Skip to content

A course in using hipFORT for accelerated computing

License

Notifications You must be signed in to change notification settings

FluidNumerics/Hipfort_Course

 
 

Repository files navigation

Accelerated computing with Fortran and Hipfort

A course in using hipfort for accelerated computing

Building the examples

This repository comes with examples that you can test drive on your own GPU accelerated systems (AMD or Nvidia). To get started, you will need

  • A 2008 Fortran standard compliant Fortran compiler
  • ROCm
  • HIPFort
  • CMake (v 3.21 or greater)

After you clone this repository and navigate into its directory on your local workstation,

  1. Create a build directory
mkdir build/
  1. Run cmake, specifying the hipfc wrapper as the Fortran compiler
FC=hipfc cmake -DCMAKE_INSTALL_PREFIX=${HOME}/.local/ ../ 

Note

The CMAKE_INSTALL_PREFIX variable sets the destination for the example applications. The above example will install binaries at ${HOME}/.local/bin

  1. Build the examples
make
  1. Install the examples
make install
  1. Add the binary and library paths to your PATH and LD_LIBRARY_PATH environment variables
export PATH=$PATH:${HOME}/.local/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:${HOME}/.local/lib

About

A course in using hipFORT for accelerated computing

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 82.7%
  • Jupyter Notebook 13.6%
  • Fortran 3.0%
  • C++ 0.5%
  • CMake 0.2%
  • Shell 0.0%