Multiscale Modeling meets AI

Materials research has been in the era of data-driven approaches over the past two decades. The need to use AI and ML to aid in multiscale modeling and simulation becomes indispensable. From training molecular force fields to running simulation to predicting physical properties to design new molecules, AI-powered tools are now part of these multidisciplinary studies. Training physics-informed data models is crucial to ensure that the models and predictions are physically meaningful, and can be tested. Importantly, since AI/ML projects are becoming more and more computationally demanding, to be able to invest wisely in infrastructure, you need to understand how the key components of a popular computing system work together in each and every stage of a AI-driven workflow from data processing, model training and inference. At Codice, we help researchers find optimal solutions to using AI tools in materials modeling problems.