src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecEwSh_VdwLJSh_GeomW3P1_avx_256_double.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014,2015,2017, 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_double kernel generator.
  37  */
  38 #include "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/gmxlib/nrnb.h"
  46
  47 #include "kernelutil_x86_avx_256_double.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwLJSh_GeomW3P1_VF_avx_256_double
  51  * Electrostatics interaction: Ewald
  52  * VdW interaction:            LennardJones
  53  * Geometry:                   Water3-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecEwSh_VdwLJSh_GeomW3P1_VF_avx_256_double
  58                     (t_nblist                    * gmx_restrict       nlist,
  59                      rvec                        * gmx_restrict          xx,
  60                      rvec                        * gmx_restrict          ff,
  61                      struct t_forcerec           * gmx_restrict          fr,
  62                      t_mdatoms                   * gmx_restrict     mdatoms,
  63                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  64                      t_nrnb                      * gmx_restrict        nrnb)
  65 {
  66     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  67      * just 0 for non-waters.
  68      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
  69      * jnr indices corresponding to data put in the four positions in the SIMD register.
  70      */
  71     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  72     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  73     int              jnrA,jnrB,jnrC,jnrD;
  74     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  75     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  76     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  77     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  78     real             rcutoff_scalar;
  79     real             *shiftvec,*fshift,*x,*f;
  80     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  81     real             scratch[4*DIM];
  82     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  83     real *           vdwioffsetptr0;
  84     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  85     real *           vdwioffsetptr1;
  86     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  87     real *           vdwioffsetptr2;
  88     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  89     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  90     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  91     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  92     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  93     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  94     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     int              nvdwtype;
  97     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  98     int              *vdwtype;
  99     real             *vdwparam;
 100     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
 101     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
 102     __m128i          ewitab;
 103     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 104     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 105     real             *ewtab;
 106     __m256d          dummy_mask,cutoff_mask;
 107     __m128           tmpmask0,tmpmask1;
 108     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 109     __m256d          one     = _mm256_set1_pd(1.0);
 110     __m256d          two     = _mm256_set1_pd(2.0);
 111     x                = xx[0];
 112     f                = ff[0];
 113
 114     nri              = nlist->nri;
 115     iinr             = nlist->iinr;
 116     jindex           = nlist->jindex;
 117     jjnr             = nlist->jjnr;
 118     shiftidx         = nlist->shift;
 119     gid              = nlist->gid;
 120     shiftvec         = fr->shift_vec[0];
 121     fshift           = fr->fshift[0];
 122     facel            = _mm256_set1_pd(fr->ic->epsfac);
 123     charge           = mdatoms->chargeA;
 124     nvdwtype         = fr->ntype;
 125     vdwparam         = fr->nbfp;
 126     vdwtype          = mdatoms->typeA;
 127
 128     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 129     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
 130     beta2            = _mm256_mul_pd(beta,beta);
 131     beta3            = _mm256_mul_pd(beta,beta2);
 132
 133     ewtab            = fr->ic->tabq_coul_FDV0;
 134     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 135     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 136
 137     /* Setup water-specific parameters */
 138     inr              = nlist->iinr[0];
 139     iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
 140     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 141     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 142     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 143
 144     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 145     rcutoff_scalar   = fr->ic->rcoulomb;
 146     rcutoff          = _mm256_set1_pd(rcutoff_scalar);
 147     rcutoff2         = _mm256_mul_pd(rcutoff,rcutoff);
 148
 149     sh_vdw_invrcut6  = _mm256_set1_pd(fr->ic->sh_invrc6);
 150     rvdw             = _mm256_set1_pd(fr->ic->rvdw);
 151
 152     /* Avoid stupid compiler warnings */
 153     jnrA = jnrB = jnrC = jnrD = 0;
 154     j_coord_offsetA = 0;
 155     j_coord_offsetB = 0;
 156     j_coord_offsetC = 0;
 157     j_coord_offsetD = 0;
 158
 159     outeriter        = 0;
 160     inneriter        = 0;
 161
 162     for(iidx=0;iidx<4*DIM;iidx++)
 163     {
 164         scratch[iidx] = 0.0;
 165     }
 166
 167     /* Start outer loop over neighborlists */
 168     for(iidx=0; iidx<nri; iidx++)
 169     {
 170         /* Load shift vector for this list */
 171         i_shift_offset   = DIM*shiftidx[iidx];
 172
 173         /* Load limits for loop over neighbors */
 174         j_index_start    = jindex[iidx];
 175         j_index_end      = jindex[iidx+1];
 176
 177         /* Get outer coordinate index */
 178         inr              = iinr[iidx];
 179         i_coord_offset   = DIM*inr;
 180
 181         /* Load i particle coords and add shift vector */
 182         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 183                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 184
 185         fix0             = _mm256_setzero_pd();
 186         fiy0             = _mm256_setzero_pd();
 187         fiz0             = _mm256_setzero_pd();
 188         fix1             = _mm256_setzero_pd();
 189         fiy1             = _mm256_setzero_pd();
 190         fiz1             = _mm256_setzero_pd();
 191         fix2             = _mm256_setzero_pd();
 192         fiy2             = _mm256_setzero_pd();
 193         fiz2             = _mm256_setzero_pd();
 194
 195         /* Reset potential sums */
 196         velecsum         = _mm256_setzero_pd();
 197         vvdwsum          = _mm256_setzero_pd();
 198
 199         /* Start inner kernel loop */
 200         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 201         {
 202
 203             /* Get j neighbor index, and coordinate index */
 204             jnrA             = jjnr[jidx];
 205             jnrB             = jjnr[jidx+1];
 206             jnrC             = jjnr[jidx+2];
 207             jnrD             = jjnr[jidx+3];
 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
 213             /* load j atom coordinates */
 214             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 215                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 216                                                  &jx0,&jy0,&jz0);
