ACES: Machine-Learned Potentials for Atomistic Simulation

Machine-learned interatomic potentials are rapidly emerging tools for atomistic simulation that offer first-principles accuracy at a fraction of the computational cost. In this short course, we will introduce the basic ideas behind machine-learned interatomic potentials and demonstrate how to use several modern models, including MACE, AIMNet2, and UMA/FairChem. Participants will learn how to set up these models, optimize molecular structures, evaluate energies and forces, and use them in practical atomistic simulation workflows. Hands-on demonstrations will include examples from both molecular and materials systems.

Read more at https://hprc.tamu.edu/training/mlAtomicSimulation.html