Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) is written in C++ and it's over 20 years old. The FORTRAN codebases seem to be centered around the finite elements/difference methods and the fluid dynamics community.
My background is in QM, so I guess that's my bias showing through :)
Somewhere under http://www.archer.ac.uk, there's a summary of the time used by various codes on that UK "Tier 1" system.