src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecCSTab_VdwLJ_GeomW4P1_avx_256_double.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_double kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_256_double.h"
  48 #include "kernelutil_x86_avx_256_double.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomW4P1_VF_avx_256_double
  52  * Electrostatics interaction: CubicSplineTable
  53  * VdW interaction:            LennardJones
  54  * Geometry:                   Water4-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecCSTab_VdwLJ_GeomW4P1_VF_avx_256_double
  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 refer to j loop unrolling done with AVX, e.g. for the four 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              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  76     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  77     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  78     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  79     real             rcutoff_scalar;
  80     real             *shiftvec,*fshift,*x,*f;
  81     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  82     real             scratch[4*DIM];
  83     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  84     real *           vdwioffsetptr0;
  85     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  86     real *           vdwioffsetptr1;
  87     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  88     real *           vdwioffsetptr2;
  89     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  90     real *           vdwioffsetptr3;
  91     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  92     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  93     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  94     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  95     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  96     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  97     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  98     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
  99     real             *charge;
 100     int              nvdwtype;
 101     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 102     int              *vdwtype;
 103     real             *vdwparam;
 104     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
 105     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
 106     __m128i          vfitab;
 107     __m128i          ifour       = _mm_set1_epi32(4);
 108     __m256d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 109     real             *vftab;
 110     __m256d          dummy_mask,cutoff_mask;
 111     __m128           tmpmask0,tmpmask1;
 112     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 113     __m256d          one     = _mm256_set1_pd(1.0);
 114     __m256d          two     = _mm256_set1_pd(2.0);
 115     x                = xx[0];
 116     f                = ff[0];
 117
 118     nri              = nlist->nri;
 119     iinr             = nlist->iinr;
 120     jindex           = nlist->jindex;
 121     jjnr             = nlist->jjnr;
 122     shiftidx         = nlist->shift;
 123     gid              = nlist->gid;
 124     shiftvec         = fr->shift_vec[0];
 125     fshift           = fr->fshift[0];
 126     facel            = _mm256_set1_pd(fr->epsfac);
 127     charge           = mdatoms->chargeA;
 128     nvdwtype         = fr->ntype;
 129     vdwparam         = fr->nbfp;
 130     vdwtype          = mdatoms->typeA;
 131
 132     vftab            = kernel_data->table_elec->data;
 133     vftabscale       = _mm256_set1_pd(kernel_data->table_elec->scale);
 134
 135     /* Setup water-specific parameters */
 136     inr              = nlist->iinr[0];
 137     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 138     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 139     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 140     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 141
 142     /* Avoid stupid compiler warnings */
 143     jnrA = jnrB = jnrC = jnrD = 0;
 144     j_coord_offsetA = 0;
 145     j_coord_offsetB = 0;
 146     j_coord_offsetC = 0;
 147     j_coord_offsetD = 0;
 148
 149     outeriter        = 0;
 150     inneriter        = 0;
 151
 152     for(iidx=0;iidx<4*DIM;iidx++)
 153     {
 154         scratch[iidx] = 0.0;
 155     }
 156
 157     /* Start outer loop over neighborlists */
 158     for(iidx=0; iidx<nri; iidx++)
 159     {
 160         /* Load shift vector for this list */
 161         i_shift_offset   = DIM*shiftidx[iidx];
 162
 163         /* Load limits for loop over neighbors */
 164         j_index_start    = jindex[iidx];
 165         j_index_end      = jindex[iidx+1];
 166
 167         /* Get outer coordinate index */
 168         inr              = iinr[iidx];
 169         i_coord_offset   = DIM*inr;
 170
 171         /* Load i particle coords and add shift vector */
 172         gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 173                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 174
 175         fix0             = _mm256_setzero_pd();
 176         fiy0             = _mm256_setzero_pd();
 177         fiz0             = _mm256_setzero_pd();
 178         fix1             = _mm256_setzero_pd();
 179         fiy1             = _mm256_setzero_pd();
 180         fiz1             = _mm256_setzero_pd();
 181         fix2             = _mm256_setzero_pd();
 182         fiy2             = _mm256_setzero_pd();
 183         fiz2             = _mm256_setzero_pd();
 184         fix3             = _mm256_setzero_pd();
 185         fiy3             = _mm256_setzero_pd();
 186         fiz3             = _mm256_setzero_pd();
 187
 188         /* Reset potential sums */
 189         velecsum         = _mm256_setzero_pd();
 190         vvdwsum          = _mm256_setzero_pd();
 191
 192         /* Start inner kernel loop */
 193         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 194         {
 195
 196             /* Get j neighbor index, and coordinate index */
 197             jnrA             = jjnr[jidx];
 198             jnrB             = jjnr[jidx+1];
 199             jnrC             = jjnr[jidx+2];
 200             jnrD             = jjnr[jidx+3];