 217
 218             /* Calculate displacement vector */
 219             dx00             = _mm256_sub_pd(ix0,jx0);
 220             dy00             = _mm256_sub_pd(iy0,jy0);
 221             dz00             = _mm256_sub_pd(iz0,jz0);
 222             dx10             = _mm256_sub_pd(ix1,jx0);
 223             dy10             = _mm256_sub_pd(iy1,jy0);
 224             dz10             = _mm256_sub_pd(iz1,jz0);
 225             dx20             = _mm256_sub_pd(ix2,jx0);
 226             dy20             = _mm256_sub_pd(iy2,jy0);
 227             dz20             = _mm256_sub_pd(iz2,jz0);
 228
 229             /* Calculate squared distance and things based on it */
 230             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 231             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 232             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 233
 234             rinv00           = avx256_invsqrt_d(rsq00);
 235             rinv10           = avx256_invsqrt_d(rsq10);
 236             rinv20           = avx256_invsqrt_d(rsq20);
 237
 238             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 239             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 240             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 241
 242             /* Load parameters for j particles */
 243             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 244                                                                  charge+jnrC+0,charge+jnrD+0);
 245             vdwjidx0A        = 2*vdwtype[jnrA+0];
 246             vdwjidx0B        = 2*vdwtype[jnrB+0];
 247             vdwjidx0C        = 2*vdwtype[jnrC+0];
 248             vdwjidx0D        = 2*vdwtype[jnrD+0];
 249
 250             fjx0             = _mm256_setzero_pd();
 251             fjy0             = _mm256_setzero_pd();
 252             fjz0             = _mm256_setzero_pd();
 253
 254             /**************************
 255              * CALCULATE INTERACTIONS *
 256              **************************/
 257
 258             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 259             {
 260
 261             r00              = _mm256_mul_pd(rsq00,rinv00);
 262
 263             /* Compute parameters for interactions between i and j atoms */
 264             qq00             = _mm256_mul_pd(iq0,jq0);
 265             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 266                                             vdwioffsetptr0+vdwjidx0B,
 267                                             vdwioffsetptr0+vdwjidx0C,
 268                                             vdwioffsetptr0+vdwjidx0D,
 269                                             &c6_00,&c12_00);
 270
 271             /* EWALD ELECTROSTATICS */
 272
 273             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 274             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 275             ewitab           = _mm256_cvttpd_epi32(ewrt);
 276             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 277             ewitab           = _mm_slli_epi32(ewitab,2);
 278             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 279             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 280             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 281             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 282             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 283             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 284             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 285             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(_mm256_sub_pd(rinv00,sh_ewald),velec));
 286             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 287
 288             /* LENNARD-JONES DISPERSION/REPULSION */
 289
 290             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 291             vvdw6            = _mm256_mul_pd(c6_00,rinvsix);
 292             vvdw12           = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
 293             vvdw             = _mm256_sub_pd(_mm256_mul_pd( _mm256_sub_pd(vvdw12 , _mm256_mul_pd(c12_00,_mm256_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
 294                                           _mm256_mul_pd( _mm256_sub_pd(vvdw6,_mm256_mul_pd(c6_00,sh_vdw_invrcut6)),one_sixth));
 295             fvdw             = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
 296
 297             cutoff_mask      = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
 298
 299             /* Update potential sum for this i atom from the interaction with this j atom. */
 300             velec            = _mm256_and_pd(velec,cutoff_mask);
 301             velecsum         = _mm256_add_pd(velecsum,velec);
 302             vvdw             = _mm256_and_pd(vvdw,cutoff_mask);
 303             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 304
 305             fscal            = _mm256_add_pd(felec,fvdw);
 306
 307             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 308
 309             /* Calculate temporary vectorial force */
 310             tx               = _mm256_mul_pd(fscal,dx00);
 311             ty               = _mm256_mul_pd(fscal,dy00);
 312             tz               = _mm256_mul_pd(fscal,dz00);
 313
 314             /* Update vectorial force */
 315             fix0             = _mm256_add_pd(fix0,tx);
 316             fiy0             = _mm256_add_pd(fiy0,ty);
 317             fiz0             = _mm256_add_pd(fiz0,tz);
 318
 319             fjx0             = _mm256_add_pd(fjx0,tx);
 320             fjy0             = _mm256_add_pd(fjy0,ty);
 321             fjz0             = _mm256_add_pd(fjz0,tz);
 322
 323             }
 324
 325             /**************************
 326              * CALCULATE INTERACTIONS *
 327              **************************/
 328
 329             if (gmx_mm256_any_lt(rsq10,rcutoff2))
 330             {
 331
 332             r10              = _mm256_mul_pd(rsq10,rinv10);
 333
 334             /* Compute parameters for interactions between i and j atoms */
 335             qq10             = _mm256_mul_pd(iq1,jq0);
 336
 337             /* EWALD ELECTROSTATICS */
 338
 339             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 340             ewrt             = _mm256_mul_pd(r10,ewtabscale);
 341             ewitab           = _mm256_cvttpd_epi32(ewrt);
 342             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 343             ewitab           = _mm_slli_epi32(ewitab,2);
 344             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 345             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 346             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 347             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 348             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 349             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 350             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 351             velec            = _mm256_mul_pd(qq10,_mm256_sub_pd(_mm256_sub_pd(rinv10,sh_ewald),velec));
 352             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
 353
 354             cutoff_mask      = _mm256_cmp_pd(rsq10,rcutoff2,_CMP_LT_OQ);
 355
 356             /* Update potential sum for this i atom from the interaction with this j atom. */
 357             velec            = _mm256_and_pd(velec,cutoff_mask);
 358             velecsum         = _mm256_add_pd(velecsum,velec);
 359
 360             fscal            = felec;
 361
