design for the indirect drive inertial confinement fusion experiments. The results of the optimization process and the initial analysis of the resulting designs will be discussed in this presentation, as well as the analysis of the optimization performance and scaling. We will show how this automated search explores the input parameter space, which is extended towards higher dimensions owing to the exascale supercomputer executing a large number of computationally intensive integrated radiation-hydrodynamics ICF simulations. The optimization process is linked to the experimental inputs through the laser-hohlraum simulations, while the robustness metric that penalizes the sensitivity to the uncontrolled input variations is leveraging ensembles of 3D capsule simulations. The variability of the inputs is determined, in part, by the previous studies of the high-performing Hybrid-E experiments at the National Ignition Facility. The initial objective of the optimization will be to identify new designs, or confirm those already established, for the 1.9 MJ laser energy. LLNL-ABS-2009020