src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecRFCut_VdwLJSh_GeomW3P1_avx_256_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_256_single kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_256_single.h"
  48 #include "kernelutil_x86_avx_256_single.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwLJSh_GeomW3P1_VF_avx_256_single
  52  * Electrostatics interaction: ReactionField
  53  * VdW interaction:            LennardJones
  54  * Geometry:                   Water3-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecRFCut_VdwLJSh_GeomW3P1_VF_avx_256_single
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB,jnrC,jnrD;
  75     int              jnrE,jnrF,jnrG,jnrH;
  76     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  77     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  78     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  79     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
  80     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  81     real             rcutoff_scalar;
  82     real             *shiftvec,*fshift,*x,*f;
  83     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
  84     real             scratch[4*DIM];
  85     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  86     real *           vdwioffsetptr0;
  87     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  88     real *           vdwioffsetptr1;
  89     __m256           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  90     real *           vdwioffsetptr2;
  91     __m256           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  92     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
  93     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  94     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  95     __m256           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  96     __m256           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  97     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
  98     real             *charge;
  99     int              nvdwtype;
 100     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 101     int              *vdwtype;
 102     real             *vdwparam;
 103     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 104     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 105     __m256           dummy_mask,cutoff_mask;
 106     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 107     __m256           one     = _mm256_set1_ps(1.0);
 108     __m256           two     = _mm256_set1_ps(2.0);
 109     x                = xx[0];
 110     f                = ff[0];
 111
 112     nri              = nlist->nri;
 113     iinr             = nlist->iinr;
 114     jindex           = nlist->jindex;
 115     jjnr             = nlist->jjnr;
 116     shiftidx         = nlist->shift;
 117     gid              = nlist->gid;
 118     shiftvec         = fr->shift_vec[0];
 119     fshift           = fr->fshift[0];
 120     facel            = _mm256_set1_ps(fr->epsfac);
 121     charge           = mdatoms->chargeA;
 122     krf              = _mm256_set1_ps(fr->ic->k_rf);
 123     krf2             = _mm256_set1_ps(fr->ic->k_rf*2.0);
 124     crf              = _mm256_set1_ps(fr->ic->c_rf);
 125     nvdwtype         = fr->ntype;
 126     vdwparam         = fr->nbfp;
 127     vdwtype          = mdatoms->typeA;
 128
 129     /* Setup water-specific parameters */
 130     inr              = nlist->iinr[0];
 131     iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 132     iq1              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
 133     iq2              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
 134     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 135
 136     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 137     rcutoff_scalar   = fr->rcoulomb;
 138     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
 139     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
 140
 141     sh_vdw_invrcut6  = _mm256_set1_ps(fr->ic->sh_invrc6);
 142     rvdw             = _mm256_set1_ps(fr->rvdw);
 143
 144     /* Avoid stupid compiler warnings */
 145     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 146     j_coord_offsetA = 0;
 147     j_coord_offsetB = 0;
 148     j_coord_offsetC = 0;
 149     j_coord_offsetD = 0;
 150     j_coord_offsetE = 0;
 151     j_coord_offsetF = 0;
 152     j_coord_offsetG = 0;
 153     j_coord_offsetH = 0;
 154
 155     outeriter        = 0;
 156     inneriter        = 0;
 157
 158     for(iidx=0;iidx<4*DIM;iidx++)
 159     {
 160         scratch[iidx] = 0.0;
 161     }
 162
 163     /* Start outer loop over neighborlists */
 164     for(iidx=0; iidx<nri; iidx++)
 165     {
 166         /* Load shift vector for this list */
 167         i_shift_offset   = DIM*shiftidx[iidx];
 168
 169         /* Load limits for loop over neighbors */
 170         j_index_start    = jindex[iidx];
 171         j_index_end      = jindex[iidx+1];
 172
 173         /* Get outer coordinate index */
 174         inr              = iinr[iidx];
 175         i_coord_offset   = DIM*inr;
 176
 177         /* Load i particle coords and add shift vector */
 178         gmx_mm256_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 179                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 180
 181         fix0             = _mm256_setzero_ps();
 182         fiy0             = _mm256_setzero_ps();
 183         fiz0             = _mm256_setzero_ps();
 184         fix1             = _mm256_setzero_ps();
 185         fiy1             = _mm256_setzero_ps();
 186         fiz1             = _mm256_setzero_ps();
 187         fix2             = _mm256_setzero_ps();
 188         fiy2             = _mm256_setzero_ps();
 189         fiz2             = _mm256_setzero_ps();
 190
 191         /* Reset potential sums */
 192         velecsum         = _mm256_setzero_ps();
 193         vvdwsum          = _mm256_setzero_ps();
 194
 195         /* Start inner kernel loop */
 196         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 197         {
 198
 199             /* Get j neighbor index, and coordinate index */
 200             jnrA             = jjnr[jidx];
 201             jnrB             = jjnr[jidx+1];
 202             jnrC             = jjnr[jidx+2];
 203             jnrD             = jjnr[jidx+3];
 204             jnrE             = jjnr[jidx+4];
 205             jnrF             = jjnr[jidx+5];
 206             jnrG             = jjnr[jidx+6];
 207             jnrH             = jjnr[jidx+7];
 208             j_coord_offsetA  = DIM*jnrA;
 209             j_coord_offsetB  = DIM*jnrB;
 210             j_coord_offsetC  = DIM*jnrC;
 211             j_coord_offsetD  = DIM*jnrD;
 212             j_coord_offsetE  = DIM*jnrE;
 213             j_coord_offsetF  = DIM*jnrF;
 214             j_coord_offsetG  = DIM*jnrG;
 215             j_coord_offsetH  = DIM*jnrH;
 216
 217             /* load j atom coordinates */
 218             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 219                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 220                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 221                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 222                                                  &jx0,&jy0,&jz0);