 201             j_coord_offsetA  = DIM*jnrA;
 202             j_coord_offsetB  = DIM*jnrB;
 203             j_coord_offsetC  = DIM*jnrC;
 204             j_coord_offsetD  = DIM*jnrD;
 205
 206             /* load j atom coordinates */
 207             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 208                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 209                                                  &jx0,&jy0,&jz0);
 210
 211             /* Calculate displacement vector */
 212             dx00             = _mm256_sub_pd(ix0,jx0);
 213             dy00             = _mm256_sub_pd(iy0,jy0);
 214             dz00             = _mm256_sub_pd(iz0,jz0);
 215             dx10             = _mm256_sub_pd(ix1,jx0);
 216             dy10             = _mm256_sub_pd(iy1,jy0);
 217             dz10             = _mm256_sub_pd(iz1,jz0);
 218             dx20             = _mm256_sub_pd(ix2,jx0);
 219             dy20             = _mm256_sub_pd(iy2,jy0);
 220             dz20             = _mm256_sub_pd(iz2,jz0);
 221             dx30             = _mm256_sub_pd(ix3,jx0);
 222             dy30             = _mm256_sub_pd(iy3,jy0);
 223             dz30             = _mm256_sub_pd(iz3,jz0);
 224
 225             /* Calculate squared distance and things based on it */
 226             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 227             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 228             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 229             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 230
 231             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 232             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 233             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 234
 235             rinvsq00         = gmx_mm256_inv_pd(rsq00);
 236
 237             /* Load parameters for j particles */
 238             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 239                                                                  charge+jnrC+0,charge+jnrD+0);
 240             vdwjidx0A        = 2*vdwtype[jnrA+0];
 241             vdwjidx0B        = 2*vdwtype[jnrB+0];
 242             vdwjidx0C        = 2*vdwtype[jnrC+0];
 243             vdwjidx0D        = 2*vdwtype[jnrD+0];
 244
 245             fjx0             = _mm256_setzero_pd();
 246             fjy0             = _mm256_setzero_pd();
 247             fjz0             = _mm256_setzero_pd();
 248
 249             /**************************
 250              * CALCULATE INTERACTIONS *
 251              **************************/
 252
 253             /* Compute parameters for interactions between i and j atoms */
 254             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 255                                             vdwioffsetptr0+vdwjidx0B,
 256                                             vdwioffsetptr0+vdwjidx0C,
 257                                             vdwioffsetptr0+vdwjidx0D,
 258                                             &c6_00,&c12_00);
 259
 260             /* LENNARD-JONES DISPERSION/REPULSION */
 261
 262             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 263             vvdw6            = _mm256_mul_pd(c6_00,rinvsix);
 264             vvdw12           = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
 265             vvdw             = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
 266             fvdw             = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
 267
 268             /* Update potential sum for this i atom from the interaction with this j atom. */
 269             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 270
 271             fscal            = fvdw;
 272
 273             /* Calculate temporary vectorial force */
 274             tx               = _mm256_mul_pd(fscal,dx00);
 275             ty               = _mm256_mul_pd(fscal,dy00);
 276             tz               = _mm256_mul_pd(fscal,dz00);
 277
 278             /* Update vectorial force */
 279             fix0             = _mm256_add_pd(fix0,tx);
 280             fiy0             = _mm256_add_pd(fiy0,ty);
 281             fiz0             = _mm256_add_pd(fiz0,tz);
 282
 283             fjx0             = _mm256_add_pd(fjx0,tx);
 284             fjy0             = _mm256_add_pd(fjy0,ty);
 285             fjz0             = _mm256_add_pd(fjz0,tz);
 286
 287             /**************************
 288              * CALCULATE INTERACTIONS *
 289              **************************/
 290
 291             r10              = _mm256_mul_pd(rsq10,rinv10);
 292
 293             /* Compute parameters for interactions between i and j atoms */
 294             qq10             = _mm256_mul_pd(iq1,jq0);
 295
 296             /* Calculate table index by multiplying r with table scale and truncate to integer */
 297             rt               = _mm256_mul_pd(r10,vftabscale);
 298             vfitab           = _mm256_cvttpd_epi32(rt);
 299             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 300             vfitab           = _mm_slli_epi32(vfitab,2);
 301
 302             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 303             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 304             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 305             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 306             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 307             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 308             Heps             = _mm256_mul_pd(vfeps,H);
 309             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 310             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 311             velec            = _mm256_mul_pd(qq10,VV);
 312             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 313             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq10,FF),_mm256_mul_pd(vftabscale,rinv10)));
 314
 315             /* Update potential sum for this i atom from the interaction with this j atom. */
 316             velecsum         = _mm256_add_pd(velecsum,velec);
 317
 318             fscal            = felec;
 319
 320             /* Calculate temporary vectorial force */
