Homec4science
Diffusion lammps (efficient_neuronet) Browse

lammps/src/USER-OMP/pair_adp_omp.cpp
3fb19bcea664efficient_neuronet

Last Change
Raw File

pair_adp_omp.cpp
View Options

/* ----------------------------------------------------------------------
LAMMPS - Large-scale Atomic/Molecular Massively Parallel Simulator
http://lammps.sandia.gov, Sandia National Laboratories
Steve Plimpton, sjplimp@sandia.gov
This software is distributed under the GNU General Public License.
See the README file in the top-level LAMMPS directory.
------------------------------------------------------------------------- */
/* ----------------------------------------------------------------------
Contributing author: Axel Kohlmeyer (Temple U)
------------------------------------------------------------------------- */
#include "math.h"
#include "string.h"
#include "pair_adp_omp.h"
#include "atom.h"
#include "comm.h"
#include "force.h"
#include "memory.h"
#include "neighbor.h"
#include "neigh_list.h"
#include "suffix.h"
using namespace LAMMPS_NS;
/* ---------------------------------------------------------------------- */
PairADPOMP::PairADPOMP(LAMMPS *lmp) :
PairADP(lmp), ThrOMP(lmp, THR_PAIR)
{
suffix_flag |= Suffix::OMP;
respa_enable = 0;
}
/* ---------------------------------------------------------------------- */
void PairADPOMP::compute(int eflag, int vflag)
{
if (eflag || vflag) {
ev_setup(eflag,vflag);
} else evflag = vflag_fdotr = eflag_global = eflag_atom = 0;
const int nlocal = atom->nlocal;
const int nall = nlocal + atom->nghost;
const int nthreads = comm->nthreads;
const int inum = list->inum;
// grow energy and fp arrays if necessary
// need to be atom->nmax in length
if (atom->nmax > nmax) {
memory->destroy(rho);
memory->destroy(fp);
memory->destroy(mu);
memory->destroy(lambda);
nmax = atom->nmax;
memory->create(rho,nthreads*nmax,"pair:rho");
memory->create(fp,nmax,"pair:fp");
memory->create(mu,nthreads*nmax,3,"pair:mu");
memory->create(lambda,nthreads*nmax,6,"pair:lambda");
}
#if defined(_OPENMP)
#pragma omp parallel default(none) shared(eflag,vflag)
#endif
{
int ifrom, ito, tid;
loop_setup_thr(ifrom, ito, tid, inum, nthreads);
ThrData *thr = fix->get_thr(tid);
ev_setup_thr(eflag, vflag, nall, eatom, vatom, thr);
if (force->newton_pair)
thr->init_adp(nall, rho, mu, lambda);
else
thr->init_adp(nlocal, rho, mu, lambda);
if (evflag) {
if (eflag) {
if (force->newton_pair) eval<1,1,1>(ifrom, ito, thr);
else eval<1,1,0>(ifrom, ito, thr);
} else {
if (force->newton_pair) eval<1,0,1>(ifrom, ito, thr);
else eval<1,0,0>(ifrom, ito, thr);
}
} else {
if (force->newton_pair) eval<0,0,1>(ifrom, ito, thr);
else eval<0,0,0>(ifrom, ito, thr);
}
reduce_thr(this, eflag, vflag, thr);
} // end of omp parallel region
}
template <int EVFLAG, int EFLAG, int NEWTON_PAIR>
void PairADPOMP::eval(int iifrom, int iito, ThrData * const thr)
{
int i,j,ii,jj,m,jnum,itype,jtype;
double xtmp,ytmp,ztmp,delx,dely,delz,evdwl,fpair;
double rsq,r,p,rhoip,rhojp,z2,z2p,recip,phip,psip,phi;
double u2,u2p,w2,w2p,nu;
double *coeff;
int *ilist,*jlist,*numneigh,**firstneigh;
double delmux,delmuy,delmuz,trdelmu,tradellam;
double adpx,adpy,adpz,fx,fy,fz;
double sumlamxx,sumlamyy,sumlamzz,sumlamyz,sumlamxz,sumlamxy;
evdwl = 0.0;
const double * const * const x = atom->x;
double * const * const f = thr->get_f();
double * const rho_t = thr->get_rho();
double * const * const mu_t = thr->get_mu();
double * const * const lambda_t = thr->get_lambda();
const int tid = thr->get_tid();
int *type = atom->type;
int nlocal = atom->nlocal;
int nall = nlocal + atom->nghost;
double fxtmp,fytmp,fztmp;
ilist = list->ilist;
