-
Module 0: sequential Execution: compile- g++ random.cpp g++ -fopenmp openmp.cpp Run the program with following command: time ./a.out
-
Module 1: openmp Execution: compile- g++ random.cpp g++ -fopenmp openmp.cpp Run the program with following command: time ./a.out Number of threads can be changed by changing the line 5 from graphDef.cpp define P N, where N is the Number of threads.
-
Module 2: mpi Execution: Edit number of processors "N" in mpiDef.h Compile- mpic++ mpi.cpp Run the program with following command: time mprun -np N -host {,} ./a.out For example: mpirun -np 3 -host 172.18.4.103,172.18.4.119,172.18.4.46 ./a.out
-
Module 3: hybrid Execution: Edit number of processors "N" in mpiDef.h Compile- mpic++ -fopenmp hybrid.cpp Run the program with following command: time mprun -np N -host {,}./a.out For example: mpirun -np 3 -host 172.18.4.103,172.18.4.119,172.18.4.46 ./a.out
-
Input format- Input.txt Each line corresponds to each source node, same representation of adjacency matrix as given in the paper. Input.txt is randomly generated.