This project uses artificial intelligence together with physics-based computer simulations to speed up the discovery of new materials. By learning from large collections of experimental and simulation data, our models will both predict important properties and suggest new compositions or structures for polymers, alloys, pharmaceuticals, glasses, and photonic or plasmonic materials. We also design advanced materials for future AI hardware to create a two-way link: AI helps discover better materials, and better materials enable more powerful and energy-efficient AI systems.