Jet noise is still a distinct noise component when a commercial aircraft is taking off. A parallel high-fidelity simulation framework for industrial jet noise prediction is presented in this paper. This framework includes complex geometry meshing and Ffowcs Williams-Hawkings (FW-H) surface placement during preprocessing, a parallel hybrid RANS-LES flow solver coupled with an FW-H acoustic solver in the simulation and mean and unsteady data processing after the simulation. The use of this framework is demonstrated through two jet noise prediction cases: in-flight heated jets and installed ultra-high bypass-ratio (UHBPR) engines. These simulations can provide more insight than experimental tests into jet flow physics for engineering model improvement. Additional advantages are also shown in the cost and turn-around time. Thus there is great potential for high-fidelity jet noise simulations to partly replace rig tests for industrial use in the future.