 362             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 363
 364             /* Calculate temporary vectorial force */
 365             tx               = _mm256_mul_pd(fscal,dx10);
 366             ty               = _mm256_mul_pd(fscal,dy10);
 367             tz               = _mm256_mul_pd(fscal,dz10);
 368
 369             /* Update vectorial force */
 370             fix1             = _mm256_add_pd(fix1,tx);
 371             fiy1             = _mm256_add_pd(fiy1,ty);
 372             fiz1             = _mm256_add_pd(fiz1,tz);
 373
 374             fjx0             = _mm256_add_pd(fjx0,tx);
 375             fjy0             = _mm256_add_pd(fjy0,ty);
 376             fjz0             = _mm256_add_pd(fjz0,tz);
 377
 378             }
 379
 380             /**************************
 381              * CALCULATE INTERACTIONS *
 382              **************************/
 383
 384             if (gmx_mm256_any_lt(rsq20,rcutoff2))
 385             {
 386
 387             r20              = _mm256_mul_pd(rsq20,rinv20);
 388
 389             /* Compute parameters for interactions between i and j atoms */
 390             qq20             = _mm256_mul_pd(iq2,jq0);
 391
 392             /* EWALD ELECTROSTATICS */
 393
 394             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 395             ewrt             = _mm256_mul_pd(r20,ewtabscale);
 396             ewitab           = _mm256_cvttpd_epi32(ewrt);
 397             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 398             ewitab           = _mm_slli_epi32(ewitab,2);
 399             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 400             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 401             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 402             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 403             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 404             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 405             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 406             velec            = _mm256_mul_pd(qq20,_mm256_sub_pd(_mm256_sub_pd(rinv20,sh_ewald),velec));
 407             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
 408
 409             cutoff_mask      = _mm256_cmp_pd(rsq20,rcutoff2,_CMP_LT_OQ);
 410
 411             /* Update potential sum for this i atom from the interaction with this j atom. */
 412             velec            = _mm256_and_pd(velec,cutoff_mask);
 413             velecsum         = _mm256_add_pd(velecsum,velec);
 414
 415             fscal            = felec;
 416
 417             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 418
 419             /* Calculate temporary vectorial force */
 420             tx               = _mm256_mul_pd(fscal,dx20);
 421             ty               = _mm256_mul_pd(fscal,dy20);
 422             tz               = _mm256_mul_pd(fscal,dz20);
 423
 424             /* Update vectorial force */
 425             fix2             = _mm256_add_pd(fix2,tx);
 426             fiy2             = _mm256_add_pd(fiy2,ty);
 427             fiz2             = _mm256_add_pd(fiz2,tz);
 428
 429             fjx0             = _mm256_add_pd(fjx0,tx);
 430             fjy0             = _mm256_add_pd(fjy0,ty);
 431             fjz0             = _mm256_add_pd(fjz0,tz);
 432
 433             }
 434
 435             fjptrA             = f+j_coord_offsetA;
 436             fjptrB             = f+j_coord_offsetB;
 437             fjptrC             = f+j_coord_offsetC;
 438             fjptrD             = f+j_coord_offsetD;
 439
 440             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 441
 442             /* Inner loop uses 159 flops */
 443         }
 444
 445         if(jidx<j_index_end)
 446         {
 447
 448             /* Get j neighbor index, and coordinate index */
 449             jnrlistA         = jjnr[jidx];
 450             jnrlistB         = jjnr[jidx+1];
 451             jnrlistC         = jjnr[jidx+2];
 452             jnrlistD         = jjnr[jidx+3];
 453             /* Sign of each element will be negative for non-real atoms.
 454              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 455              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 456              */
 457             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 458
 459             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 460             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 461             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 462
 463             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 464             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 465             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 466             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 467             j_coord_offsetA  = DIM*jnrA;
 468             j_coord_offsetB  = DIM*jnrB;
 469             j_coord_offsetC  = DIM*jnrC;
 470             j_coord_offsetD  = DIM*jnrD;
 471
 472             /* load j atom coordinates */
 473             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 474                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 475                                                  &jx0,&jy0,&jz0);
 476
 477             /* Calculate displacement vector */
 478             dx00             = _mm256_sub_pd(ix0,jx0);
 479             dy00             = _mm256_sub_pd(iy0,jy0);
 480             dz00             = _mm256_sub_pd(iz0,jz0);
 481             dx10             = _mm256_sub_pd(ix1,jx0);
 482             dy10             = _mm256_sub_pd(iy1,jy0);
 483             dz10             = _mm256_sub_pd(iz1,jz0);
 484             dx20             = _mm256_sub_pd(ix2,jx0);
 485             dy20             = _mm256_sub_pd(iy2,jy0);
 486             dz20             = _mm256_sub_pd(iz2,jz0);
 487
 488             /* Calculate squared distance and things based on it */
 489             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 490             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 491             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 492
 493             rinv00           = avx256_invsqrt_d(rsq00);
 494             rinv10           = avx256_invsqrt_d(rsq10);
 495             rinv20           = avx256_invsqrt_d(rsq20);
 496
 497             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 498             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 499             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 500
 501             /* Load parameters for j particles */
 502             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 503                                                                  charge+jnrC+0,charge+jnrD+0);
 504             vdwjidx0A        = 2*vdwtype[jnrA+0];
 505             vdwjidx0B        = 2*vdwtype[jnrB+0];
 506             vdwjidx0C        = 2*vdwtype[jnrC+0];
 507             vdwjidx0D        = 2*vdwtype[jnrD+0];
 508
 509             fjx0             = _mm256_setzero_pd();
 510             fjy0             = _mm256_setzero_pd();
 511             fjz0             = _mm256_setzero_pd();
 512
 513             /**************************
 514              * CALCULATE INTERACTIONS *
 515              **************************/
 516
 517             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 518             {
 519
 520             r00              = _mm256_mul_pd(rsq00,rinv00);
 521             r00              = _mm256_andnot_pd(dummy_mask,r00);
 522
 523             /* Compute parameters for interactions between i and j atoms */
 524             qq00             = _mm256_mul_pd(iq0,jq0);
