performance test for dgemv

test_dgemv.c

#include <stdio.h>
#include <stdlib.h>
#include <sys/time.h>

//#define ITERS 1000000
void dgemv_(char *, int*, int *, double*, double*, int*, double*, int*, double*, double*, int*);
int main(int argc, char * argv[])
{

  int m, n;
  double alpha=1.1;
  double beta=1.0;
  double * a, *x, *y;
  int inc=1;
  int i;
  int iters;
  char trans='N';
  struct timeval start,finish;
  double duration;

  if(argc<4){printf("input error\n"); return 0;}

  m=atoi(argv[1]);
  n=atoi(argv[2]);
  iters=atoi(argv[3]);

  a=(double*)malloc(m*n*sizeof(double));
  x=(double*)malloc(n*sizeof(double));
  y=(double*)malloc(m*sizeof(double));

  for(i=0; i<m*n; i++){
    a[i]=1.1;
  }

  for(i=0; i<n; i++){
    x[i]=1.0;
  }

  for(i=0; i<m; i++){
    y[i]=1.0;
  }

  gettimeofday(&start, NULL);
  for(i=0; i<iters;i++){

    dgemv_(&trans, &m, &n, &alpha, a, &m, x, &inc, &beta, y, &inc);
  }
  gettimeofday(&finish, NULL);

  duration = ((double)(finish.tv_sec-start.tv_sec)*1000000 + (double)(finish.tv_usec-start.tv_usec)) / 1000000;
  double gflops = 2.0 * m *n*iters;
  gflops = gflops/duration*1.0e-6;

  printf("%dx%d\t%d loops\t%lf s\t%lf MGFLOPS\n",m, n, iters, duration, gflops);

  free(a);
  free(x);
  free(y);
  return 0;
}

Hi, I follow this https://gist.github.com/xianyi/65aef3c2e5bc32049806 to come here.
I don't know how could get higher performance than 6.9 GFLOPS.
My CPU is i7-4790 and I tested dgemv with m=6400, n=64 but the highest performance was only 6.9GFLOPS when using 4 or 8 threads.