In an internal combustion engine, the residual energy remaining after combustion in the exhaust gasses can be partially recovered by a downstream arranged device. The exhaust port represents the passage guiding the exhaust gasses from the combustion chamber to the energy recovering device, e.g. a turbocharger. Thus, energy losses in the course of transmission shall be reduced as much as possible. However, in one-dimensional engine models used for engine design, the exhaust port is reduced to its discharge coefficient, which is commonly measured under constant inflow conditions neglecting engine-like flow pulsation. In this present study, the influence of different boundary conditions on the energy losses and flow development during the exhaust stroke are analyzed numerically regarding two cases, i.e. using simple constant and pulsating boundary conditions. The compressible flow in an exhaust port geometry of a truck engine is investigated using three-dimensional Large Eddy Simulations (LES). The results contrast the importance of applying engine-like boundary conditions in order to estimate accurately the flow induced losses and the discharge coefficient of the exhaust port. The instantaneous flow field alters significantly when pulsating boundary conditions are applied. Thus, the induced losses by the unsteady flow motion and the secondary flow motion are increased with inflow pulsations. The discharge coefficient decreased about 2% with flow pulsation. A modal flow decomposition method, i.e. Proper Orthogonal Decomposition (POD), is used to analyze the coherent structures induced with the particular inflow and outflow conditions. The differences in the flow field for different boundary conditions suggest to incorporate a modeling parameter accounting for the quality of the flow at the turbocharger turbine inlet in one-dimensional simulations.