IPARS Parallel Implementation

1. Source file: IPARSv2/parall/
2. Documentation: IPARSv2/doc/mem.txt
3. Basic principles parallel execution of IPARSv2
   • Parallel decomposition of grid
   • Encapsulated calls to
     - UPDATE (exchange values in ghost cells)
     - GLOBAL_REDUCE operations (compute scalar values across all processors)
   • Main files
     - many.df or manyf.f and manyc.c to handle UPDATE calls
     - putil.df to handle GLOBAL_REDUCE and BCAST operations
4. Load MPI compilers and library paths:
   gxl@bevo2:~> module load intel
   gxl@bevo2:~> module load mvapich
   gxl@bevo2:~> module load mkl
5. Load tecplot for visualization:
   gxl@bevo2:~> module load tecplot
   Note: tecplot is launched by ‘tec360 -mesa’