Recent advancements in machine learned potentials (MLPs) allow researchers to perform simulations that access nanosecond-scale dynamics with atomistic models that describe domain sizes of the order of several nanometers. Such simulations can now be performed with reasonable computational resources whereby the interatomic interactions are described with the accuracy of first-principles methods such as standard density functional methods or beyond.

While MLPs are the key to modeling real materials, integrating them into existing R&D workflows remains a significant challenge for many researchers. In response to this, the MedeA 3.12 multiscale simulation environment delivers a comprehensive, end-to-end machine learning workflow for materials modeling.

This webinar will share several case studies showcasing how the integrated machine-learning workflow in MedeA 3.12 can accelerate your research, streamline the adoption of MLPs, and move beyond the limitations of standard first-principles methods.