AI creates quasi-atoms: A new approach to materials study

Understanding how a crack grows in metal requires simultaneously calculating the behavior of hundreds of billions of atoms: At the crack tip, where atomic bonds break, and in the surrounding volume of material, where the stress, driving the crack propagation, is distributed. For modern supercomputers, an accurate modeling of this process remains beyond reach. Researchers from the Skoltech Artificial Intelligence Center have proposed a solution, now published in the journal Computer Physics Communications.