static double HPCC_Stream()
{
double scalar = 3.0, times[NTIMES], curGBs, mintime = FLT_MAX;
int j, k;
for(k=0;k<NTIMES;k++) {
#pragma xmp barrier
times[k] = -xmp_wtime();
#ifdef _FUJITSU
#pragma loop xfill
#pragma loop noalias
#endif
#pragma omp parallel for
for(j=0;j<array_elements;j++)
a[j] = b[j] + scalar*c[j];
#pragma xmp barrier
times[k] += xmp_wtime();
}
for(k=1;k<NTIMES;k++)
mintime = Mmin(mintime, times[k]);
curGBs = (mintime > 0.0 ? 1.0 / mintime : -1.0);
curGBs *= 1e-9 * 3 * sizeof(double) * array_elements;
#pragma xmp reduction(+:curGBs)
return curGBs;
}
int
main()
{
int i, j, k, nn;
float gosa;
double cpu, cpu0, cpu1, flop, target;
int myrank = xmp_node_num() - 1;
target = 60.0;
omega = 0.8;
imax = MIMAX;
jmax = MJMAX;
kmax = MKMAX;
/*
* Initializing matrixes
*/
initmt();
if (myrank == 0) {
printf("mimax = %d mjmax = %d mkmax = %d\n",MIMAX, MJMAX, MKMAX);
printf("imax = %d jmax = %d kmax =%d\n",imax,jmax,kmax);
}
nn= 3;
if (myrank == 0) {
printf(" Start rehearsal measurement process.\n");
printf(" Measure the performance in %d times.\n\n",nn);
}
cpu0= xmp_wtime();
gosa= jacobi(nn);
cpu1= xmp_wtime();
cpu= cpu1 - cpu0;
flop= fflop(imax,jmax,kmax);
if (myrank == 0) {
printf(" MFLOPS: %f time(s): %f %e\n\n", mflops(nn,cpu,flop),cpu,gosa);
}
nn= (int)(target/(cpu/3.0));
#pragma xmp reduction (max:nn)
if (myrank == 0) {
printf(" Now, start the actual measurement process.\n");
printf(" The loop will be excuted in %d times\n",nn);
printf(" This will take about one minute.\n");
printf(" Wait for a while\n\n");
}
/*
* Start measuring
*/
cpu0 = xmp_wtime();
gosa = jacobi(nn);
cpu1 = xmp_wtime();
cpu= cpu1 - cpu0;
if (myrank == 0) {
printf(" Loop executed for %d times\n",nn);
printf(" Gosa : %e \n",gosa);
printf(" MFLOPS measured : %f\tcpu : %f\n",mflops(nn,cpu,flop),cpu);
printf(" Score based on Pentium III 600MHz : %f\n",
mflops(nn,cpu,flop)/82.84);
}
return (0);
}