 223
 224             /* Calculate displacement vector */
 225             dx00             = _mm256_sub_ps(ix0,jx0);
 226             dy00             = _mm256_sub_ps(iy0,jy0);
 227             dz00             = _mm256_sub_ps(iz0,jz0);
 228             dx10             = _mm256_sub_ps(ix1,jx0);
 229             dy10             = _mm256_sub_ps(iy1,jy0);
 230             dz10             = _mm256_sub_ps(iz1,jz0);
 231             dx20             = _mm256_sub_ps(ix2,jx0);
 232             dy20             = _mm256_sub_ps(iy2,jy0);
 233             dz20             = _mm256_sub_ps(iz2,jz0);
 234
 235             /* Calculate squared distance and things based on it */
 236             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 237             rsq10            = gmx_mm256_calc_rsq_ps(dx10,dy10,dz10);
 238             rsq20            = gmx_mm256_calc_rsq_ps(dx20,dy20,dz20);
 239
 240             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 241             rinv10           = gmx_mm256_invsqrt_ps(rsq10);
 242             rinv20           = gmx_mm256_invsqrt_ps(rsq20);
 243
 244             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 245             rinvsq10         = _mm256_mul_ps(rinv10,rinv10);
 246             rinvsq20         = _mm256_mul_ps(rinv20,rinv20);
 247
 248             /* Load parameters for j particles */
 249             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 250                                                                  charge+jnrC+0,charge+jnrD+0,
 251                                                                  charge+jnrE+0,charge+jnrF+0,
 252                                                                  charge+jnrG+0,charge+jnrH+0);
 253             vdwjidx0A        = 2*vdwtype[jnrA+0];
 254             vdwjidx0B        = 2*vdwtype[jnrB+0];
 255             vdwjidx0C        = 2*vdwtype[jnrC+0];
 256             vdwjidx0D        = 2*vdwtype[jnrD+0];
 257             vdwjidx0E        = 2*vdwtype[jnrE+0];
 258             vdwjidx0F        = 2*vdwtype[jnrF+0];
 259             vdwjidx0G        = 2*vdwtype[jnrG+0];
 260             vdwjidx0H        = 2*vdwtype[jnrH+0];
 261
 262             fjx0             = _mm256_setzero_ps();
 263             fjy0             = _mm256_setzero_ps();
 264             fjz0             = _mm256_setzero_ps();
 265
 266             /**************************
 267              * CALCULATE INTERACTIONS *
 268              **************************/
 269
 270             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 271             {
 272
 273             /* Compute parameters for interactions between i and j atoms */
 274             qq00             = _mm256_mul_ps(iq0,jq0);
 275             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 276                                             vdwioffsetptr0+vdwjidx0B,
 277                                             vdwioffsetptr0+vdwjidx0C,
 278                                             vdwioffsetptr0+vdwjidx0D,
 279                                             vdwioffsetptr0+vdwjidx0E,
 280                                             vdwioffsetptr0+vdwjidx0F,
 281                                             vdwioffsetptr0+vdwjidx0G,
 282                                             vdwioffsetptr0+vdwjidx0H,
 283                                             &c6_00,&c12_00);
 284
 285             /* REACTION-FIELD ELECTROSTATICS */
 286             velec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_add_ps(rinv00,_mm256_mul_ps(krf,rsq00)),crf));
 287             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 288
 289             /* LENNARD-JONES DISPERSION/REPULSION */
 290
 291             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 292             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 293             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 294             vvdw             = _mm256_sub_ps(_mm256_mul_ps( _mm256_sub_ps(vvdw12 , _mm256_mul_ps(c12_00,_mm256_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
 295                                           _mm256_mul_ps( _mm256_sub_ps(vvdw6,_mm256_mul_ps(c6_00,sh_vdw_invrcut6)),one_sixth));
 296             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 297
 298             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 299
 300             /* Update potential sum for this i atom from the interaction with this j atom. */
 301             velec            = _mm256_and_ps(velec,cutoff_mask);
 302             velecsum         = _mm256_add_ps(velecsum,velec);
 303             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
 304             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 305
 306             fscal            = _mm256_add_ps(felec,fvdw);
 307
 308             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 309
 310             /* Calculate temporary vectorial force */
 311             tx               = _mm256_mul_ps(fscal,dx00);
 312             ty               = _mm256_mul_ps(fscal,dy00);
 313             tz               = _mm256_mul_ps(fscal,dz00);
 314
 315             /* Update vectorial force */
 316             fix0             = _mm256_add_ps(fix0,tx);
 317             fiy0             = _mm256_add_ps(fiy0,ty);
 318             fiz0             = _mm256_add_ps(fiz0,tz);
 319
 320             fjx0             = _mm256_add_ps(fjx0,tx);
 321             fjy0             = _mm256_add_ps(fjy0,ty);
 322             fjz0             = _mm256_add_ps(fjz0,tz);
 323
 324             }
 325
 326             /**************************
 327              * CALCULATE INTERACTIONS *
 328              **************************/
 329
 330             if (gmx_mm256_any_lt(rsq10,rcutoff2))
 331             {
 332
 333             /* Compute parameters for interactions between i and j atoms */
 334             qq10             = _mm256_mul_ps(iq1,jq0);
 335
 336             /* REACTION-FIELD ELECTROSTATICS */
 337             velec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_add_ps(rinv10,_mm256_mul_ps(krf,rsq10)),crf));
 338             felec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_mul_ps(rinv10,rinvsq10),krf2));
 339
 340             cutoff_mask      = _mm256_cmp_ps(rsq10,rcutoff2,_CMP_LT_OQ);
 341
 342             /* Update potential sum for this i atom from the interaction with this j atom. */
 343             velec            = _mm256_and_ps(velec,cutoff_mask);
 344             velecsum         = _mm256_add_ps(velecsum,velec);
 345
 346             fscal            = felec;
 347
 348             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 349
 350             /* Calculate temporary vectorial force */
 351             tx               = _mm256_mul_ps(fscal,dx10);
 352             ty               = _mm256_mul_ps(fscal,dy10);
 353             tz               = _mm256_mul_ps(fscal,dz10);
 354
 355             /* Update vectorial force */
 356             fix1             = _mm256_add_ps(fix1,tx);
 357             fiy1             = _mm256_add_ps(fiy1,ty);
 358             fiz1             = _mm256_add_ps(fiz1,tz);
 359
 360             fjx0             = _mm256_add_ps(fjx0,tx);