 321             tx               = _mm256_mul_pd(fscal,dx10);
 322             ty               = _mm256_mul_pd(fscal,dy10);
 323             tz               = _mm256_mul_pd(fscal,dz10);
 324
 325             /* Update vectorial force */
 326             fix1             = _mm256_add_pd(fix1,tx);
 327             fiy1             = _mm256_add_pd(fiy1,ty);
 328             fiz1             = _mm256_add_pd(fiz1,tz);
 329
 330             fjx0             = _mm256_add_pd(fjx0,tx);
 331             fjy0             = _mm256_add_pd(fjy0,ty);
 332             fjz0             = _mm256_add_pd(fjz0,tz);
 333
 334             /**************************
 335              * CALCULATE INTERACTIONS *
 336              **************************/
 337
 338             r20              = _mm256_mul_pd(rsq20,rinv20);
 339
 340             /* Compute parameters for interactions between i and j atoms */
 341             qq20             = _mm256_mul_pd(iq2,jq0);
 342
 343             /* Calculate table index by multiplying r with table scale and truncate to integer */
 344             rt               = _mm256_mul_pd(r20,vftabscale);
 345             vfitab           = _mm256_cvttpd_epi32(rt);
 346             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 347             vfitab           = _mm_slli_epi32(vfitab,2);
 348
 349             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 350             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 351             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 352             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 353             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 354             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 355             Heps             = _mm256_mul_pd(vfeps,H);
 356             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 357             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 358             velec            = _mm256_mul_pd(qq20,VV);
 359             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 360             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq20,FF),_mm256_mul_pd(vftabscale,rinv20)));
 361
 362             /* Update potential sum for this i atom from the interaction with this j atom. */
 363             velecsum         = _mm256_add_pd(velecsum,velec);
 364
 365             fscal            = felec;
 366
 367             /* Calculate temporary vectorial force */
 368             tx               = _mm256_mul_pd(fscal,dx20);
 369             ty               = _mm256_mul_pd(fscal,dy20);
 370             tz               = _mm256_mul_pd(fscal,dz20);
 371
 372             /* Update vectorial force */
 373             fix2             = _mm256_add_pd(fix2,tx);
 374             fiy2             = _mm256_add_pd(fiy2,ty);
 375             fiz2             = _mm256_add_pd(fiz2,tz);
 376
 377             fjx0             = _mm256_add_pd(fjx0,tx);
 378             fjy0             = _mm256_add_pd(fjy0,ty);
 379             fjz0             = _mm256_add_pd(fjz0,tz);
 380
 381             /**************************
 382              * CALCULATE INTERACTIONS *
 383              **************************/
 384
 385             r30              = _mm256_mul_pd(rsq30,rinv30);
 386
 387             /* Compute parameters for interactions between i and j atoms */
 388             qq30             = _mm256_mul_pd(iq3,jq0);
 389
 390             /* Calculate table index by multiplying r with table scale and truncate to integer */
 391             rt               = _mm256_mul_pd(r30,vftabscale);
 392             vfitab           = _mm256_cvttpd_epi32(rt);
 393             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 394             vfitab           = _mm_slli_epi32(vfitab,2);
 395
 396             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 397             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 398             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 399             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 400             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 401             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 402             Heps             = _mm256_mul_pd(vfeps,H);
 403             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 404             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 405             velec            = _mm256_mul_pd(qq30,VV);
 406             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 407             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq30,FF),_mm256_mul_pd(vftabscale,rinv30)));
 408
 409             /* Update potential sum for this i atom from the interaction with this j atom. */
 410             velecsum         = _mm256_add_pd(velecsum,velec);
 411
 412             fscal            = felec;
 413
 414             /* Calculate temporary vectorial force */
 415             tx               = _mm256_mul_pd(fscal,dx30);
 416             ty               = _mm256_mul_pd(fscal,dy30);
 417             tz               = _mm256_mul_pd(fscal,dz30);
 418
 419             /* Update vectorial force */
 420             fix3             = _mm256_add_pd(fix3,tx);
 421             fiy3             = _mm256_add_pd(fiy3,ty);
 422             fiz3             = _mm256_add_pd(fiz3,tz);
 423
 424             fjx0             = _mm256_add_pd(fjx0,tx);
 425             fjy0             = _mm256_add_pd(fjy0,ty);
 426             fjz0             = _mm256_add_pd(fjz0,tz);
 427
 428             fjptrA             = f+j_coord_offsetA;
 429             fjptrB             = f+j_coord_offsetB;
 430             fjptrC             = f+j_coord_offsetC;
 431             fjptrD             = f+j_coord_offsetD;
 432
 433             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 434
 435             /* Inner loop uses 164 flops */
 436         }
 437
 438         if(jidx<j_index_end)
 439         {
 440
 441             /* Get j neighbor index, and coordinate index */
 442             jnrlistA         = jjnr[jidx];
 443             jnrlistB         = jjnr[jidx+1];
 444             jnrlistC         = jjnr[jidx+2];
 445             jnrlistD         = jjnr[jidx+3];
 446             /* Sign of each element will be negative for non-real atoms.
