The prediction of jet and fan noise is among the most challenging tasks for computational fluid mechanics. The noise generation mechanisms are not fully understood yet — even on a theoretical level. Prediction methods based on traditional CFD technology require use of advanced LES turbulence models, for which productive use for complex real-life engine geometries is still many years away. Therefore, propulsion noise engineers must rely almost exclusively on experimental tests, which have significant limitations. Most tests cannot be performed at installation velocities and are generally limited to stand-alone configurations of the engine only. In addition, installation noise can be tested only during late-stage flight tests, which leads to high safety margins and overdesign to avoid noise-related certification issues.