numneigh = list->numneigh;
firstneigh = list->firstneigh;
// rho = density at each atom
// loop over neighbors of my atoms
for (ii = iifrom; ii < iito; ii++) {
i = ilist[ii];
xtmp = x[i][0];
ytmp = x[i][1];
ztmp = x[i][2];
itype = type[i];
jlist = firstneigh[i];
jnum = numneigh[i];
for (jj = 0; jj < jnum; jj++) {
j = jlist[jj];
j &= NEIGHMASK;
delx = xtmp - x[j][0];
dely = ytmp - x[j][1];
delz = ztmp - x[j][2];
rsq = delx*delx + dely*dely + delz*delz;
if (rsq < cutforcesq) {
jtype = type[j];
p = sqrt(rsq)*rdr + 1.0;
m = static_cast<int> (p);
m = MIN(m,nr-1);
p -= m;
p = MIN(p,1.0);
coeff = rhor_spline[type2rhor[jtype][itype]][m];
rho_t[i] += ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
coeff = u2r_spline[type2u2r[jtype][itype]][m];
u2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
mu_t[i][0] += u2*delx;
mu_t[i][1] += u2*dely;
mu_t[i][2] += u2*delz;
coeff = w2r_spline[type2w2r[jtype][itype]][m];
w2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
lambda_t[i][0] += w2*delx*delx;
lambda_t[i][1] += w2*dely*dely;
lambda_t[i][2] += w2*delz*delz;
lambda_t[i][3] += w2*dely*delz;
lambda_t[i][4] += w2*delx*delz;
lambda_t[i][5] += w2*delx*dely;
if (NEWTON_PAIR || j < nlocal) {
// verify sign difference for mu and lambda
coeff = rhor_spline[type2rhor[itype][jtype]][m];
rho_t[j] += ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
coeff = u2r_spline[type2u2r[itype][jtype]][m];
u2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
mu_t[j][0] -= u2*delx;
mu_t[j][1] -= u2*dely;
mu_t[j][2] -= u2*delz;
coeff = w2r_spline[type2w2r[itype][jtype]][m];
w2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
lambda_t[j][0] += w2*delx*delx;
lambda_t[j][1] += w2*dely*dely;
lambda_t[j][2] += w2*delz*delz;
lambda_t[j][3] += w2*dely*delz;
lambda_t[j][4] += w2*delx*delz;
lambda_t[j][5] += w2*delx*dely;
}
}
}
}
// wait until all threads are done with computation
sync_threads();
// communicate and sum densities
if (NEWTON_PAIR) {
// reduce per thread density
data_reduce_thr(&(rho[0]), nall, comm->nthreads, 1, tid);
data_reduce_thr(&(mu[0][0]), nall, comm->nthreads, 3, tid);
data_reduce_thr(&(lambda[0][0]), nall, comm->nthreads, 6, tid);
// wait until reduction is complete
sync_threads();
#if defined(_OPENMP)
#pragma omp master
#endif
{ comm->reverse_comm_pair(this); }
// wait until master thread is done with communication
sync_threads();
} else {
// reduce per thread density
data_reduce_thr(&(rho[0]), nlocal, comm->nthreads, 1, tid);
data_reduce_thr(&(mu[0][0]), nlocal, comm->nthreads, 3, tid);
data_reduce_thr(&(lambda[0][0]), nlocal, comm->nthreads, 6, tid);
// wait until reduction is complete
sync_threads();
}
// fp = derivative of embedding energy at each atom
// phi = embedding energy at each atom
for (ii = iifrom; ii < iito; ii++) {
i = ilist[ii];
p = rho[i]*rdrho + 1.0;
m = static_cast<int> (p);
m = MAX(1,MIN(m,nrho-1));
p -= m;
p = MIN(p,1.0);
coeff = frho_spline[type2frho[type[i]]][m];
fp[i] = (coeff[0]*p + coeff[1])*p + coeff[2];
if (EFLAG) {
phi = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
phi += 0.5*(mu[i][0]*mu[i][0]+mu[i][1]*mu[i][1]+mu[i][2]*mu[i][2]);
phi += 0.5*(lambda[i][0]*lambda[i][0]+lambda[i][1]*
lambda[i][1]+lambda[i][2]*lambda[i][2]);
phi += 1.0*(lambda[i][3]*lambda[i][3]+lambda[i][4]*
lambda[i][4]+lambda[i][5]*lambda[i][5]);
phi -= 1.0/6.0*(lambda[i][0]+lambda[i][1]+lambda[i][2])*
(lambda[i][0]+lambda[i][1]+lambda[i][2]);
e_tally_thr(this,i,i,nlocal,/* newton_pair */ 1, phi, 0.0, thr);
}
}
// wait until all theads are done with computation
sync_threads();
// communicate derivative of embedding function
// MPI communication only on master thread