 361             fjy0             = _mm256_add_ps(fjy0,ty);
 362             fjz0             = _mm256_add_ps(fjz0,tz);
 363
 364             }
 365
 366             /**************************
 367              * CALCULATE INTERACTIONS *
 368              **************************/
 369
 370             if (gmx_mm256_any_lt(rsq20,rcutoff2))
 371             {
 372
 373             /* Compute parameters for interactions between i and j atoms */
 374             qq20             = _mm256_mul_ps(iq2,jq0);
 375
 376             /* REACTION-FIELD ELECTROSTATICS */
 377             velec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_add_ps(rinv20,_mm256_mul_ps(krf,rsq20)),crf));
 378             felec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_mul_ps(rinv20,rinvsq20),krf2));
 379
 380             cutoff_mask      = _mm256_cmp_ps(rsq20,rcutoff2,_CMP_LT_OQ);
 381
 382             /* Update potential sum for this i atom from the interaction with this j atom. */
 383             velec            = _mm256_and_ps(velec,cutoff_mask);
 384             velecsum         = _mm256_add_ps(velecsum,velec);
 385
 386             fscal            = felec;
 387
 388             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 389
 390             /* Calculate temporary vectorial force */
 391             tx               = _mm256_mul_ps(fscal,dx20);
 392             ty               = _mm256_mul_ps(fscal,dy20);
 393             tz               = _mm256_mul_ps(fscal,dz20);
 394
 395             /* Update vectorial force */
 396             fix2             = _mm256_add_ps(fix2,tx);
 397             fiy2             = _mm256_add_ps(fiy2,ty);
 398             fiz2             = _mm256_add_ps(fiz2,tz);
 399
 400             fjx0             = _mm256_add_ps(fjx0,tx);
 401             fjy0             = _mm256_add_ps(fjy0,ty);
 402             fjz0             = _mm256_add_ps(fjz0,tz);
 403
 404             }
 405
 406             fjptrA             = f+j_coord_offsetA;
 407             fjptrB             = f+j_coord_offsetB;
 408             fjptrC             = f+j_coord_offsetC;
 409             fjptrD             = f+j_coord_offsetD;
 410             fjptrE             = f+j_coord_offsetE;
 411             fjptrF             = f+j_coord_offsetF;
 412             fjptrG             = f+j_coord_offsetG;
 413             fjptrH             = f+j_coord_offsetH;
 414
 415             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,fjx0,fjy0,fjz0);
 416
 417             /* Inner loop uses 129 flops */
 418         }
 419
 420         if(jidx<j_index_end)
 421         {
 422
 423             /* Get j neighbor index, and coordinate index */
 424             jnrlistA         = jjnr[jidx];
 425             jnrlistB         = jjnr[jidx+1];
 426             jnrlistC         = jjnr[jidx+2];
 427             jnrlistD         = jjnr[jidx+3];
 428             jnrlistE         = jjnr[jidx+4];
 429             jnrlistF         = jjnr[jidx+5];
 430             jnrlistG         = jjnr[jidx+6];
 431             jnrlistH         = jjnr[jidx+7];
 432             /* Sign of each element will be negative for non-real atoms.
 433              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 434              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 435              */
 436             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 437                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 438
 439             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 440             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 441             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 442             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 443             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 444             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 445             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 446             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 447             j_coord_offsetA  = DIM*jnrA;
 448             j_coord_offsetB  = DIM*jnrB;
 449             j_coord_offsetC  = DIM*jnrC;
 450             j_coord_offsetD  = DIM*jnrD;
 451             j_coord_offsetE  = DIM*jnrE;
 452             j_coord_offsetF  = DIM*jnrF;
 453             j_coord_offsetG  = DIM*jnrG;
 454             j_coord_offsetH  = DIM*jnrH;
 455
 456             /* load j atom coordinates */
 457             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 458                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 459                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 460                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 461                                                  &jx0,&jy0,&jz0);
 462
 463             /* Calculate displacement vector */
 464             dx00             = _mm256_sub_ps(ix0,jx0);
 465             dy00             = _mm256_sub_ps(iy0,jy0);
 466             dz00             = _mm256_sub_ps(iz0,jz0);
 467             dx10             = _mm256_sub_ps(ix1,jx0);
 468             dy10             = _mm256_sub_ps(iy1,jy0);
 469             dz10             = _mm256_sub_ps(iz1,jz0);
 470             dx20             = _mm256_sub_ps(ix2,jx0);
 471             dy20             = _mm256_sub_ps(iy2,jy0);
 472             dz20             = _mm256_sub_ps(iz2,jz0);
 473
 474             /* Calculate squared distance and things based on it */
 475             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 476             rsq10            = gmx_mm256_calc_rsq_ps(dx10,dy10,dz10);
 477             rsq20            = gmx_mm256_calc_rsq_ps(dx20,dy20,dz20);
 478
 479             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 480             rinv10           = gmx_mm256_invsqrt_ps(rsq10);
 481             rinv20           = gmx_mm256_invsqrt_ps(rsq20);
 482
 483             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 484             rinvsq10         = _mm256_mul_ps(rinv10,rinv10);
 485             rinvsq20         = _mm256_mul_ps(rinv20,rinv20);
 486
 487             /* Load parameters for j particles */
 488             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 489                                                                  charge+jnrC+0,charge+jnrD+0,
 490                                                                  charge+jnrE+0,charge+jnrF+0,
 491                                                                  charge+jnrG+0,charge+jnrH+0);
 492             vdwjidx0A        = 2*vdwtype[jnrA+0];
 493             vdwjidx0B        = 2*vdwtype[jnrB+0];
 494             vdwjidx0C        = 2*vdwtype[jnrC+0];
 495             vdwjidx0D        = 2*vdwtype[jnrD+0];
 496             vdwjidx0E        = 2*vdwtype[jnrE+0];
 497             vdwjidx0F        = 2*vdwtype[jnrF+0];
 498             vdwjidx0G        = 2*vdwtype[jnrG+0];
 499             vdwjidx0H        = 2*vdwtype[jnrH+0];
 500
 501             fjx0             = _mm256_setzero_ps();
 502             fjy0             = _mm256_setzero_ps();
 503             fjz0             = _mm256_setzero_ps();
 504
 505             /**************************
 506              * CALCULATE INTERACTIONS *
