As new multi-modal nanoscale capabilities accelerated the discovery of new fundamental concepts and phenomena in materials science across multiple length and time scales, there is a significant amount of complex multidimensional new data being generated as these tools are developed. This increasingly complex data requires new comprehensive digital approaches combining AI/ML and physics-based modeling to analyze multi-dimensional correlations and enable new scientific discoveries as well as improved capabilities for autonomous experiments. Exabyte.io proposes to develop a new capability in our current open-source, cross-platform software framework to bridge digital approaches across multiple lengths and timescales for selected materials systems. We deploy scalable, python-based workflows that employ first-principles and/or atomistic-MD simulation packages that scale well on DOE supercomputers (e.g., VASP, Quantum Espresso, GROMACS, etc.) and provide atomistic calculations of 2d or 3d structural models to develop data-driven AI/ML models. Our approach enables an accessible (âblack boxâ) usage pattern so that the codes are easily accessible even with a rudimentary understanding of the different methods underlying the software. We will build an open-source, python-based workflow design engine to make it accessible to the community at large. We will focus on NWChem(Ex) and LAMMPS because of their ability to study vdW and molecular solids, as well as heterostructures formed from - a major unique addition to our existing platform. Our aim is to focus on the workflow engine as the key component and reuse other parts of our existing infrastructure to demonstrate the proof-of-concept. The integration of this open-source multi-scale modeling with our platform improves the accessibility, speed, and efficiency of materials research and development (R&D) and enables the development of new kinds of products in industrial sectors. Computational materials design drives the transformation of industrial materials R&D and will become a multi-billion market by 20
