unimore/hpc-2022-g3
unimore/hpc-2022-g3
1
Fork 0
mirror of https://github.com/Steffo99/unimore-hpc-assignments.git synced 2024-11-25 17:44:23 +00:00
Code Activity

Remove OpenMP pragmas

Browse source
This commit is contained in:
Steffo 2022-11-28 14:38:13 +01:00
parent a2a070bb3a
commit 43deb504c9
Signed by: steffo
GPG key ID: 6965406171929D01
1 changed files with 0 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
OpenMP/linear-algebra/kernels/atax/atax.cu
View file

@ -25,17 +25,11 @@ static void init_array(int nx, int ny,
int i, j; int i, j;
/// Initialize the `x` array with PI and its multiples. /// Initialize the `x` array with PI and its multiples.
#ifdef TOGGLE_INIT_ARRAY_1
#pragma omp parallel for num_threads(THREAD_COUNT) schedule(static)
#endif
for (i = 0; i < ny; i++) { for (i = 0; i < ny; i++) {
x[i] = i * M_PI; x[i] = i * M_PI;
} }
/// Initialize the `A` matrix /// Initialize the `A` matrix
#ifdef TOGGLE_INIT_ARRAY_2
#pragma omp parallel for num_threads(THREAD_COUNT) schedule(static)
#endif
for (i = 0; i < nx; i++) { for (i = 0; i < nx; i++) {
for (j = 0; j < ny; j++) { for (j = 0; j < ny; j++) {
A[i][j] = ((DATA_TYPE)i * (j + 1)) / nx; A[i][j] = ((DATA_TYPE)i * (j + 1)) / nx;
@ -67,18 +61,12 @@ static void kernel_atax(int nx, int ny,
{ {
int i, j; int i, j;
#ifdef TOGGLE_KERNEL_ATAX_1
#pragma omp parallel for num_threads(THREAD_COUNT) schedule(static)
#endif
for (i = 0; i < _PB_NY; i++) for (i = 0; i < _PB_NY; i++)
y[i] = 0; y[i] = 0;
/// This computes... something? I guess whatever ATAX is? /// This computes... something? I guess whatever ATAX is?
// Now this gives a nice speedup, especially with a lot more threads than the count! // Now this gives a nice speedup, especially with a lot more threads than the count!
// THREAD_COUNT * 4 seems to be the best on my local computer. What's the best for the Jetson Nano? // THREAD_COUNT * 4 seems to be the best on my local computer. What's the best for the Jetson Nano?
#ifdef TOGGLE_KERNEL_ATAX_2
#pragma omp parallel for num_threads(THREAD_COUNT) schedule(static)
#endif
for (i = 0; i < _PB_NX; i++) for (i = 0; i < _PB_NX; i++)
{ {
/// Every iteration has its own tmp variable /// Every iteration has its own tmp variable