1
0
mirror of https://github.com/TREX-CoE/qmckl.git synced 2024-11-19 20:42:50 +01:00

Improved HPC of jastrow deriv

This commit is contained in:
Anthony Scemama 2023-04-11 19:16:14 +02:00
parent b0bec3bc6c
commit 7bec8b7984

View File

@ -5540,45 +5540,45 @@ integer function qmckl_compute_een_rescaled_e_doc_f( &
! Prepare table of exponentiated distances raised to appropriate power
een_rescaled_e = 0.0d0
do nw = 1, walk_num
een_rescaled_e_ij = 0.0d0
een_rescaled_e_ij(:, 1) = 1.0d0
een_rescaled_e_ij = 0.0d0
een_rescaled_e_ij(:, 1) = 1.0d0
k = 0
do j = 1, elec_num
do i = 1, j - 1
k = k + 1
een_rescaled_e_ij(k, 2) = dexp(-rescale_factor_ee * ee_distance(i, j, nw))
end do
end do
k = 0
do j = 1, elec_num
do i = 1, j - 1
k = k + 1
een_rescaled_e_ij(k, 2) = dexp(-rescale_factor_ee * ee_distance(i, j, nw))
end do
end do
do l = 2, cord_num
do k = 1, elec_num * (elec_num - 1)/2
een_rescaled_e_ij(k, l + 1) = een_rescaled_e_ij(k, l) * een_rescaled_e_ij(k, 2)
end do
end do
do l = 2, cord_num
do k = 1, elec_num * (elec_num - 1)/2
een_rescaled_e_ij(k, l + 1) = een_rescaled_e_ij(k, l) * een_rescaled_e_ij(k, 2)
end do
end do
! prepare the actual een table
een_rescaled_e(:, :, 0, nw) = 1.0d0
! prepare the actual een table
een_rescaled_e(:, :, 0, nw) = 1.0d0
do l = 1, cord_num
k = 0
do j = 1, elec_num
do i = 1, j - 1
k = k + 1
x = een_rescaled_e_ij(k, l + 1)
een_rescaled_e(i, j, l, nw) = x
een_rescaled_e(j, i, l, nw) = x
end do
end do
end do
do l = 1, cord_num
k = 0
do j = 1, elec_num
do i = 1, j - 1
k = k + 1
x = een_rescaled_e_ij(k, l + 1)
een_rescaled_e(i, j, l, nw) = x
een_rescaled_e(j, i, l, nw) = x
end do
end do
end do
do l = 0, cord_num
do j = 1, elec_num
een_rescaled_e(j, j, l, nw) = 0.0d0
end do
end do
do l = 0, cord_num
do j = 1, elec_num
een_rescaled_e(j, j, l, nw) = 0.0d0
end do
end do
end do
@ -5654,9 +5654,7 @@ qmckl_exit_code qmckl_compute_een_rescaled_e_hpc (
// Prepare table of exponentiated distances raised to appropriate power
// init
for (int kk = 0; kk < walk_num*(cord_num+1)*elec_num*elec_num; ++kk) {
een_rescaled_e[kk]= 0.0;
}
memset(een_rescaled_e,0,walk_num*(cord_num+1)*elec_num*elec_num*sizeof(double));
const size_t elec_pairs = (size_t) (elec_num * (elec_num - 1)) / 2;
const size_t len_een_ij = (size_t) elec_pairs * (cord_num + 1);
@ -5665,41 +5663,43 @@ qmckl_exit_code qmckl_compute_een_rescaled_e_hpc (
// probably in C is better [cord+1, Ne*(Ne-1)/2]
// elec_pairs = (elec_num * (elec_num - 1)) / 2;
// len_een_ij = elec_pairs * (cord_num + 1);
qmckl_memory_info_struct mem_info = qmckl_memory_info_struct_zero;
mem_info.size = len_een_ij * sizeof(double);
double* const restrict een_rescaled_e_ij = (double*) qmckl_malloc(context, mem_info);
if (een_rescaled_e_ij == NULL) {
return qmckl_failwith( context,
QMCKL_ALLOCATION_FAILED,
"qmckl_compute_een_rescaled_e_hpc",
NULL);
}
const size_t e2 = elec_num*elec_num;
