Available Software

Efforts have been made to ensure that the versions of the available software are identical on both clusters but, at times, there will be differences.

Base Modules

ModuleDescription
cmake/3.20.4Portable Build Process Manager – cmake.org
This tool is useful if you want to be able to build your software on Linux, Mac or Windows.
gcc/8.4.0The GNU GCC suite of C, C++ and Fortran compilers – gcc.gnu.org
gcc/10.2.0 (D)The GNU GCC suite of C, C++ and Fortran compilers – gcc.gnu.org
git/2.32.0 (D)Git – Distributed Version Control System – git-scm.com
intel/2021.2The Intel Compiler suite of C, C++ and Fortran compilers – software.intel.com
matlab/r2020aMatlab R2020a – uk.mathworks.com
python/3.8.5Python v3.8.5 – python.org
python/3.9.5 (D)Python v3.9.5 – python.org
Base Software Modules

In general, the Intel compilers will give you slightly better performance (5-10% faster). However, the GNU suite of compilers tend to insist your software more strictly adheres to relevant language standards. This means if you want to ensure your software will compile on any other Linux platform then choose the GNU compilers.

During development, it is suggested to use the GNU compilers as they will identify more potential issues with your software. In production however, if you want unltimate performance, then the Intel compilers will likely produce quicker run-times.

Compiler based Modules

These modules all require one of the compiler suites to be already loaded. The compiler suites that are compatible are listed in the table below.

ModuleCompatible
Compilers
Description
boost/1.76.0GCC (all)
Intel v16 (enhpc)
Boost C++ libraries – www.boost.org
mvapich2/2.3.6GCC (all)
Intel (all)
The MPICH implementation of MPI – mvapich.cse.ohio-state.edu
openblas/0.3.12GCC (all)
Intel (all)
The BLAS libraries – www.openblas.net
openmpi/4.1.1GCC (all)
Intel (all)
The OpenMPI implementation of MPI – www.open-mpi.org
valgrind/3.17.0GCC (all)
Intel v16 (enhpc)
Debugging and profiling tools – valgrind.org

MPI based Modules

These modules all require a compiler suite and an MPI library to already be loaded.

ModuleCompatible
Compilers / MPI
Description
parmetis/4.0.3GCC (all)
Intel (all)
MPI (all)
Parallel Graph Partitioning Library – glaros.dtc.umn.edu