Programming Environment

GNU Compiler Collection (GCC) version 8.4.1 is in the default user environment. Version 12.2.0 is also available; load version 12.2.0 with the command:

module load gcc/12.2.0

Compiler Commands

Serial

To build (compile and link) a serial program in Fortran, C, and C++ enter:

Compiling Serial Code With GCC
GCC	Build Commands
Fortran 77	`gfortran myprog.f`
Fortran 90	`gfortran myprog.f900`
C	`gcc myprog.c`
C++	`g++ myprog.cc`

MPI

(coming soon)

To build (compile and link) a MPI program in Fortran, C, and C++:

Compiling With OpenMPI
MPI Implementation	modulefile for MPI/Compiler	Build Commands
Open MPI	openmpi/4.1.4-gcc	`Fortran 77: mpif77 myprog.f` `Fortran 90: mpif90 myprog.f90` `C: mpicc myprog.c` `C++: mpicxx myprog.cc`

OpenMP

To build an OpenMP program, use the -fopenmp option:

Compiling OpenMP Code
GCC	Build Commands
Fortran 77	`gfortran -fopenmp myprog.f`
Fortran 90	`gfortran -fopenmp myprog.f900`
C	`gcc -fopenmp myprog.c`
C++	`g++ -fopenmp myprog.cc`

Hybrid MPI/OpenMP

(coming soon)

To build an MPI/OpenMP hybrid program, use the -fopenmp option with the MPI compiling commands:

Compiling Hybrid MPI/OpenMP Code
MPI Implementation	modulefile for MPI/Compiler	Build Commands
Open MPI	openmpi/4.1.4-gcc	`Fortran 77: mpif77 -fopenmp myprog.f` `Fortran 90: mpif90 -fopenmp myprog.f90` `C: mpicc -fopenmp myprog.c` `C++: mpicxx -fopenmp myprog.cc`

CUDA

NVIDIA GPUs are available as part of the Nightingale compute cluster. CUDA is a parallel computing platform and programming model from NVIDIA for use on the GPUs. These GPUs support CUDA compute capability 2.0.

Load the CUDA Toolkit into your environment using the following module command:

module load cuda/11.4.2