#ifdef HAVE_OPENMP
#pragma omp parallel for
#endif
for (int nw = 0; nw < walk_num; ++nw) {
for (size_t nw = 0; nw < (size_t) walk_num; ++nw) {
for (int kk = 0; kk < len_een_ij; ++kk) {
// this array initialized at 0 except een_rescaled_e_ij(:, 1) = 1.0d0
// and the arrangement of indices is [cord_num+1, ne*(ne-1)/2]
een_rescaled_e_ij[kk]= ( kk < (elec_pairs) ? 1.0 : 0.0 );
double een_rescaled_e_ij[len_een_ij];
memset(&(een_rescaled_e_ij[0]),0,len_een_ij*sizeof(double));
for (size_t kk = 0; kk < elec_pairs ; ++kk) {
een_rescaled_e_ij[kk]= 1.0;
}
size_t kk = 0;
for (size_t i = 0; i < elec_num; ++i) {
for (size_t i = 0; i < (size_t) elec_num; ++i) {
#ifdef HAVE_OPENMP
#pragma omp simd
#endif
for (size_t j = 0; j < i; ++j) {
// een_rescaled_e_ij(kk, 2) = dexp(-rescale_factor_ee * ee_distance(i, j, nw));
een_rescaled_e_ij[kk + elec_pairs] = exp(-rescale_factor_ee * \
ee_distance[j + i*elec_num + nw*e2]);
kk += 1;
een_rescaled_e_ij[j + kk + elec_pairs] = -rescale_factor_ee * ee_distance[j + i*elec_num + nw*e2];
}
kk += i;
}
#ifdef HAVE_OPENMP
#pragma omp simd
#endif
for (size_t k = elec_pairs; k < 2*elec_pairs; ++k) {
een_rescaled_e_ij[k] = exp(een_rescaled_e_ij[k]);
}
for (size_t l = 2; l < (cord_num+1); ++l) {
for (size_t l = 2; l < (size_t) (cord_num+1); ++l) {
#ifdef HAVE_OPENMP
#pragma omp simd
#endif
for (size_t k = 0; k < elec_pairs; ++k) {
// een_rescaled_e_ij(k, l + 1) = een_rescaled_e_ij(k, l + 1 - 1) * een_rescaled_e_ij(k, 2)
een_rescaled_e_ij[k+l*elec_pairs] = een_rescaled_e_ij[k + (l - 1)*elec_pairs] * \
@ -5709,16 +5709,18 @@ qmckl_exit_code qmckl_compute_een_rescaled_e_hpc (
double* const een_rescaled_e_ = &(een_rescaled_e[nw*(cord_num+1)*e2]);
// prepare the actual een table
#ifdef HAVE_OPENMP
#pragma omp simd
#endif
for (size_t i = 0; i < e2; ++i){
een_rescaled_e_[i] = 1.0;
}
// Up to here it should work.
for ( size_t l = 1; l < (cord_num+1); ++l) {
for ( size_t l = 1; l < (size_t) (cord_num+1); ++l) {
double* x = een_rescaled_e_ij + l*elec_pairs;
double* const een_rescaled_e__ = &(een_rescaled_e_[l*e2]);
double* een_rescaled_e_i = een_rescaled_e__;
for (size_t i = 0; i < elec_num; ++i) {
for (size_t i = 0; i < (size_t) elec_num; ++i) {
for (size_t j = 0; j < i; ++j) {
een_rescaled_e_i[j] = *x;
een_rescaled_e__[i + j*elec_num] = *x;
@ -5729,9 +5731,9 @@ qmckl_exit_code qmckl_compute_een_rescaled_e_hpc (
}
double* const x0 = &(een_rescaled_e[nw*e2*(cord_num+1)]);
for (size_t l = 0; l < (cord_num + 1); ++l) {
for (size_t l = 0; l < (size_t) (cord_num + 1); ++l) {
double* x1 = &(x0[l*e2]);
for (size_t j = 0; j < elec_num; ++j) {
for (size_t j = 0; j < (size_t) elec_num; ++j) {
,*x1 = 0.0;
x1 += 1+elec_num;
}
@ -5739,9 +5741,7 @@ qmckl_exit_code qmckl_compute_een_rescaled_e_hpc (
}
qmckl_exit_code rc = qmckl_free(context,een_rescaled_e_ij);
return rc;
return QMCKL_SUCCESS;
}
#+end_src