Modeling and simulation of feature scale profiles, including damage distributions and film properties in dry etching (continuous wave, pulse, cycle, and atomic layer etching) and deposition processes (chemical vapor, physical vapor, and atomic layer deposition) using string, shock tracking, level-set, and cell removable (or voxel) methods, are useful for predicting process properties and gain insights into the variation mechanisms for realizing high performance of advanced complementary metal-oxide semiconductor devices and have been comprehensively reviewed in this work. The future perspective has also been discussed from the viewpoint of a fusion physical model with machine learning using real-time monitoring of the manufacturing equipment, so-called process informatics. Furthermore, after the introduction of quantum computing for process simulations on atomic scales, the epoch of full digital twins might be realized.