Several studies have been conducted on hypervapotron (HV) cooling channels for the application of the fusion tokamak cooling system; however, most of them have focused on the critical heat flux (CHF). To predict the thermal efficiency of the tokamak system of a DEMO or commercial plant, it is necessary to predict the heat transfer performance of the HV. Therefore, in this study, the heat transfer characteristics of a one-side heated hypervapotron cooling channel under subcooled flow conditions were analysed. The subcooled flow boiling experiment was conducted as a method of gradually increasing the applied heat, and the heat transfer performance in all flow regimes from the single-phase regime to the CHF was measured. In subcooled flow boiling, it was found that three flow regimes exist: the single-phase (SP), partially developed nucleate boiling (PDB), and fully developed nucleate boiling (FDB) regimes. Additionally, we analysed different heat transfer mechanisms for each flow regime. The effects of system parameters such as the system pressure, sub-cooling, and mass flow rate on heat transfer were analysed, and the prediction performance of existing heat transfer correlations developed for application to the three flow regimes was evaluated. However, it was difficult to find a correlation reflecting the heat transfer performance enhanced by the fin structure of the HV, and the authors of this study developed heat transfer correlations using Python code and an artificial intelligence regression method