 507              **************************/
 508
 509             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 510             {
 511
 512             /* Compute parameters for interactions between i and j atoms */
 513             qq00             = _mm256_mul_ps(iq0,jq0);
 514             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 515                                             vdwioffsetptr0+vdwjidx0B,
 516                                             vdwioffsetptr0+vdwjidx0C,
 517                                             vdwioffsetptr0+vdwjidx0D,
 518                                             vdwioffsetptr0+vdwjidx0E,
 519                                             vdwioffsetptr0+vdwjidx0F,
 520                                             vdwioffsetptr0+vdwjidx0G,
 521                                             vdwioffsetptr0+vdwjidx0H,
 522                                             &c6_00,&c12_00);
 523
 524             /* REACTION-FIELD ELECTROSTATICS */
 525             velec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_add_ps(rinv00,_mm256_mul_ps(krf,rsq00)),crf));
 526             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 527
 528             /* LENNARD-JONES DISPERSION/REPULSION */
 529
 530             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 531             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 532             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 533             vvdw             = _mm256_sub_ps(_mm256_mul_ps( _mm256_sub_ps(vvdw12 , _mm256_mul_ps(c12_00,_mm256_mul_ps(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
 534                                           _mm256_mul_ps( _mm256_sub_ps(vvdw6,_mm256_mul_ps(c6_00,sh_vdw_invrcut6)),one_sixth));
 535             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 536
 537             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 538
 539             /* Update potential sum for this i atom from the interaction with this j atom. */
 540             velec            = _mm256_and_ps(velec,cutoff_mask);
 541             velec            = _mm256_andnot_ps(dummy_mask,velec);
 542             velecsum         = _mm256_add_ps(velecsum,velec);
 543             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
 544             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
 545             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 546
 547             fscal            = _mm256_add_ps(felec,fvdw);
 548
 549             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 550
 551             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 552
 553             /* Calculate temporary vectorial force */
 554             tx               = _mm256_mul_ps(fscal,dx00);
 555             ty               = _mm256_mul_ps(fscal,dy00);
 556             tz               = _mm256_mul_ps(fscal,dz00);
 557
 558             /* Update vectorial force */
 559             fix0             = _mm256_add_ps(fix0,tx);
 560             fiy0             = _mm256_add_ps(fiy0,ty);
 561             fiz0             = _mm256_add_ps(fiz0,tz);
 562
 563             fjx0             = _mm256_add_ps(fjx0,tx);
 564             fjy0             = _mm256_add_ps(fjy0,ty);
 565             fjz0             = _mm256_add_ps(fjz0,tz);
 566
 567             }
 568
 569             /**************************
 570              * CALCULATE INTERACTIONS *
 571              **************************/
 572
 573             if (gmx_mm256_any_lt(rsq10,rcutoff2))
 574             {
 575
 576             /* Compute parameters for interactions between i and j atoms */
 577             qq10             = _mm256_mul_ps(iq1,jq0);
 578
 579             /* REACTION-FIELD ELECTROSTATICS */
 580             velec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_add_ps(rinv10,_mm256_mul_ps(krf,rsq10)),crf));
 581             felec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_mul_ps(rinv10,rinvsq10),krf2));
 582
 583             cutoff_mask      = _mm256_cmp_ps(rsq10,rcutoff2,_CMP_LT_OQ);
 584
 585             /* Update potential sum for this i atom from the interaction with this j atom. */
 586             velec            = _mm256_and_ps(velec,cutoff_mask);
 587             velec            = _mm256_andnot_ps(dummy_mask,velec);
 588             velecsum         = _mm256_add_ps(velecsum,velec);
 589
 590             fscal            = felec;
 591
 592             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 593
 594             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 595
 596             /* Calculate temporary vectorial force */
 597             tx               = _mm256_mul_ps(fscal,dx10);
 598             ty               = _mm256_mul_ps(fscal,dy10);
 599             tz               = _mm256_mul_ps(fscal,dz10);
 600
 601             /* Update vectorial force */
 602             fix1             = _mm256_add_ps(fix1,tx);
 603             fiy1             = _mm256_add_ps(fiy1,ty);
 604             fiz1             = _mm256_add_ps(fiz1,tz);
 605
 606             fjx0             = _mm256_add_ps(fjx0,tx);
 607             fjy0             = _mm256_add_ps(fjy0,ty);
 608             fjz0             = _mm256_add_ps(fjz0,tz);
 609
 610             }
 611
 612             /**************************
 613              * CALCULATE INTERACTIONS *
 614              **************************/
 615
 616             if (gmx_mm256_any_lt(rsq20,rcutoff2))
 617             {
 618
 619             /* Compute parameters for interactions between i and j atoms */
 620             qq20             = _mm256_mul_ps(iq2,jq0);
 621
 622             /* REACTION-FIELD ELECTROSTATICS */
 623             velec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_add_ps(rinv20,_mm256_mul_ps(krf,rsq20)),crf));
 624             felec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_mul_ps(rinv20,rinvsq20),krf2));
 625
 626             cutoff_mask      = _mm256_cmp_ps(rsq20,rcutoff2,_CMP_LT_OQ);
 627
 628             /* Update potential sum for this i atom from the interaction with this j atom. */
 629             velec            = _mm256_and_ps(velec,cutoff_mask);
 630             velec            = _mm256_andnot_ps(dummy_mask,velec);
 631             velecsum         = _mm256_add_ps(velecsum,velec);
 632
 633             fscal            = felec;
 634
 635             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 636
 637             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 638
 639             /* Calculate temporary vectorial force */
 640             tx               = _mm256_mul_ps(fscal,dx20);
 641             ty               = _mm256_mul_ps(fscal,dy20);
 642             tz               = _mm256_mul_ps(fscal,dz20);
 643
 644             /* Update vectorial force */
 645             fix2             = _mm256_add_ps(fix2,tx);
 646             fiy2             = _mm256_add_ps(fiy2,ty);
 647             fiz2             = _mm256_add_ps(fiz2,tz);
 648
 649             fjx0             = _mm256_add_ps(fjx0,tx);
 650             fjy0             = _mm256_add_ps(fjy0,ty);
 651             fjz0             = _mm256_add_ps(fjz0,tz);
 652
 653             }
 654