 525             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 526                                             vdwioffsetptr0+vdwjidx0B,
 527                                             vdwioffsetptr0+vdwjidx0C,
 528                                             vdwioffsetptr0+vdwjidx0D,
 529                                             &c6_00,&c12_00);
 530
 531             /* EWALD ELECTROSTATICS */
 532
 533             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 534             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 535             ewitab           = _mm256_cvttpd_epi32(ewrt);
 536             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 537             ewitab           = _mm_slli_epi32(ewitab,2);
 538             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 539             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 540             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 541             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 542             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 543             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 544             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 545             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(_mm256_sub_pd(rinv00,sh_ewald),velec));
 546             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 547
 548             /* LENNARD-JONES DISPERSION/REPULSION */
 549
 550             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 551             vvdw6            = _mm256_mul_pd(c6_00,rinvsix);
 552             vvdw12           = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
 553             vvdw             = _mm256_sub_pd(_mm256_mul_pd( _mm256_sub_pd(vvdw12 , _mm256_mul_pd(c12_00,_mm256_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
 554                                           _mm256_mul_pd( _mm256_sub_pd(vvdw6,_mm256_mul_pd(c6_00,sh_vdw_invrcut6)),one_sixth));
 555             fvdw             = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
 556
 557             cutoff_mask      = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
 558
 559             /* Update potential sum for this i atom from the interaction with this j atom. */
 560             velec            = _mm256_and_pd(velec,cutoff_mask);
 561             velec            = _mm256_andnot_pd(dummy_mask,velec);
 562             velecsum         = _mm256_add_pd(velecsum,velec);
 563             vvdw             = _mm256_and_pd(vvdw,cutoff_mask);
 564             vvdw             = _mm256_andnot_pd(dummy_mask,vvdw);
 565             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 566
 567             fscal            = _mm256_add_pd(felec,fvdw);
 568
 569             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 570
 571             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 572
 573             /* Calculate temporary vectorial force */
 574             tx               = _mm256_mul_pd(fscal,dx00);
 575             ty               = _mm256_mul_pd(fscal,dy00);
 576             tz               = _mm256_mul_pd(fscal,dz00);
 577
 578             /* Update vectorial force */
 579             fix0             = _mm256_add_pd(fix0,tx);
 580             fiy0             = _mm256_add_pd(fiy0,ty);
 581             fiz0             = _mm256_add_pd(fiz0,tz);
 582
 583             fjx0             = _mm256_add_pd(fjx0,tx);
 584             fjy0             = _mm256_add_pd(fjy0,ty);
 585             fjz0             = _mm256_add_pd(fjz0,tz);
 586
 587             }
 588
 589             /**************************
 590              * CALCULATE INTERACTIONS *
 591              **************************/
 592
 593             if (gmx_mm256_any_lt(rsq10,rcutoff2))
 594             {
 595
 596             r10              = _mm256_mul_pd(rsq10,rinv10);
 597             r10              = _mm256_andnot_pd(dummy_mask,r10);
 598
 599             /* Compute parameters for interactions between i and j atoms */
 600             qq10             = _mm256_mul_pd(iq1,jq0);
 601
 602             /* EWALD ELECTROSTATICS */
 603
 604             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 605             ewrt             = _mm256_mul_pd(r10,ewtabscale);
 606             ewitab           = _mm256_cvttpd_epi32(ewrt);
 607             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 608             ewitab           = _mm_slli_epi32(ewitab,2);
 609             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 610             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 611             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 612             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 613             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 614             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 615             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 616             velec            = _mm256_mul_pd(qq10,_mm256_sub_pd(_mm256_sub_pd(rinv10,sh_ewald),velec));
 617             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
 618
 619             cutoff_mask      = _mm256_cmp_pd(rsq10,rcutoff2,_CMP_LT_OQ);
 620
 621             /* Update potential sum for this i atom from the interaction with this j atom. */
 622             velec            = _mm256_and_pd(velec,cutoff_mask);
 623             velec            = _mm256_andnot_pd(dummy_mask,velec);
 624             velecsum         = _mm256_add_pd(velecsum,velec);
 625
 626             fscal            = felec;
 627
 628             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 629
 630             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 631
 632             /* Calculate temporary vectorial force */
 633             tx               = _mm256_mul_pd(fscal,dx10);
 634             ty               = _mm256_mul_pd(fscal,dy10);
 635             tz               = _mm256_mul_pd(fscal,dz10);
 636
 637             /* Update vectorial force */
 638             fix1             = _mm256_add_pd(fix1,tx);
 639             fiy1             = _mm256_add_pd(fiy1,ty);
 640             fiz1             = _mm256_add_pd(fiz1,tz);
 641
 642             fjx0             = _mm256_add_pd(fjx0,tx);
 643             fjy0             = _mm256_add_pd(fjy0,ty);
 644             fjz0             = _mm256_add_pd(fjz0,tz);
 645
 646             }
 647
 648             /**************************
 649              * CALCULATE INTERACTIONS *
 650              **************************/
 651
 652             if (gmx_mm256_any_lt(rsq20,rcutoff2))
 653             {
 654
 655             r20              = _mm256_mul_pd(rsq20,rinv20);
 656             r20              = _mm256_andnot_pd(dummy_mask,r20);
 657
 658             /* Compute parameters for interactions between i and j atoms */
 659             qq20             = _mm256_mul_pd(iq2,jq0);
 660
 661             /* EWALD ELECTROSTATICS */
 662
 663             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 664             ewrt             = _mm256_mul_pd(r20,ewtabscale);
 665             ewitab           = _mm256_cvttpd_epi32(ewrt);
 666             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 667             ewitab           = _mm_slli_epi32(ewitab,2);
 668             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 669             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 670             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 671             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 672             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 673             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 674             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 675             velec            = _mm256_mul_pd(qq20,_mm256_sub_pd(_mm256_sub_pd(rinv20,sh_ewald),velec));
