 Deep learning with E3, Equivariant Neural Networks for AB Inicio Electronic Structure Calculation is a new approach to solving the problem of efficiently calculating the electronic structure of materials using density functional theory, DFT. This method uses a neural network to learn from DFT data of small sized structures, allowing it to accurately predict the electronic structure of larger structures like supercells. It achieves this by preserving the Euclidean symmetry of the system while accounting for spin orbit coupling. This method is able to achieve sub-mivvy prediction accuracy at high training efficiency, creating opportunities for materials research, such as building a Muat twisted material database. This article was authored by Xiaoshu and Gong, He Li, Nian Longzu, and others.