1)Your application details.
code fragment such as below.
void __attribute__((noinline)) MatrixMulOpenMpGpuOffloading ()
{
int i, j, k;
// a is identity matrix
for (i=0; i<M; i++)
for (j=0; j<N; j++)
a[i][j] = 1.0;
// each column of b is the sequence 1,2,...,N
for (i=0; i<N; i++)
for (j=0; j<P; j++)
b[i][j] = i+1.;
// c is initialized to zero
for (i=0; i<M; i++)
for (j=0; j<P; j++)
c[i][j] = 0;
auto t1 = chrono::steady_clock::now();
for (int i = 0 ; i < 1000; i++)
{
// Parallelize on target device
#pragma omp target teams distribute parallel for simd map(to: a, b) map(tofrom: c) //thread_limit(128)
{
for (i=0; i<M; i++) {
// Each element of the product is just the sum 1+2+...+n
for (j=0; j<P; j++) {
for (k=0; k<N; k++) {
c[i][j] += a[i][k] * b[k][j];
}
}
}
}
}
auto t2 = chrono::steady_clock::now();
cout << "matmul default takes :" << chrono::duration_cast<chrono::microseconds>(t2 - t1).count() / 1000 << "us" << std::endl;
}
2)Commands used.
vtune -collect gpu-hotspot
3)How the results are being viewed(CLI or GUI)
GUI VTUNE
