CUDA

for (i=0; i

q[i] = 0;

for (j=0; j

s[j] = s[j] + r[i] * A[i][j];

q[i] = q[i] + A[i][j] * p[j];

}

}

Recall that one approach to parallelizing this code is to parallelize the iterations of the i loop, and protect updates to s[j] across threads with atomic operations.

(a) Provide a CUDA kernel (thread program only) for the parallelized code.