Developing new materials with specific properties requires extensive and time-consuming atomic-scale simulations. Traditional methods for these simulations often lack the accuracy needed to model complex reactions in bulk materials and at interfaces, hindering the discovery of novel materials.

Physics Inverted Materials is developing PHIN-atomic, a software tool that leverages machine learning to generate highly accurate interatomic potentials. These potentials are then used in molecular dynamics simulations, significantly accelerating the process of materials discovery and development. By providing a more precise and efficient simulation environment, PHIN-atomic enables researchers to explore a wider range of material compositions and reaction pathways, leading to the faster identification of materials with desired characteristics.

Machine learning-driven generation of interatomic potentials

High-accuracy molecular dynamics simulations

Accelerated atomic-scale simulations for materials development

Capability to model reactions in bulk materials and at interfaces

The primary users are materials scientists, engineers, and researchers in academia and industry who require accurate and efficient atomic-scale simulations for materials discovery and development.