 655             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 656             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 657             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 658             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 659             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 660             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 661             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 662             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 663
 664             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,fjx0,fjy0,fjz0);
 665
 666             /* Inner loop uses 129 flops */
 667         }
 668
 669         /* End of innermost loop */
 670
 671         gmx_mm256_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
 672                                                  f+i_coord_offset,fshift+i_shift_offset);
 673
 674         ggid                        = gid[iidx];
 675         /* Update potential energies */
 676         gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 677         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 678
 679         /* Increment number of inner iterations */
 680         inneriter                  += j_index_end - j_index_start;
 681
 682         /* Outer loop uses 20 flops */
 683     }
 684
 685     /* Increment number of outer iterations */
 686     outeriter        += nri;
 687
 688     /* Update outer/inner flops */
 689
 690     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*20 + inneriter*129);
 691 }
 692 /*
 693  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwLJSh_GeomW3P1_F_avx_256_single
 694  * Electrostatics interaction: ReactionField
 695  * VdW interaction:            LennardJones
 696  * Geometry:                   Water3-Particle
 697  * Calculate force/pot:        Force
 698  */
 699 void
 700 nb_kernel_ElecRFCut_VdwLJSh_GeomW3P1_F_avx_256_single
 701                     (t_nblist                    * gmx_restrict       nlist,
 702                      rvec                        * gmx_restrict          xx,
 703                      rvec                        * gmx_restrict          ff,
 704                      t_forcerec                  * gmx_restrict          fr,
 705                      t_mdatoms                   * gmx_restrict     mdatoms,
 706                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 707                      t_nrnb                      * gmx_restrict        nrnb)
 708 {
 709     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 710      * just 0 for non-waters.
 711      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
 712      * jnr indices corresponding to data put in the four positions in the SIMD register.
 713      */
 714     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 715     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 716     int              jnrA,jnrB,jnrC,jnrD;
 717     int              jnrE,jnrF,jnrG,jnrH;
 718     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 719     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 720     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 721     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
 722     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 723     real             rcutoff_scalar;
 724     real             *shiftvec,*fshift,*x,*f;
 725     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
 726     real             scratch[4*DIM];
 727     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 728     real *           vdwioffsetptr0;
 729     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 730     real *           vdwioffsetptr1;
 731     __m256           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 732     real *           vdwioffsetptr2;
 733     __m256           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 734     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
 735     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 736     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 737     __m256           dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 738     __m256           dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 739     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
 740     real             *charge;
 741     int              nvdwtype;
 742     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 743     int              *vdwtype;
 744     real             *vdwparam;
 745     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 746     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 747     __m256           dummy_mask,cutoff_mask;
 748     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 749     __m256           one     = _mm256_set1_ps(1.0);
 750     __m256           two     = _mm256_set1_ps(2.0);
 751     x                = xx[0];
 752     f                = ff[0];
 753
 754     nri              = nlist->nri;
 755     iinr             = nlist->iinr;
 756     jindex           = nlist->jindex;
 757     jjnr             = nlist->jjnr;
 758     shiftidx         = nlist->shift;
 759     gid              = nlist->gid;
 760     shiftvec         = fr->shift_vec[0];
 761     fshift           = fr->fshift[0];
 762     facel            = _mm256_set1_ps(fr->epsfac);
 763     charge           = mdatoms->chargeA;
 764     krf              = _mm256_set1_ps(fr->ic->k_rf);
 765     krf2             = _mm256_set1_ps(fr->ic->k_rf*2.0);
 766     crf              = _mm256_set1_ps(fr->ic->c_rf);
 767     nvdwtype         = fr->ntype;
 768     vdwparam         = fr->nbfp;
 769     vdwtype          = mdatoms->typeA;
 770
 771     /* Setup water-specific parameters */
 772     inr              = nlist->iinr[0];
 773     iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 774     iq1              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
 775     iq2              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
 776     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 777
 778     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 779     rcutoff_scalar   = fr->rcoulomb;
 780     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
 781     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
 782
 783     sh_vdw_invrcut6  = _mm256_set1_ps(fr->ic->sh_invrc6);
 784     rvdw             = _mm256_set1_ps(fr->rvdw);
 785
 786     /* Avoid stupid compiler warnings */
 787     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 788     j_coord_offsetA = 0;
 789     j_coord_offsetB = 0;
 790     j_coord_offsetC = 0;
 791     j_coord_offsetD = 0;
 792     j_coord_offsetE = 0;
 793     j_coord_offsetF = 0;
 794     j_coord_offsetG = 0;
 795     j_coord_offsetH = 0;
 796
 797     outeriter        = 0;
 798     inneriter        = 0;
 799
 800     for(iidx=0;iidx<4*DIM;iidx++)
 801     {
 802         scratch[iidx] = 0.0;
 803     }
 804
 805     /* Start outer loop over neighborlists */
 806     for(iidx=0; iidx<nri; iidx++)
 807     {
 808         /* Load shift vector for this list */
 809         i_shift_offset   = DIM*shiftidx[iidx];
 810
 811         /* Load limits for loop over neighbors */
 812         j_index_start    = jindex[iidx];
 813         j_index_end      = jindex[iidx+1];
