Instructions:

To compile code run

make all

To clean compiled code run

make clean

To generate test data run (Note this can take up to 5 minutes)

make data

To delete generated test data run

make clean-data

To run the sequential sort run

./bin/seq_sort <input_file>

e.g

./bin/seq_sort random_1048576

To run the mpi sort run

mpiexec -n <num_processes> bin/mpi_sort <input_file>

To run the mp sort run

./bin/mp_sort <num_threads> <input_file>

To run the hybrid sort run

mpiexec -n <num_processes> bin/hybrid_sort <num_threads> <input_file>

You can run

make test

to compile the code and run the code on an array of size 1 million. This will use

• 8 threads for OpenMP
• 8 processes for MPI
• 4 processes and 2 threads for hybrid

You can use the create_sbatch_script.py to create sbatch scripts to run the code on the hpc.

On the HPC:

Make sure to edit the .sh files for your HPC environment. That is change the user name and path.

To make data do

sbatch make_data.sh

To compile code do

sbatch compile.sh