 676             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
 677
 678             cutoff_mask      = _mm256_cmp_pd(rsq20,rcutoff2,_CMP_LT_OQ);
 679
 680             /* Update potential sum for this i atom from the interaction with this j atom. */
 681             velec            = _mm256_and_pd(velec,cutoff_mask);
 682             velec            = _mm256_andnot_pd(dummy_mask,velec);
 683             velecsum         = _mm256_add_pd(velecsum,velec);
 684
 685             fscal            = felec;
 686
 687             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 688
 689             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 690
 691             /* Calculate temporary vectorial force */
 692             tx               = _mm256_mul_pd(fscal,dx20);
 693             ty               = _mm256_mul_pd(fscal,dy20);
 694             tz               = _mm256_mul_pd(fscal,dz20);
 695
 696             /* Update vectorial force */
 697             fix2             = _mm256_add_pd(fix2,tx);
 698             fiy2             = _mm256_add_pd(fiy2,ty);
 699             fiz2             = _mm256_add_pd(fiz2,tz);
 700
 701             fjx0             = _mm256_add_pd(fjx0,tx);
 702             fjy0             = _mm256_add_pd(fjy0,ty);
 703             fjz0             = _mm256_add_pd(fjz0,tz);
 704
 705             }
 706
 707             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 708             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 709             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 710             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 711
 712             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 713
 714             /* Inner loop uses 162 flops */
 715         }
 716
 717         /* End of innermost loop */
 718
 719         gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
 720                                                  f+i_coord_offset,fshift+i_shift_offset);
 721
 722         ggid                        = gid[iidx];
 723         /* Update potential energies */
 724         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 725         gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
 726
 727         /* Increment number of inner iterations */
 728         inneriter                  += j_index_end - j_index_start;
 729
 730         /* Outer loop uses 20 flops */
 731     }
 732
 733     /* Increment number of outer iterations */
 734     outeriter        += nri;
 735
 736     /* Update outer/inner flops */
 737
 738     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*20 + inneriter*162);
 739 }
 740 /*
 741  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwLJSh_GeomW3P1_F_avx_256_double
 742  * Electrostatics interaction: Ewald
 743  * VdW interaction:            LennardJones
 744  * Geometry:                   Water3-Particle
 745  * Calculate force/pot:        Force
 746  */
 747 void
 748 nb_kernel_ElecEwSh_VdwLJSh_GeomW3P1_F_avx_256_double
 749                     (t_nblist                    * gmx_restrict       nlist,
 750                      rvec                        * gmx_restrict          xx,
 751                      rvec                        * gmx_restrict          ff,
 752                      struct t_forcerec           * gmx_restrict          fr,
 753                      t_mdatoms                   * gmx_restrict     mdatoms,
 754                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 755                      t_nrnb                      * gmx_restrict        nrnb)
 756 {
 757     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 758      * just 0 for non-waters.
 759      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
 760      * jnr indices corresponding to data put in the four positions in the SIMD register.
 761      */
 762     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 763     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 764     int              jnrA,jnrB,jnrC,jnrD;
 765     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 766     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 767     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 768     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 769     real             rcutoff_scalar;
 770     real             *shiftvec,*fshift,*x,*f;
 771     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 772     real             scratch[4*DIM];
 773     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 774     real *           vdwioffsetptr0;
 775     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 776     real *           vdwioffsetptr1;
 777     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 778     real *           vdwioffsetptr2;
 779     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 780     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 781     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 782     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 783     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 784     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 785     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 786     real             *charge;
 787     int              nvdwtype;
 788     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 789     int              *vdwtype;
 790     real             *vdwparam;
 791     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
 792     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
 793     __m128i          ewitab;
 794     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 795     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 796     real             *ewtab;
 797     __m256d          dummy_mask,cutoff_mask;
 798     __m128           tmpmask0,tmpmask1;
 799     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 800     __m256d          one     = _mm256_set1_pd(1.0);
 801     __m256d          two     = _mm256_set1_pd(2.0);
 802     x                = xx[0];
 803     f                = ff[0];
 804
 805     nri              = nlist->nri;
 806     iinr             = nlist->iinr;
 807     jindex           = nlist->jindex;
 808     jjnr             = nlist->jjnr;
 809     shiftidx         = nlist->shift;
 810     gid              = nlist->gid;
 811     shiftvec         = fr->shift_vec[0];
 812     fshift           = fr->fshift[0];
 813     facel            = _mm256_set1_pd(fr->ic->epsfac);
 814     charge           = mdatoms->chargeA;
 815     nvdwtype         = fr->ntype;
 816     vdwparam         = fr->nbfp;
 817     vdwtype          = mdatoms->typeA;
 818
 819     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 820     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
 821     beta2            = _mm256_mul_pd(beta,beta);
 822     beta3            = _mm256_mul_pd(beta,beta2);
 823
 824     ewtab            = fr->ic->tabq_coul_F;
 825     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 826     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 827
 828     /* Setup water-specific parameters */
 829     inr              = nlist->iinr[0];
 830     iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
 831     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 832     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 833     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 834
 835     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 836     rcutoff_scalar   = fr->ic->rcoulomb;
