This study modeled deposition processes using statistical ensemble and feature scale voxel methods to predict the coverage and film properties on a large-scale pattern for the first time. Certain new concepts, such as the use of probabilities to express physical and chemical phenomena occurring on the surface, interaction among voxels, and super particles for fast calculations, were introduced. Further, the model was used to analyze the experimental characteristic variation between the morphology and film density of SiN in low-temperature plasma-enhanced chemical vapor deposition using a SiH4/NH3/N2 gas mixture with different SiH4 flow rates, which has not been found in SiO2 films. The simulation results demonstrated the effect of precursor mass in the gas phase on the surface migration and morphology formation. In addition, a short residence time was required to prevent generation of large and heavy precursors for realizing good SiN coverage and film properties at the low temperature.