Nonadiabatic molecular dynamics can provide atomistic insights into photochemical and photophysical properties of solar energy and photocatalytic materials but modeling such processes for realistic nanoscale materials, comparable to experimental studies, is computationally expensive. In our recent publication, we implemented an optimized and parallelized code to study excited states dynamics in large nanoscale and periodic systems with thousands of atoms such as silicon quantum dots and 2D graphitic carbon nitride.