The Erich Schmid Institute of Materials Science (Austrian Academy of Sciences) was selected in the EUROfusion 9th Cycle Call (Pitagora AMD/Booster program, Cineca) with the project “Machine Learning Enhanced Atomistic Simulations of Self-Healing in Reactor Materials.” This collaborative effort, conducted jointly with Montanuniversität Leoben and the Materialsforschung Center Leoben under the lead of the Erich Schmid Institute, addresses the global challenge of irradiation-induced cracking in fusion reactor materials. 

The project employs a multiscale simulation approach that combines Molecular Dynamics (LAMMPS) with Ab Initio (VASP) calculations. The allocation of 5120000 CPU core hours and 7680000 equivalent core GPU hours on the Pitagora supercomputer (Cineca) is essential to its success: CPU resources enable the generation of extensive databases from ab initio and molecular dynamics simulations, while GPU resources allow efficient simulations with complex interatomic potentials that require large-scale parallelization. In addition, the GPU allocation supports the development of machine learning pipelines designed to predict irradiation damage and the potential for self-healing in advanced reactor materials.