 814
 815         /* Get outer coordinate index */
 816         inr              = iinr[iidx];
 817         i_coord_offset   = DIM*inr;
 818
 819         /* Load i particle coords and add shift vector */
 820         gmx_mm256_load_shift_and_3rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 821                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 822
 823         fix0             = _mm256_setzero_ps();
 824         fiy0             = _mm256_setzero_ps();
 825         fiz0             = _mm256_setzero_ps();
 826         fix1             = _mm256_setzero_ps();
 827         fiy1             = _mm256_setzero_ps();
 828         fiz1             = _mm256_setzero_ps();
 829         fix2             = _mm256_setzero_ps();
 830         fiy2             = _mm256_setzero_ps();
 831         fiz2             = _mm256_setzero_ps();
 832
 833         /* Start inner kernel loop */
 834         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 835         {
 836
 837             /* Get j neighbor index, and coordinate index */
 838             jnrA             = jjnr[jidx];
 839             jnrB             = jjnr[jidx+1];
 840             jnrC             = jjnr[jidx+2];
 841             jnrD             = jjnr[jidx+3];
 842             jnrE             = jjnr[jidx+4];
 843             jnrF             = jjnr[jidx+5];
 844             jnrG             = jjnr[jidx+6];
 845             jnrH             = jjnr[jidx+7];
 846             j_coord_offsetA  = DIM*jnrA;
 847             j_coord_offsetB  = DIM*jnrB;
 848             j_coord_offsetC  = DIM*jnrC;
 849             j_coord_offsetD  = DIM*jnrD;
 850             j_coord_offsetE  = DIM*jnrE;
 851             j_coord_offsetF  = DIM*jnrF;
 852             j_coord_offsetG  = DIM*jnrG;
 853             j_coord_offsetH  = DIM*jnrH;
 854
 855             /* load j atom coordinates */
 856             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 857                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 858                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 859                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 860                                                  &jx0,&jy0,&jz0);
 861
 862             /* Calculate displacement vector */
 863             dx00             = _mm256_sub_ps(ix0,jx0);
 864             dy00             = _mm256_sub_ps(iy0,jy0);
 865             dz00             = _mm256_sub_ps(iz0,jz0);
 866             dx10             = _mm256_sub_ps(ix1,jx0);
 867             dy10             = _mm256_sub_ps(iy1,jy0);
 868             dz10             = _mm256_sub_ps(iz1,jz0);
 869             dx20             = _mm256_sub_ps(ix2,jx0);
 870             dy20             = _mm256_sub_ps(iy2,jy0);
 871             dz20             = _mm256_sub_ps(iz2,jz0);
 872
 873             /* Calculate squared distance and things based on it */
 874             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 875             rsq10            = gmx_mm256_calc_rsq_ps(dx10,dy10,dz10);
 876             rsq20            = gmx_mm256_calc_rsq_ps(dx20,dy20,dz20);
 877
 878             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 879             rinv10           = gmx_mm256_invsqrt_ps(rsq10);
 880             rinv20           = gmx_mm256_invsqrt_ps(rsq20);
 881
 882             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 883             rinvsq10         = _mm256_mul_ps(rinv10,rinv10);
 884             rinvsq20         = _mm256_mul_ps(rinv20,rinv20);
 885
 886             /* Load parameters for j particles */
 887             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 888                                                                  charge+jnrC+0,charge+jnrD+0,
 889                                                                  charge+jnrE+0,charge+jnrF+0,
 890                                                                  charge+jnrG+0,charge+jnrH+0);
 891             vdwjidx0A        = 2*vdwtype[jnrA+0];
 892             vdwjidx0B        = 2*vdwtype[jnrB+0];
 893             vdwjidx0C        = 2*vdwtype[jnrC+0];
 894             vdwjidx0D        = 2*vdwtype[jnrD+0];
 895             vdwjidx0E        = 2*vdwtype[jnrE+0];
 896             vdwjidx0F        = 2*vdwtype[jnrF+0];
 897             vdwjidx0G        = 2*vdwtype[jnrG+0];
 898             vdwjidx0H        = 2*vdwtype[jnrH+0];
 899
 900             fjx0             = _mm256_setzero_ps();
 901             fjy0             = _mm256_setzero_ps();
 902             fjz0             = _mm256_setzero_ps();
 903
 904             /**************************
 905              * CALCULATE INTERACTIONS *
 906              **************************/
 907
 908             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 909             {
 910
 911             /* Compute parameters for interactions between i and j atoms */
 912             qq00             = _mm256_mul_ps(iq0,jq0);
 913             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 914                                             vdwioffsetptr0+vdwjidx0B,
 915                                             vdwioffsetptr0+vdwjidx0C,
 916                                             vdwioffsetptr0+vdwjidx0D,
 917                                             vdwioffsetptr0+vdwjidx0E,
 918                                             vdwioffsetptr0+vdwjidx0F,
 919                                             vdwioffsetptr0+vdwjidx0G,
 920                                             vdwioffsetptr0+vdwjidx0H,
 921                                             &c6_00,&c12_00);
 922
 923             /* REACTION-FIELD ELECTROSTATICS */
 924             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 925
 926             /* LENNARD-JONES DISPERSION/REPULSION */
 927
 928             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 929             fvdw             = _mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),c6_00),_mm256_mul_ps(rinvsix,rinvsq00));
 930
 931             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 932
 933             fscal            = _mm256_add_ps(felec,fvdw);
 934
 935             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 936
 937             /* Calculate temporary vectorial force */
 938             tx               = _mm256_mul_ps(fscal,dx00);
 939             ty               = _mm256_mul_ps(fscal,dy00);
 940             tz               = _mm256_mul_ps(fscal,dz00);
 941
 942             /* Update vectorial force */
 943             fix0             = _mm256_add_ps(fix0,tx);
 944             fiy0             = _mm256_add_ps(fiy0,ty);
 945             fiz0             = _mm256_add_ps(fiz0,tz);
 946
 947             fjx0             = _mm256_add_ps(fjx0,tx);
 948             fjy0             = _mm256_add_ps(fjy0,ty);
 949             fjz0             = _mm256_add_ps(fjz0,tz);
 950
 951             }
 952
 953             /**************************
 954              * CALCULATE INTERACTIONS *
 955              **************************/
 956
 957             if (gmx_mm256_any_lt(rsq10,rcutoff2))
 958             {
 959
 960             /* Compute parameters for interactions between i and j atoms */
 961             qq10             = _mm256_mul_ps(iq1,jq0);
