OpenACC is a directive-based parallel programming model for C, C++, and Fortran.
It is most often associated with directive-based programming for GPU computers,
but has also been used for programming multi-core CPUs and FPGAs. OpenACC is
developed and maintained by members of the OpenACC standard organization.
Exascale simulation of multiphase/physics fluid dynamics
GPUFORT: S2S translation tool for CUDA Fortran and Fortran+X in the spirit of hipify
A solver for the coupled and decoupled electron and phonon Boltzmann transport equations.
POT3D: High Performance Potential Field Solver
The repository containing everything you need to compete in the IHPCSS 2019 programming challenge.
jacobi - a benchmark by solving 2D laplace equation with jacobi iterative method. GPU or Xeon Phi can be used.
Various benchmarks used to inform PSyclone optimisations
This code mimics the basic MPI+GPU tasks of PSI's codes, for use with testing multi-GPU multi-node clusters
can - a simple dense matrix-matrix multiplication benchmark with MPI/OpenMP/OpenACC. MPI version is based on Cannon's algorithm.
Mandelbrot set by MPI/OpenMP/OpenACC.
OpenACC GPU parallelization for various numerical methods and miscellaneous problems using FORTRAN
Some Fortran codes to practice programming in Fortran.
A micro MFC and CFD mini-app
Released November 2011