Thermal rectification is an asymmetric heat transport phenomenon where thermal conductance changes depending on the temperature gradient direction. The experimentally reported efficiency of thermal rectification materials and devices, which are applicable for a wide range of temperatures, is relatively low. Here we report a giant thermal rectification efficiency of 218% by maximizing asymmetry in parameters of the Stefan-Boltzmann law for highly non-linear thermal radiation. The asymmetry in emissivity is realized by sputter-depositing manganese (ε = ∼0.38) on the top right half surface of a polyurethane specimen (ε = ∼0.98). The surface area of the polyurethane side is also dramatically increased (1302%) by 3D printing to realize asymmetry in geometry. There is an excellent agreement between the experimentally measured temperature profiles and finite element simulation results, demonstrating the reliability of the analysis. Machine learning analysis reveals that the surface area is a dominant factor for thermal rectification and suggests novel light-weight designs with high efficiencies. This work may find applications in energy efficient thermal rectification management of electronic devices and housings.