 962
 963             /* REACTION-FIELD ELECTROSTATICS */
 964             felec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_mul_ps(rinv10,rinvsq10),krf2));
 965
 966             cutoff_mask      = _mm256_cmp_ps(rsq10,rcutoff2,_CMP_LT_OQ);
 967
 968             fscal            = felec;
 969
 970             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 971
 972             /* Calculate temporary vectorial force */
 973             tx               = _mm256_mul_ps(fscal,dx10);
 974             ty               = _mm256_mul_ps(fscal,dy10);
 975             tz               = _mm256_mul_ps(fscal,dz10);
 976
 977             /* Update vectorial force */
 978             fix1             = _mm256_add_ps(fix1,tx);
 979             fiy1             = _mm256_add_ps(fiy1,ty);
 980             fiz1             = _mm256_add_ps(fiz1,tz);
 981
 982             fjx0             = _mm256_add_ps(fjx0,tx);
 983             fjy0             = _mm256_add_ps(fjy0,ty);
 984             fjz0             = _mm256_add_ps(fjz0,tz);
 985
 986             }
 987
 988             /**************************
 989              * CALCULATE INTERACTIONS *
 990              **************************/
 991
 992             if (gmx_mm256_any_lt(rsq20,rcutoff2))
 993             {
 994
 995             /* Compute parameters for interactions between i and j atoms */
 996             qq20             = _mm256_mul_ps(iq2,jq0);
 997
 998             /* REACTION-FIELD ELECTROSTATICS */
 999             felec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_mul_ps(rinv20,rinvsq20),krf2));
1000
1001             cutoff_mask      = _mm256_cmp_ps(rsq20,rcutoff2,_CMP_LT_OQ);
1002
1003             fscal            = felec;
1004
1005             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1006
1007             /* Calculate temporary vectorial force */
1008             tx               = _mm256_mul_ps(fscal,dx20);
1009             ty               = _mm256_mul_ps(fscal,dy20);
1010             tz               = _mm256_mul_ps(fscal,dz20);
1011
1012             /* Update vectorial force */
1013             fix2             = _mm256_add_ps(fix2,tx);
1014             fiy2             = _mm256_add_ps(fiy2,ty);
1015             fiz2             = _mm256_add_ps(fiz2,tz);
1016
1017             fjx0             = _mm256_add_ps(fjx0,tx);
1018             fjy0             = _mm256_add_ps(fjy0,ty);
1019             fjz0             = _mm256_add_ps(fjz0,tz);
1020
1021             }
1022
1023             fjptrA             = f+j_coord_offsetA;
1024             fjptrB             = f+j_coord_offsetB;
1025             fjptrC             = f+j_coord_offsetC;
1026             fjptrD             = f+j_coord_offsetD;
1027             fjptrE             = f+j_coord_offsetE;
1028             fjptrF             = f+j_coord_offsetF;
1029             fjptrG             = f+j_coord_offsetG;
1030             fjptrH             = f+j_coord_offsetH;
1031
1032             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,fjx0,fjy0,fjz0);
1033
1034             /* Inner loop uses 100 flops */
1035         }
1036
1037         if(jidx<j_index_end)
1038         {
1039
1040             /* Get j neighbor index, and coordinate index */
1041             jnrlistA         = jjnr[jidx];
1042             jnrlistB         = jjnr[jidx+1];
1043             jnrlistC         = jjnr[jidx+2];
1044             jnrlistD         = jjnr[jidx+3];
1045             jnrlistE         = jjnr[jidx+4];
1046             jnrlistF         = jjnr[jidx+5];
1047             jnrlistG         = jjnr[jidx+6];
1048             jnrlistH         = jjnr[jidx+7];
1049             /* Sign of each element will be negative for non-real atoms.
1050              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1051              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
1052              */
1053             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
1054                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
1055
1056             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1057             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1058             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1059             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1060             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
1061             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
1062             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
1063             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
1064             j_coord_offsetA  = DIM*jnrA;
1065             j_coord_offsetB  = DIM*jnrB;
1066             j_coord_offsetC  = DIM*jnrC;
1067             j_coord_offsetD  = DIM*jnrD;
1068             j_coord_offsetE  = DIM*jnrE;
1069             j_coord_offsetF  = DIM*jnrF;
1070             j_coord_offsetG  = DIM*jnrG;
1071             j_coord_offsetH  = DIM*jnrH;
1072
1073             /* load j atom coordinates */
1074             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1075                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1076                                                  x+j_coord_offsetE,x+j_coord_offsetF,
1077                                                  x+j_coord_offsetG,x+j_coord_offsetH,
1078                                                  &jx0,&jy0,&jz0);
1079
1080             /* Calculate displacement vector */
1081             dx00             = _mm256_sub_ps(ix0,jx0);
1082             dy00             = _mm256_sub_ps(iy0,jy0);
1083             dz00             = _mm256_sub_ps(iz0,jz0);
1084             dx10             = _mm256_sub_ps(ix1,jx0);
1085             dy10             = _mm256_sub_ps(iy1,jy0);
1086             dz10             = _mm256_sub_ps(iz1,jz0);
1087             dx20             = _mm256_sub_ps(ix2,jx0);
1088             dy20             = _mm256_sub_ps(iy2,jy0);
1089             dz20             = _mm256_sub_ps(iz2,jz0);
1090
1091             /* Calculate squared distance and things based on it */
1092             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
1093             rsq10            = gmx_mm256_calc_rsq_ps(dx10,dy10,dz10);
1094             rsq20            = gmx_mm256_calc_rsq_ps(dx20,dy20,dz20);
1095
1096             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
1097             rinv10           = gmx_mm256_invsqrt_ps(rsq10);
1098             rinv20           = gmx_mm256_invsqrt_ps(rsq20);
1099
1100             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
1101             rinvsq10         = _mm256_mul_ps(rinv10,rinv10);
1102             rinvsq20         = _mm256_mul_ps(rinv20,rinv20);
1103
1104             /* Load parameters for j particles */
1105             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
1106                                                                  charge+jnrC+0,charge+jnrD+0,
1107                                                                  charge+jnrE+0,charge+jnrF+0,
1108                                                                  charge+jnrG+0,charge+jnrH+0);
1109             vdwjidx0A        = 2*vdwtype[jnrA+0];
1110             vdwjidx0B        = 2*vdwtype[jnrB+0];
1111             vdwjidx0C        = 2*vdwtype[jnrC+0];
