Description
1. (10 pts.) Implement a “hello-world” program using OpenMP. Your program should
create as many threads as their are processors/cores on your current machine. From
each thread that you create, your program should print:
Hello from thread i.
This is ’s first parallel program
There are currently n threads running
Implement your program so that the text given above is printed in one block for each
thread.
2. (10 pts.) Consider the following problem. Given an n×n matrix of numbers, an integer
m, and a modulus q, find the largest product of m modulo q numbers in a row, either
vertically, horizontally, or diagonally. Write a program that outputs the value of the
largest product on a line by itself. For example, consider the following 5 × 5 matrix of
values with the value m = 4 and q = 1000000 :
10 1 1 1 1
1 10 11 12 13
1 1 10 11 12
1 1 1 10 1
1 1 1 1 10
Your program should produce the answer “17160” (10 × 11 × 12 × 13). When m = 5,
your program should answer “100000”
Implement your program in C++ and provide adequate documentation. Test input/output examples will be provided to help you debug your code.
3. (20 pts.) Use OpenMP to implement a parallel version of the program that you implemented in problem #2. Implement your program using #pragma omp parallel.
Compare your program’s running time to your original program. Your program should
be at least 75% efficient on a four core machine on put of a 1000 x 1000 array with
m = 500.
Implement your program in C++ and provide adequate documentation.
