Introduction to Fortran

Fortran (Formula Translation) is the oldest high-level programming language still in active use.

1.1 What is Fortran?

Fortran (Formula Translation) is the oldest high-level programming language still in active use. It is a compiled, statically typed, imperative language designed for numerical and scientific computing. Modern Fortran (90 onwards) is a clean, powerful language with:

First-class multidimensional arrays and whole-array operations.

Excellent performance — compilers produce highly optimized, vectorized code.

Modules, derived types, and object-oriented features.

Built-in parallelism: do concurrent, coarrays; plus OpenMP, MPI, and GPU offloading.

Standard C interoperability.

Backward compatibility with decades of tested scientific code.

1.2 History

Standard / Milestone FORTRAN (IBM, John Backus) first optimizing compiler subroutines, functions first ANSI standard FORTRAN 77 block IF, CHARACTER type Fortran 90 free form, modules, array syntax, allocatables, derived types, pointers, recursion

forall, pure/elemental, minor fixes Fortran 2003 OOP, C interoperability, IEEE arithmetic, stream I/O Fortran 2008 coarrays, submodules, do concurrent, block teams, collective subroutines, improved C interop conditional expressions, enumeration type, typeof/classof, more

1.3 Where is Fortran used?

Weather & climate (ECMWF IFS, WRF, CESM, ICON, MOM6) Computational fluid dynamics (OpenFOAM's predecessors, NASA codes, FUN3D)

Quantum chemistry & materials (Gaussian, VASP, Quantum ESPRESSO, CP2K, NWChem)

Physics & astrophysics (nuclear simulations, lattice QCD, MESA stellar evolution)

Structural engineering (finite element codes, NASTRAN, Abaqus internals)

Numerical libraries (BLAS, LAPACK, ARPACK, MINPACK, QUADPACK) — used under NumPy, SciPy, R, MATLAB, Julia

HPC benchmarks (LINPACK/HPL ranks the TOP500 supercomputers)

1.4 Why learn Fortran?

Maintain or extend critical scientific codes.

Write extremely fast numerical kernels with simple syntax.

Understand the numerical libraries underlying every data-science stack.

HPC careers (national labs, weather centers, aerospace, energy).