1112             vdwjidx0D        = 2*vdwtype[jnrD+0];
1113             vdwjidx0E        = 2*vdwtype[jnrE+0];
1114             vdwjidx0F        = 2*vdwtype[jnrF+0];
1115             vdwjidx0G        = 2*vdwtype[jnrG+0];
1116             vdwjidx0H        = 2*vdwtype[jnrH+0];
1117
1118             fjx0             = _mm256_setzero_ps();
1119             fjy0             = _mm256_setzero_ps();
1120             fjz0             = _mm256_setzero_ps();
1121
1122             /**************************
1123              * CALCULATE INTERACTIONS *
1124              **************************/
1125
1126             if (gmx_mm256_any_lt(rsq00,rcutoff2))
1127             {
1128
1129             /* Compute parameters for interactions between i and j atoms */
1130             qq00             = _mm256_mul_ps(iq0,jq0);
1131             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
1132                                             vdwioffsetptr0+vdwjidx0B,
1133                                             vdwioffsetptr0+vdwjidx0C,
1134                                             vdwioffsetptr0+vdwjidx0D,
1135                                             vdwioffsetptr0+vdwjidx0E,
1136                                             vdwioffsetptr0+vdwjidx0F,
1137                                             vdwioffsetptr0+vdwjidx0G,
1138                                             vdwioffsetptr0+vdwjidx0H,
1139                                             &c6_00,&c12_00);
1140
1141             /* REACTION-FIELD ELECTROSTATICS */
1142             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
1143
1144             /* LENNARD-JONES DISPERSION/REPULSION */
1145
1146             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1147             fvdw             = _mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),c6_00),_mm256_mul_ps(rinvsix,rinvsq00));
1148
1149             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
1150
1151             fscal            = _mm256_add_ps(felec,fvdw);
1152
1153             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1154
1155             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1156
1157             /* Calculate temporary vectorial force */
1158             tx               = _mm256_mul_ps(fscal,dx00);
1159             ty               = _mm256_mul_ps(fscal,dy00);
1160             tz               = _mm256_mul_ps(fscal,dz00);
1161
1162             /* Update vectorial force */
1163             fix0             = _mm256_add_ps(fix0,tx);
1164             fiy0             = _mm256_add_ps(fiy0,ty);
1165             fiz0             = _mm256_add_ps(fiz0,tz);
1166
1167             fjx0             = _mm256_add_ps(fjx0,tx);
1168             fjy0             = _mm256_add_ps(fjy0,ty);
1169             fjz0             = _mm256_add_ps(fjz0,tz);
1170
1171             }
1172
1173             /**************************
1174              * CALCULATE INTERACTIONS *
1175              **************************/
1176
1177             if (gmx_mm256_any_lt(rsq10,rcutoff2))
1178             {
1179
1180             /* Compute parameters for interactions between i and j atoms */
1181             qq10             = _mm256_mul_ps(iq1,jq0);
1182
1183             /* REACTION-FIELD ELECTROSTATICS */
1184             felec            = _mm256_mul_ps(qq10,_mm256_sub_ps(_mm256_mul_ps(rinv10,rinvsq10),krf2));
1185
1186             cutoff_mask      = _mm256_cmp_ps(rsq10,rcutoff2,_CMP_LT_OQ);
1187
1188             fscal            = felec;
1189
1190             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1191
1192             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1193
1194             /* Calculate temporary vectorial force */
1195             tx               = _mm256_mul_ps(fscal,dx10);
1196             ty               = _mm256_mul_ps(fscal,dy10);
1197             tz               = _mm256_mul_ps(fscal,dz10);
1198
1199             /* Update vectorial force */
1200             fix1             = _mm256_add_ps(fix1,tx);
1201             fiy1             = _mm256_add_ps(fiy1,ty);
1202             fiz1             = _mm256_add_ps(fiz1,tz);
1203
1204             fjx0             = _mm256_add_ps(fjx0,tx);
1205             fjy0             = _mm256_add_ps(fjy0,ty);
1206             fjz0             = _mm256_add_ps(fjz0,tz);
1207
1208             }
1209
1210             /**************************
1211              * CALCULATE INTERACTIONS *
1212              **************************/
1213
1214             if (gmx_mm256_any_lt(rsq20,rcutoff2))
1215             {
1216
1217             /* Compute parameters for interactions between i and j atoms */
1218             qq20             = _mm256_mul_ps(iq2,jq0);
1219
1220             /* REACTION-FIELD ELECTROSTATICS */
1221             felec            = _mm256_mul_ps(qq20,_mm256_sub_ps(_mm256_mul_ps(rinv20,rinvsq20),krf2));
1222
1223             cutoff_mask      = _mm256_cmp_ps(rsq20,rcutoff2,_CMP_LT_OQ);
1224
1225             fscal            = felec;
1226
1227             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1228
1229             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1230
1231             /* Calculate temporary vectorial force */
1232             tx               = _mm256_mul_ps(fscal,dx20);
1233             ty               = _mm256_mul_ps(fscal,dy20);
1234             tz               = _mm256_mul_ps(fscal,dz20);
1235
1236             /* Update vectorial force */
1237             fix2             = _mm256_add_ps(fix2,tx);
1238             fiy2             = _mm256_add_ps(fiy2,ty);
1239             fiz2             = _mm256_add_ps(fiz2,tz);
1240
1241             fjx0             = _mm256_add_ps(fjx0,tx);
1242             fjy0             = _mm256_add_ps(fjy0,ty);
1243             fjz0             = _mm256_add_ps(fjz0,tz);
1244
1245             }
1246
1247             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1248             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1249             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1250             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1251             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
1252             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
1253             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
1254             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
1255
1256             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,fjx0,fjy0,fjz0);
1257
1258             /* Inner loop uses 100 flops */
1259         }
1260
1261         /* End of innermost loop */
1262
1263         gmx_mm256_update_iforce_3atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1264                                                  f+i_coord_offset,fshift+i_shift_offset);
1265
1266         /* Increment number of inner iterations */
1267         inneriter                  += j_index_end - j_index_start;
1268
1269         /* Outer loop uses 18 flops */
1270     }
1271
1272     /* Increment number of outer iterations */
1273     outeriter        += nri;
1274
1275     /* Update outer/inner flops */
1276
1277     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*18 + inneriter*100);
1278 }