#if defined(_OPENMP)
#pragma omp master
#endif
{ comm->forward_comm_pair(this); }
// wait until master thread is done with communication
sync_threads();
// compute forces on each atom
// loop over neighbors of my atoms
for (ii = iifrom; ii < iito; ii++) {
i = ilist[ii];
xtmp = x[i][0];
ytmp = x[i][1];
ztmp = x[i][2];
itype = type[i];
fxtmp = fytmp = fztmp = 0.0;
jlist = firstneigh[i];
jnum = numneigh[i];
for (jj = 0; jj < jnum; jj++) {
j = jlist[jj];
j &= NEIGHMASK;
delx = xtmp - x[j][0];
dely = ytmp - x[j][1];
delz = ztmp - x[j][2];
rsq = delx*delx + dely*dely + delz*delz;
if (rsq < cutforcesq) {
jtype = type[j];
r = sqrt(rsq);
p = r*rdr + 1.0;
m = static_cast<int> (p);
m = MIN(m,nr-1);
p -= m;
p = MIN(p,1.0);
// rhoip = derivative of (density at atom j due to atom i)
// rhojp = derivative of (density at atom i due to atom j)
// phi = pair potential energy
// phip = phi'
// z2 = phi * r
// z2p = (phi * r)' = (phi' r) + phi
// u2 = u
// u2p = u'
// w2 = w
// w2p = w'
// psip needs both fp[i] and fp[j] terms since r_ij appears in two
// terms of embed eng: Fi(sum rho_ij) and Fj(sum rho_ji)
// hence embed' = Fi(sum rho_ij) rhojp + Fj(sum rho_ji) rhoip
coeff = rhor_spline[type2rhor[itype][jtype]][m];
rhoip = (coeff[0]*p + coeff[1])*p + coeff[2];
coeff = rhor_spline[type2rhor[jtype][itype]][m];
rhojp = (coeff[0]*p + coeff[1])*p + coeff[2];
coeff = z2r_spline[type2z2r[itype][jtype]][m];
z2p = (coeff[0]*p + coeff[1])*p + coeff[2];
z2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
coeff = u2r_spline[type2u2r[itype][jtype]][m];
u2p = (coeff[0]*p + coeff[1])*p + coeff[2];
u2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
coeff = w2r_spline[type2w2r[itype][jtype]][m];
w2p = (coeff[0]*p + coeff[1])*p + coeff[2];
w2 = ((coeff[3]*p + coeff[4])*p + coeff[5])*p + coeff[6];
recip = 1.0/r;
phi = z2*recip;
phip = z2p*recip - phi*recip;
psip = fp[i]*rhojp + fp[j]*rhoip + phip;
fpair = -psip*recip;
delmux = mu[i][0]-mu[j][0];
delmuy = mu[i][1]-mu[j][1];
delmuz = mu[i][2]-mu[j][2];
trdelmu = delmux*delx+delmuy*dely+delmuz*delz;
sumlamxx = lambda[i][0]+lambda[j][0];
sumlamyy = lambda[i][1]+lambda[j][1];
sumlamzz = lambda[i][2]+lambda[j][2];
sumlamyz = lambda[i][3]+lambda[j][3];
sumlamxz = lambda[i][4]+lambda[j][4];
sumlamxy = lambda[i][5]+lambda[j][5];
tradellam = sumlamxx*delx*delx+sumlamyy*dely*dely+
sumlamzz*delz*delz+2.0*sumlamxy*delx*dely+
2.0*sumlamxz*delx*delz+2.0*sumlamyz*dely*delz;
nu = sumlamxx+sumlamyy+sumlamzz;
adpx = delmux*u2 + trdelmu*u2p*delx*recip +
2.0*w2*(sumlamxx*delx+sumlamxy*dely+sumlamxz*delz) +
w2p*delx*recip*tradellam - 1.0/3.0*nu*(w2p*r+2.0*w2)*delx;
adpy = delmuy*u2 + trdelmu*u2p*dely*recip +
2.0*w2*(sumlamxy*delx+sumlamyy*dely+sumlamyz*delz) +
w2p*dely*recip*tradellam - 1.0/3.0*nu*(w2p*r+2.0*w2)*dely;
adpz = delmuz*u2 + trdelmu*u2p*delz*recip +
2.0*w2*(sumlamxz*delx+sumlamyz*dely+sumlamzz*delz) +
w2p*delz*recip*tradellam - 1.0/3.0*nu*(w2p*r+2.0*w2)*delz;
adpx*=-1.0; adpy*=-1.0; adpz*=-1.0;
fx = delx*fpair+adpx;
fy = dely*fpair+adpy;
fz = delz*fpair+adpz;
fxtmp += fx;
fytmp += fy;
fztmp += fz;
if (NEWTON_PAIR || j < nlocal) {
f[j][0] -= fx;
f[j][1] -= fy;
f[j][2] -= fz;
}
if (EFLAG) evdwl = phi;
if (EVFLAG) ev_tally_xyz_thr(this,i,j,nlocal,NEWTON_PAIR,evdwl,0.0,
fx,fy,fz,delx,dely,delz,thr);
}
}
f[i][0] += fxtmp;
f[i][1] += fytmp;
f[i][2] += fztmp;
}
}
/* ---------------------------------------------------------------------- */
double PairADPOMP::memory_usage()
{
double bytes = memory_usage_thr();
bytes += PairADP::memory_usage();
bytes += (comm->nthreads-1) * nmax * (10*sizeof(double) + 3*sizeof(double *));
return bytes;
}
c4science · Help