int main(int argc, char* argv[])
{
int i, j;
double tstart, tstop;
double nflop;
double tmmult, tdgemm;
for( i=0; i<SIZE_M; i++ ) {
for( j=0; j<SIZE_N; j++ ) {
A[i][j]=(double)(i)+(double)(j);
}
}
for( i=0; i<SIZE_N; i++ ) {
for( j=0; j<SIZE_K; j++ ) {
B[i][j]=(double)(i)+(double)(j);
}
}
nflop = 2.0*(double)SIZE_M*(double)SIZE_N*(double)SIZE_K;
MYTIMESTAMP(tstart);
mmult( A, B, C);
MYTIMESTAMP(tstop);
tmmult = tstop-tstart;
MYTIMESTAMP(tstart);
cblas_dgemm(CblasRowMajor,
CblasNoTrans, CblasNoTrans, SIZE_M, SIZE_N, SIZE_K,
1.0, (const double*)A, SIZE_M, (const double*)B,
SIZE_N, 0.0, (double*)C, SIZE_K);
MYTIMESTAMP(tstop);
tdgemm = tstop-tstart;
fprintf(stderr, "#M,N,K,tmmult,tdgemm,gflops_mmult,gflops_dgemm\n");
fprintf(stderr, "%d,%d,%d,%f,%f,%f,%f\n",
SIZE_M, SIZE_N, SIZE_K, tmmult, tdgemm,
1.0e-6*nflop/tmmult,
1.0e-6*nflop/tdgemm);
executeUnloopMethod(nflop, mmult_unroll4, "Unloop 4");
executeUnloopMethod(nflop, mmult_unroll8, "Unloop 8");
executeUnloopMethod(nflop, mmult_unroll16, "Unloop 16");
executeUnloopMethod(nflop, mmult_unroll24, "Unloop 24");
executeUnloopMethod(nflop, mmult_unroll28, "Unloop 28");
executeUnloopMethod(nflop, mmult_unroll222, "Unloop 222");
return 0;
}
void executeUnloopMethod(double nflops, void (*unrollMethod) (double (*)[SIZE_N], double (*)[SIZE_N], double (*)[SIZE_N]), char *name) {
double tstart, tstop, tmmult;
MYTIMESTAMP(tstart);
(*unrollMethod)( A, B, C);
MYTIMESTAMP(tstop);
tmmult = tstop-tstart;
fprintf(stderr, "%s: %fsec, %fgflops\n", name, tmmult, 1.0e-6*nflops/tmmult);
if(CHECK == 1){
fprintf(stderr, "%s\t", name);
compare(C);
}
}
int main(int argc, char *argv[])
{
long i;
long count = 0;
double tstart, tstop, time;
//printf ("Max Threads: %d\n", omp_get_max_threads());
//printf ("Num Threads: %d\n", omp_get_num_threads());
MYTIMESTAMP(tstart);
#pragma omp parallel for reduction(+: count)
for (i = 2; i <= NUM_ITERATIONS; i++)
{
count += isprime(i);
}
MYTIMESTAMP(tstop);
time = tstop-tstart;
printf("prime count = %ld\n", count);
printf("Time: %.*f\n", 2, time);
return 0;
}
int main(int argc, char** argv)
{
if (argc != 2)
{
fprintf(stderr, "Usage: avg num_elements_per_proc\n");
exit(1);
}
double start, stop;
int num_elements_per_proc = atoi(argv[1]);
// Seed the random number generator to get different results each time
srand(time(NULL));
MYTIMESTAMP(start);
MPI_Init(NULL, NULL);
int world_rank;
MPI_Comm_rank(MPI_COMM_WORLD, &world_rank);
int world_size;
MPI_Comm_size(MPI_COMM_WORLD, &world_size);
// Create a random array of elements on all processes.
srand(world_rank);
float *rand_nums = NULL;
rand_nums = create_rand_nums(num_elements_per_proc);
// Sum the numbers locally
float local_sum = 0;
int i;
for (i = 0; i < num_elements_per_proc; i++)
{
local_sum += rand_nums[i];
}
// Reduce all of the local sums into the global sum in order to
// calculate the mean
float global_sum;
MPI_Allreduce(&local_sum, &global_sum, 1, MPI_FLOAT, MPI_SUM,
MPI_COMM_WORLD);
float mean = global_sum / (num_elements_per_proc * world_size);
// Compute the local sum of the squared differences from the mean
float local_sq_diff = 0;
for (i = 0; i < num_elements_per_proc; i++)
{
local_sq_diff += (rand_nums[i] - mean) * (rand_nums[i] - mean);
}
// Reduce the global sum of the squared differences to the root process
// and print off the answer
float global_sq_diff;
MPI_Reduce(&local_sq_diff, &global_sq_diff, 1, MPI_FLOAT, MPI_SUM, 0,
MPI_COMM_WORLD);
// The standard deviation is the square root of the mean of the squared
// differences.
if (world_rank == 0)
{
float stddev = sqrt(global_sq_diff /
(num_elements_per_proc * world_size));
printf("Mean - %f, Standard deviation = %f\n", mean, stddev);
}
// Clean up
free(rand_nums);
MPI_Barrier(MPI_COMM_WORLD);
MPI_Finalize();
if (world_rank == 0)
{
MYTIMESTAMP(stop);
printf("number of seconds: %f\n", stop - start);
}
}
