I encountered the QGSJET-II model which we used for modeling cosmic ray showers in the atmosphere. At one point I was asked about finding a way to parallelize the code, but the 17k+ line fortran file for the model, which I recall also included self modifying code, was too deep for an undergrad to penetrate.

https://gitlab.iap.kit.edu/AirShowerPhysics/crmc/-/blob/mast...