Semiconductor devices are scaled down to the level which constituent materials are no longer considered continuous. To account for atomistic disorder, process variations and quantum mechanical effects, an atomistic modeling approach needs to be pursued. We developed NEMO 3D-peta is a massively parallel Schroedinger equation solver based on semi-empirical atomistic tight-binding approach (TB). With 10~20 of bases per atom, TB parameters are fit to reconstruct bandstructure traits of crystalline solids, which significantly reduces the computational burden compared to ab-initio calculations. With the support of increasing supercomputer infrastructures, it is possible to compute the eigenstates of million atom quantum dots as well as quantum transport of nano-transistors. As a modeling example, a STM controlled impurity doped system in Silicon for solid-state quantum computing application is presented. Finally, limitations of TB and bridging gap between TB and ab-initio approaches will be discussed.