Nature publication
A generative model for inorganic materials design
MatterGen
MatterGen is a diffusion model specifically designed for generating stable inorganic materials across the periodic table. Crucially, the model can be fine-tuned to steer the generation towards a broad range of property constraints, including desired chemistry, symmetry as well as mechanical, electronic and magnetic properties. MatterGen reaches state-of-the-art performance in the de-novo generation of novel materials, and outperforms traditional computational methods like screening for inverse design.
MatterSim
MatterSim is a deep-learning model for accurate and efficient materials simulation and property prediction over a broad range of elements, temperatures, and pressures to enable the in silico materials design. MatterSim employs deep learning to understand atomic interactions from the very fundamental principles of quantum mechanics, across a comprehensive spectrum of elements and conditions—from 0 to 5,000 Kelvin (K), and from standard atmospheric pressure to 10,000,000 atmospheres. In our experiment, MatterSim efficiently handles simulations for a variety of materials, including metals, oxides, sulfides, halides, and their various states such as crystals, amorphous solids, and liquids. Additionally, it offers customization options for intricate prediction tasks by incorporating user-provided data.
Work with us
Check out our open roles
Check out our open roles