 837     rcutoff          = _mm256_set1_pd(rcutoff_scalar);
 838     rcutoff2         = _mm256_mul_pd(rcutoff,rcutoff);
 839
 840     sh_vdw_invrcut6  = _mm256_set1_pd(fr->ic->sh_invrc6);
 841     rvdw             = _mm256_set1_pd(fr->ic->rvdw);
 842
 843     /* Avoid stupid compiler warnings */
 844     jnrA = jnrB = jnrC = jnrD = 0;
 845     j_coord_offsetA = 0;
 846     j_coord_offsetB = 0;
 847     j_coord_offsetC = 0;
 848     j_coord_offsetD = 0;
 849
 850     outeriter        = 0;
 851     inneriter        = 0;
 852
 853     for(iidx=0;iidx<4*DIM;iidx++)
 854     {
 855         scratch[iidx] = 0.0;
 856     }
 857
 858     /* Start outer loop over neighborlists */
 859     for(iidx=0; iidx<nri; iidx++)
 860     {
 861         /* Load shift vector for this list */
 862         i_shift_offset   = DIM*shiftidx[iidx];
 863
 864         /* Load limits for loop over neighbors */
 865         j_index_start    = jindex[iidx];
 866         j_index_end      = jindex[iidx+1];
 867
 868         /* Get outer coordinate index */
 869         inr              = iinr[iidx];
 870         i_coord_offset   = DIM*inr;
 871
 872         /* Load i particle coords and add shift vector */
 873         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 874                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 875
 876         fix0             = _mm256_setzero_pd();
 877         fiy0             = _mm256_setzero_pd();
 878         fiz0             = _mm256_setzero_pd();
 879         fix1             = _mm256_setzero_pd();
 880         fiy1             = _mm256_setzero_pd();
 881         fiz1             = _mm256_setzero_pd();
 882         fix2             = _mm256_setzero_pd();
 883         fiy2             = _mm256_setzero_pd();
 884         fiz2             = _mm256_setzero_pd();
 885
 886         /* Start inner kernel loop */
 887         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 888         {
 889
 890             /* Get j neighbor index, and coordinate index */
 891             jnrA             = jjnr[jidx];
 892             jnrB             = jjnr[jidx+1];
 893             jnrC             = jjnr[jidx+2];
 894             jnrD             = jjnr[jidx+3];
 895             j_coord_offsetA  = DIM*jnrA;
 896             j_coord_offsetB  = DIM*jnrB;
 897             j_coord_offsetC  = DIM*jnrC;
 898             j_coord_offsetD  = DIM*jnrD;
 899
 900             /* load j atom coordinates */
 901             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 902                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 903                                                  &jx0,&jy0,&jz0);
 904
 905             /* Calculate displacement vector */
 906             dx00             = _mm256_sub_pd(ix0,jx0);
 907             dy00             = _mm256_sub_pd(iy0,jy0);
 908             dz00             = _mm256_sub_pd(iz0,jz0);
 909             dx10             = _mm256_sub_pd(ix1,jx0);
 910             dy10             = _mm256_sub_pd(iy1,jy0);
 911             dz10             = _mm256_sub_pd(iz1,jz0);
 912             dx20             = _mm256_sub_pd(ix2,jx0);
 913             dy20             = _mm256_sub_pd(iy2,jy0);
 914             dz20             = _mm256_sub_pd(iz2,jz0);
 915
 916             /* Calculate squared distance and things based on it */
 917             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 918             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 919             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 920
 921             rinv00           = avx256_invsqrt_d(rsq00);
 922             rinv10           = avx256_invsqrt_d(rsq10);
 923             rinv20           = avx256_invsqrt_d(rsq20);
 924
 925             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 926             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 927             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 928
 929             /* Load parameters for j particles */
 930             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 931                                                                  charge+jnrC+0,charge+jnrD+0);
 932             vdwjidx0A        = 2*vdwtype[jnrA+0];
 933             vdwjidx0B        = 2*vdwtype[jnrB+0];
 934             vdwjidx0C        = 2*vdwtype[jnrC+0];
 935             vdwjidx0D        = 2*vdwtype[jnrD+0];
 936
 937             fjx0             = _mm256_setzero_pd();
 938             fjy0             = _mm256_setzero_pd();
 939             fjz0             = _mm256_setzero_pd();
 940
 941             /**************************
 942              * CALCULATE INTERACTIONS *
 943              **************************/
 944
 945             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 946             {
 947
 948             r00              = _mm256_mul_pd(rsq00,rinv00);
 949
 950             /* Compute parameters for interactions between i and j atoms */
 951             qq00             = _mm256_mul_pd(iq0,jq0);
 952             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 953                                             vdwioffsetptr0+vdwjidx0B,
 954                                             vdwioffsetptr0+vdwjidx0C,
 955                                             vdwioffsetptr0+vdwjidx0D,
 956                                             &c6_00,&c12_00);
 957
 958             /* EWALD ELECTROSTATICS */
 959
 960             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 961             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 962             ewitab           = _mm256_cvttpd_epi32(ewrt);
 963             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 964             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
 965                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
 966                                             &ewtabF,&ewtabFn);
 967             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
 968             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 969
 970             /* LENNARD-JONES DISPERSION/REPULSION */
 971
 972             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 973             fvdw             = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
 974
 975             cutoff_mask      = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
 976
 977             fscal            = _mm256_add_pd(felec,fvdw);
 978
 979             fscal            = _mm256_and_pd(fscal,cutoff_mask);
 980
 981             /* Calculate temporary vectorial force */
 982             tx               = _mm256_mul_pd(fscal,dx00);
 983             ty               = _mm256_mul_pd(fscal,dy00);
 984             tz               = _mm256_mul_pd(fscal,dz00);
 985
 986             /* Update vectorial force */
 987             fix0             = _mm256_add_pd(fix0,tx);
 988             fiy0             = _mm256_add_pd(fiy0,ty);
 989             fiz0             = _mm256_add_pd(fiz0,tz);
 990
 991             fjx0             = _mm256_add_pd(fjx0,tx);
 992             fjy0             = _mm256_add_pd(fjy0,ty);
 993             fjz0             = _mm256_add_pd(fjz0,tz);
 994
 995             }
 996
 997             /**************************
 998              * CALCULATE INTERACTIONS *
 999              **************************/
1000
1001             if (gmx_mm256_any_lt(rsq10,rcutoff2))
1002             {
1003
1004             r10              = _mm256_mul_pd(rsq10,rinv10);
1005
1006             /* Compute parameters for interactions between i and j atoms */
1007             qq10             = _mm256_mul_pd(iq1,jq0);
1008
1009             /* EWALD ELECTROSTATICS */
1010
