Issues on prefetch function

["Jim Jia" <mydpdk-KN7UnAbNpbg@public.gmane.org>]

Hello, everyone
         I'm doing an experiment to test the performance of rte_prefetch0. I find the speed of that function is quit slow.  When executing the following code, tit will take about 8,000,000 cycles (about 45 cycles each loop ). What's wrong with it?

      Thanks in advance.
    

 struct A
{
  char a[64];
};

struct A foo[180000];


int
main ()
{
  int i, j;

  long long a = rte_rdtsc ();
  for (j = 0; j < 180000; j++)
    {
      rte_prefetcht0 (&foo[j]);
    }
  long long b = rte_rdtsc ();
  printf ("%lld\n", b - a);
  return 0;
}