This paper focuses on the optimization driving strategies for GaN HEMTs (Gallium Nitride High Electron Mobility Transistors) with the aim of predicting their behavior in power electronics applications. A numerical non-linear datasheet-based model is proposed to accurately estimate switching losses. The results show improved accuracy in the estimation of power losses compared to conventional methods. Moreover, the study explores the influence of inductive coupling and parasitic inductance in the driving and power loops aided by Ansys Q3D extractor, an electromagnetic field simulator and analysis to optimize driving loop designs and ultimately take more informed decisions in the optimization process.