1011             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1012             ewrt             = _mm256_mul_pd(r10,ewtabscale);
1013             ewitab           = _mm256_cvttpd_epi32(ewrt);
1014             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1015             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1016                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1017                                             &ewtabF,&ewtabFn);
1018             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1019             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
1020
1021             cutoff_mask      = _mm256_cmp_pd(rsq10,rcutoff2,_CMP_LT_OQ);
1022
1023             fscal            = felec;
1024
1025             fscal            = _mm256_and_pd(fscal,cutoff_mask);
1026
1027             /* Calculate temporary vectorial force */
1028             tx               = _mm256_mul_pd(fscal,dx10);
1029             ty               = _mm256_mul_pd(fscal,dy10);
1030             tz               = _mm256_mul_pd(fscal,dz10);
1031
1032             /* Update vectorial force */
1033             fix1             = _mm256_add_pd(fix1,tx);
1034             fiy1             = _mm256_add_pd(fiy1,ty);
1035             fiz1             = _mm256_add_pd(fiz1,tz);
1036
1037             fjx0             = _mm256_add_pd(fjx0,tx);
1038             fjy0             = _mm256_add_pd(fjy0,ty);
1039             fjz0             = _mm256_add_pd(fjz0,tz);
1040
1041             }
1042
1043             /**************************
1044              * CALCULATE INTERACTIONS *
1045              **************************/
1046
1047             if (gmx_mm256_any_lt(rsq20,rcutoff2))
1048             {
1049
1050             r20              = _mm256_mul_pd(rsq20,rinv20);
1051
1052             /* Compute parameters for interactions between i and j atoms */
1053             qq20             = _mm256_mul_pd(iq2,jq0);
1054
1055             /* EWALD ELECTROSTATICS */
1056
1057             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1058             ewrt             = _mm256_mul_pd(r20,ewtabscale);
1059             ewitab           = _mm256_cvttpd_epi32(ewrt);
1060             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1061             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1062                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1063                                             &ewtabF,&ewtabFn);
1064             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1065             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
1066
1067             cutoff_mask      = _mm256_cmp_pd(rsq20,rcutoff2,_CMP_LT_OQ);
1068
1069             fscal            = felec;
1070
1071             fscal            = _mm256_and_pd(fscal,cutoff_mask);
1072
1073             /* Calculate temporary vectorial force */
1074             tx               = _mm256_mul_pd(fscal,dx20);
1075             ty               = _mm256_mul_pd(fscal,dy20);
1076             tz               = _mm256_mul_pd(fscal,dz20);
1077
1078             /* Update vectorial force */
1079             fix2             = _mm256_add_pd(fix2,tx);
1080             fiy2             = _mm256_add_pd(fiy2,ty);
1081             fiz2             = _mm256_add_pd(fiz2,tz);
1082
1083             fjx0             = _mm256_add_pd(fjx0,tx);
1084             fjy0             = _mm256_add_pd(fjy0,ty);
1085             fjz0             = _mm256_add_pd(fjz0,tz);
1086
1087             }
1088
1089             fjptrA             = f+j_coord_offsetA;
1090             fjptrB             = f+j_coord_offsetB;
1091             fjptrC             = f+j_coord_offsetC;
1092             fjptrD             = f+j_coord_offsetD;
1093
1094             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1095
1096             /* Inner loop uses 127 flops */
1097         }
1098
1099         if(jidx<j_index_end)
1100         {
1101
1102             /* Get j neighbor index, and coordinate index */
1103             jnrlistA         = jjnr[jidx];
1104             jnrlistB         = jjnr[jidx+1];
1105             jnrlistC         = jjnr[jidx+2];
1106             jnrlistD         = jjnr[jidx+3];
1107             /* Sign of each element will be negative for non-real atoms.
1108              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1109              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1110              */
1111             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1112
1113             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1114             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1115             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1116
1117             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1118             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1119             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1120             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1121             j_coord_offsetA  = DIM*jnrA;
1122             j_coord_offsetB  = DIM*jnrB;
1123             j_coord_offsetC  = DIM*jnrC;
1124             j_coord_offsetD  = DIM*jnrD;
1125
1126             /* load j atom coordinates */
1127             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1128                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1129                                                  &jx0,&jy0,&jz0);
1130
1131             /* Calculate displacement vector */
1132             dx00             = _mm256_sub_pd(ix0,jx0);
1133             dy00             = _mm256_sub_pd(iy0,jy0);
1134             dz00             = _mm256_sub_pd(iz0,jz0);
1135             dx10             = _mm256_sub_pd(ix1,jx0);
1136             dy10             = _mm256_sub_pd(iy1,jy0);
1137             dz10             = _mm256_sub_pd(iz1,jz0);
1138             dx20             = _mm256_sub_pd(ix2,jx0);
1139             dy20             = _mm256_sub_pd(iy2,jy0);
1140             dz20             = _mm256_sub_pd(iz2,jz0);
1141
1142             /* Calculate squared distance and things based on it */
1143             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1144             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1145             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1146
1147             rinv00           = avx256_invsqrt_d(rsq00);
1148             rinv10           = avx256_invsqrt_d(rsq10);
1149             rinv20           = avx256_invsqrt_d(rsq20);
1150
1151             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
1152             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
1153             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
1154
1155             /* Load parameters for j particles */
1156             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
1157                                                                  charge+jnrC+0,charge+jnrD+0);
1158             vdwjidx0A        = 2*vdwtype[jnrA+0];
1159             vdwjidx0B        = 2*vdwtype[jnrB+0];
1160             vdwjidx0C        = 2*vdwtype[jnrC+0];
1161             vdwjidx0D        = 2*vdwtype[jnrD+0];
1162
1163             fjx0             = _mm256_setzero_pd();
1164             fjy0             = _mm256_setzero_pd();
1165             fjz0             = _mm256_setzero_pd();
1166
1167             /**************************
1168              * CALCULATE INTERACTIONS *
1169              **************************/
1170
1171             if (gmx_mm256_any_lt(rsq00,rcutoff2))
1172             {
1173
1174             r00              = _mm256_mul_pd(rsq00,rinv00);
1175             r00              = _mm256_andnot_pd(dummy_mask,r00);
1176
