This thesis presents a unifying description of a variety of experiments on micro- and nano-scale heat transport in semiconductors like silicon or germanium. A hydrodynamic-like heat transport model is used to predict the non-diffusive thermal response of complex systems in technologically relevant situations, like the process of energy release from nanostructured heat sources towards a semiconductor substrate. The model does not use geometry-dependent or fitted parameters, but use intrinsic material properties that can be calculated from first principles. Small-size and high-frequency effects are captured through the use of specific boundary conditions, thus resulting in a practical tool for complex microelectronic device design. Since hydrodynamic modeling is not the state-of-the-art approach to describe standard semiconductors like silicon, special care is devoted to quantitatively determine the applicability of the model, and multiple experiments using different techniques are considered and studied in a unifying way. As a result, previously unreported non-Fourier phenomena in materials like silicon or germanium is identified and demonstrated (e.g. second sound in rapidly varying thermal fields or multiple decay times characterizing the evolution of nano-structured heaters). Furthermore, the hydrodynamic description is compared with alternative modern frameworks describing size and frequency effects in semiconductor heat transport, and a summarized overview of the theoretical background, namely non-equilibrium thermodynamics and kinetic theory, is presented. In light of the extensive experimental evidence provided, this thesis demonstrate the predictive capability of hydrodynamic-like thermal transport modeling in semiconductors within a certain range of applicability that is well beyond the diffusive regime.