In C++:

Exercise 1. [ 8 points] Write either a single program or two separate C/C++ programs that use(s) MPI blocking and non-blocking commands MPI_Send, MPI_Rcvd, MPI_Isend and MPI_Ircvd to exchange one double value between process with rank 0 and process with rank 1. Calculate the execution time using MPI_Wtime to compute the execution time and write it down in the table below as follows:

	Execution Time
A single transmission using blocking communication
A single transmission using non blocking communication
Two transmissions (round trip) using blocking communication
Two transmissions (round trip) using non blocking communication

Exercise 2. [ 4 points] Modify the ring example given in class to calculate the execution time using MPI_Wtime of the transmission of the value 5 from process with rank 0 to process with rank 1, etc. until the value 5 is received back at the process with rank 0. Use only blocking communication. Launch the execution of the program with a varied number of parallel processes (mpirun -n 10 ./a.out to launch the executable a.out for 10 processes) and write down the execution time of the ring example as follows:

Ring Example:

int token;

if (rank != 0) {

MPI_Recv(&token, 1, MPI_INT, (rank - 1) % size, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);

printf("Process %d received token %d from process %d ", rank, token, (rank - 1) % size);

} else {

// Set the token's value if you are process 0

token = -1;

}

MPI_Send(&token, 1, MPI_INT, (rank + 1) % size, 0, MPI_COMM_WORLD);

// Now process 0 can receive from the last process.

if (rank == 0) {

MPI_Recv(&token, 1, MPI_INT, size - 1, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE);

printf("Process %d received token %d from process %d ", rank, token, size - 1);

}

	Execution Time
Ring with 4 nodes
Ring with 8 nodes
Ring with 10 nodes
Ring with 12 nodes