/ParaSolve_CUDA

Primary LanguageFORTRANOtherNOASSERTION

Fortran/C based Hybrid MPI-CUDA Computational Physics Solver

Execution instructions

  1. Download the source code

  2. Make sure you have make utility and the latest version of GNU and MPI installed

  3. Edit the Makefile to include your MPI path.

       make
       mpirun -n [number_of_procs] ./Solver 
    
  4. Note that the number of processes in each direction is HK defined in the header file Solver.h. And for 2D cases HD = 2. Therefore, the total number of processes will be HK^HD.

  5. To plot results, edit the python file (plot.py) to match your grid size and simply type (make sure k = HK and d = HD)

       python plot.py
    

Author - Akash V. Dhruv
License - Refer LICENSE.md