Add array example

This program is an example to see how to use the S_GPU library.

1. Make the program

cp Makefile.standalone Makefile
set your own environment variable in Makefile (cuda path, s_gpu path)
make

2. Set the S_GPU environment

Cf documentation to learn how to build the GPU.config file

3. Run the program

The matrices height must be a multiple of the number of MPI tasks - 1. You can change the matrices size in add_array.c (MATRIX_H and MATRIX_W). By default, MATRIX_H is set to 320. The program works well with 6 MPI tasks:

mpirun -n 6 add_array
