int main(int argc, char **argv)
{
int i;
float *a, *b;
double t;
a = (float *)_mm_malloc(sizeof(float) * N, 16);
b = (float *)_mm_malloc(sizeof(float) * N, 16);
for (i = 0; i < N; i++) {
a[i] = 1.0;
}
t = hpctimer_getwtime();
for (i = 0; i < NREPS; i++) {
fun_def(a, b, N);
// fun_sse(a, b, N);
}
t = hpctimer_getwtime() - t;
t = t / NREPS;
//print_vec(b, N);
printf("Elapsed time: %.6f sec.\n", t);
_mm_free(a);
_mm_free(b);
return 0;
}
int main(void)
{
int i, size;
struct Foo *Bar;
struct Foo *Bar1;
double t;
size = sizeof (struct Foo) * TN;
if ( (Bar = malloc(size)) == NULL) {
fprintf(stderr, "No enough memory\n");
exit(EXIT_FAILURE);
}
if ( (Bar1 = malloc(size)) == NULL) {
fprintf(stderr, "No enough memory\n");
exit(EXIT_FAILURE);
}
memset (Bar, 0 ,sizeof (struct Foo));
memset (Bar1, 0 ,sizeof (struct Foo));
t = hpctimer_getwtime();
int j;
for (i = 0; i < TN; i++)
{
for (j = 0; j < T_SIZE; j++)
{
Bar1[i].arr[j] = Bar[i].arr[j];
Bar1[i].arr1[j] = Bar[i].arr1[j];
Bar1[i].arr2[j] = Bar[i].arr2[j];
Bar1[i].arr3[j] = Bar[i].arr3[j];
Bar1[i].arr4[j] = Bar[i].arr4[j];
Bar1[i].arr5[j] = Bar[i].arr5[j];
}
}
t = hpctimer_getwtime() - t;
printf("T_SIZE: %d\nElapsed time (sec.): %.6f\n", T_SIZE, t);
free(Bar);
free(Bar1);
return 0;
}
int main()
{
int i;
float *a;
double t, sum;
a = (float *)_mm_malloc(sizeof(*a) * N, 16);
for (i = 0; i < N; i++)
a[i] = 1.0;
t = hpctimer_getwtime();
for (i = 0; i < NREPS; i++) {
sum = reduction_sum(a, N);
// sum = reduction_sum_sse(a, N);
}
t = (hpctimer_getwtime() - t) / NREPS;
printf("Reduction sum: %.4f (real %.4f)\n", sum, (float)N);
printf("Elapsed time: %.6f sec.\n", t);
_mm_free(a);
return 0;
}
int main(int argc, char **argv)
{
double PI25DT = 3.141592653589793238462643;
int i, nsteps;
double pi, x, step, sum;
double t;
int nthreads = 1;
nsteps = (argc > 1) ? atoi(argv[1]) : 1000000;
step = 1.0 / (double)nsteps;
t = hpctimer_getwtime();
sum = 0.0;
#pragma omp parallel shared(sum)
{
#pragma omp master
#ifdef _OPENMP
nthreads = omp_get_num_threads();
#endif
#pragma omp for private(x)
for (i = 1; i <= nsteps; i++) {
x = (i - 0.5) * step;
#pragma omp critical
sum = sum + 4.0 / (1.0 + x * x);
}
}
pi = step * sum;
t = hpctimer_getwtime() - t;
printf("PI is approximately %.16f, Error is %.16f\n",
pi, fabs(pi - PI25DT));
printf("(nsteps = %d, step = %f, nthreads = %d)\n", nsteps, step, nthreads);
printf("Elapsed time = %.6f sec.\n", t);
return EXIT_SUCCESS;
}
int main(int argc, char **argv)
{
int i;
double t;
init_matrix(A, B, C, N);
t = hpctimer_getwtime();
for (i = 0; i < NREPS; i++) {
dgemm_def(A, B, C, N);
//dgemm_transpose(A, B, C, N);
//dgemm_transpose2(A, B, C, N);
//dgemm_block(A, B, C, N);
}
t = hpctimer_getwtime() - t;
t = t / NREPS;
/*print_matrix(C, N);*/
printf("Elapsed time: %.6f sec.\n", t);
return 0;
}