1177             /* Compute parameters for interactions between i and j atoms */
1178             qq00             = _mm256_mul_pd(iq0,jq0);
1179             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
1180                                             vdwioffsetptr0+vdwjidx0B,
1181                                             vdwioffsetptr0+vdwjidx0C,
1182                                             vdwioffsetptr0+vdwjidx0D,
1183                                             &c6_00,&c12_00);
1184
1185             /* EWALD ELECTROSTATICS */
1186
1187             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1188             ewrt             = _mm256_mul_pd(r00,ewtabscale);
1189             ewitab           = _mm256_cvttpd_epi32(ewrt);
1190             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1191             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1192                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1193                                             &ewtabF,&ewtabFn);
1194             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1195             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
1196
1197             /* LENNARD-JONES DISPERSION/REPULSION */
1198
1199             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1200             fvdw             = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1201
1202             cutoff_mask      = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
1203
1204             fscal            = _mm256_add_pd(felec,fvdw);
1205
1206             fscal            = _mm256_and_pd(fscal,cutoff_mask);
1207
1208             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1209
1210             /* Calculate temporary vectorial force */
1211             tx               = _mm256_mul_pd(fscal,dx00);
1212             ty               = _mm256_mul_pd(fscal,dy00);
1213             tz               = _mm256_mul_pd(fscal,dz00);
1214
1215             /* Update vectorial force */
1216             fix0             = _mm256_add_pd(fix0,tx);
1217             fiy0             = _mm256_add_pd(fiy0,ty);
1218             fiz0             = _mm256_add_pd(fiz0,tz);
1219
1220             fjx0             = _mm256_add_pd(fjx0,tx);
1221             fjy0             = _mm256_add_pd(fjy0,ty);
1222             fjz0             = _mm256_add_pd(fjz0,tz);
1223
1224             }
1225
1226             /**************************
1227              * CALCULATE INTERACTIONS *
1228              **************************/
1229
1230             if (gmx_mm256_any_lt(rsq10,rcutoff2))
1231             {
1232
1233             r10              = _mm256_mul_pd(rsq10,rinv10);
1234             r10              = _mm256_andnot_pd(dummy_mask,r10);
1235
1236             /* Compute parameters for interactions between i and j atoms */
1237             qq10             = _mm256_mul_pd(iq1,jq0);
1238
1239             /* EWALD ELECTROSTATICS */
1240
1241             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1242             ewrt             = _mm256_mul_pd(r10,ewtabscale);
1243             ewitab           = _mm256_cvttpd_epi32(ewrt);
1244             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1245             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1246                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1247                                             &ewtabF,&ewtabFn);
1248             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1249             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
1250
1251             cutoff_mask      = _mm256_cmp_pd(rsq10,rcutoff2,_CMP_LT_OQ);
1252
1253             fscal            = felec;
1254
1255             fscal            = _mm256_and_pd(fscal,cutoff_mask);
1256
1257             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1258
1259             /* Calculate temporary vectorial force */
1260             tx               = _mm256_mul_pd(fscal,dx10);
1261             ty               = _mm256_mul_pd(fscal,dy10);
1262             tz               = _mm256_mul_pd(fscal,dz10);
1263
1264             /* Update vectorial force */
1265             fix1             = _mm256_add_pd(fix1,tx);
1266             fiy1             = _mm256_add_pd(fiy1,ty);
1267             fiz1             = _mm256_add_pd(fiz1,tz);
1268
1269             fjx0             = _mm256_add_pd(fjx0,tx);
1270             fjy0             = _mm256_add_pd(fjy0,ty);
1271             fjz0             = _mm256_add_pd(fjz0,tz);
1272
1273             }
1274
1275             /**************************
1276              * CALCULATE INTERACTIONS *
1277              **************************/
1278
1279             if (gmx_mm256_any_lt(rsq20,rcutoff2))
1280             {
1281
1282             r20              = _mm256_mul_pd(rsq20,rinv20);
1283             r20              = _mm256_andnot_pd(dummy_mask,r20);
1284
1285             /* Compute parameters for interactions between i and j atoms */
1286             qq20             = _mm256_mul_pd(iq2,jq0);
1287
1288             /* EWALD ELECTROSTATICS */
1289
1290             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1291             ewrt             = _mm256_mul_pd(r20,ewtabscale);
1292             ewitab           = _mm256_cvttpd_epi32(ewrt);
1293             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1294             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1295                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1296                                             &ewtabF,&ewtabFn);
1297             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1298             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
1299
1300             cutoff_mask      = _mm256_cmp_pd(rsq20,rcutoff2,_CMP_LT_OQ);
1301
1302             fscal            = felec;
1303
1304             fscal            = _mm256_and_pd(fscal,cutoff_mask);
1305
1306             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1307
1308             /* Calculate temporary vectorial force */
1309             tx               = _mm256_mul_pd(fscal,dx20);
1310             ty               = _mm256_mul_pd(fscal,dy20);
1311             tz               = _mm256_mul_pd(fscal,dz20);
1312
1313             /* Update vectorial force */
1314             fix2             = _mm256_add_pd(fix2,tx);
1315             fiy2             = _mm256_add_pd(fiy2,ty);
1316             fiz2             = _mm256_add_pd(fiz2,tz);
1317
1318             fjx0             = _mm256_add_pd(fjx0,tx);
1319             fjy0             = _mm256_add_pd(fjy0,ty);
1320             fjz0             = _mm256_add_pd(fjz0,tz);
1321
1322             }
1323
1324             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1325             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1326             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1327             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1328
1329             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1330
1331             /* Inner loop uses 130 flops */
1332         }
1333
1334         /* End of innermost loop */
1335
1336         gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1337                                                  f+i_coord_offset,fshift+i_shift_offset);
1338
1339         /* Increment number of inner iterations */
1340         inneriter                  += j_index_end - j_index_start;
1341
1342         /* Outer loop uses 18 flops */
1343     }
1344
1345     /* Increment number of outer iterations */
1346     outeriter        += nri;
1347
1348     /* Update outer/inner flops */
1349
1350     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*18 + inneriter*130);
1351 }