 447              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 448              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 449              */
 450             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 451
 452             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 453             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 454             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 455
 456             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 457             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 458             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 459             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 460             j_coord_offsetA  = DIM*jnrA;
 461             j_coord_offsetB  = DIM*jnrB;
 462             j_coord_offsetC  = DIM*jnrC;
 463             j_coord_offsetD  = DIM*jnrD;
 464
 465             /* load j atom coordinates */
 466             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 467                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 468                                                  &jx0,&jy0,&jz0);
 469
 470             /* Calculate displacement vector */
 471             dx00             = _mm256_sub_pd(ix0,jx0);
 472             dy00             = _mm256_sub_pd(iy0,jy0);
 473             dz00             = _mm256_sub_pd(iz0,jz0);
 474             dx10             = _mm256_sub_pd(ix1,jx0);
 475             dy10             = _mm256_sub_pd(iy1,jy0);
 476             dz10             = _mm256_sub_pd(iz1,jz0);
 477             dx20             = _mm256_sub_pd(ix2,jx0);
 478             dy20             = _mm256_sub_pd(iy2,jy0);
 479             dz20             = _mm256_sub_pd(iz2,jz0);
 480             dx30             = _mm256_sub_pd(ix3,jx0);
 481             dy30             = _mm256_sub_pd(iy3,jy0);
 482             dz30             = _mm256_sub_pd(iz3,jz0);
 483
 484             /* Calculate squared distance and things based on it */
 485             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 486             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 487             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 488             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 489
 490             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 491             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 492             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 493
 494             rinvsq00         = gmx_mm256_inv_pd(rsq00);
 495
 496             /* Load parameters for j particles */
 497             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 498                                                                  charge+jnrC+0,charge+jnrD+0);
 499             vdwjidx0A        = 2*vdwtype[jnrA+0];
 500             vdwjidx0B        = 2*vdwtype[jnrB+0];
 501             vdwjidx0C        = 2*vdwtype[jnrC+0];
 502             vdwjidx0D        = 2*vdwtype[jnrD+0];
 503
 504             fjx0             = _mm256_setzero_pd();
 505             fjy0             = _mm256_setzero_pd();
 506             fjz0             = _mm256_setzero_pd();
 507
 508             /**************************
 509              * CALCULATE INTERACTIONS *
 510              **************************/
 511
 512             /* Compute parameters for interactions between i and j atoms */
 513             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 514                                             vdwioffsetptr0+vdwjidx0B,
 515                                             vdwioffsetptr0+vdwjidx0C,
 516                                             vdwioffsetptr0+vdwjidx0D,
 517                                             &c6_00,&c12_00);
 518
 519             /* LENNARD-JONES DISPERSION/REPULSION */
 520
 521             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 522             vvdw6            = _mm256_mul_pd(c6_00,rinvsix);
 523             vvdw12           = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
 524             vvdw             = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
 525             fvdw             = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
 526
 527             /* Update potential sum for this i atom from the interaction with this j atom. */
 528             vvdw             = _mm256_andnot_pd(dummy_mask,vvdw);
 529             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 530
 531             fscal            = fvdw;
 532
 533             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 534
 535             /* Calculate temporary vectorial force */
 536             tx               = _mm256_mul_pd(fscal,dx00);
 537             ty               = _mm256_mul_pd(fscal,dy00);
 538             tz               = _mm256_mul_pd(fscal,dz00);
 539
 540             /* Update vectorial force */
 541             fix0             = _mm256_add_pd(fix0,tx);
 542             fiy0             = _mm256_add_pd(fiy0,ty);
 543             fiz0             = _mm256_add_pd(fiz0,tz);
 544
 545             fjx0             = _mm256_add_pd(fjx0,tx);
 546             fjy0             = _mm256_add_pd(fjy0,ty);
 547             fjz0             = _mm256_add_pd(fjz0,tz);
 548
 549             /**************************
 550              * CALCULATE INTERACTIONS *
 551              **************************/
 552
 553             r10              = _mm256_mul_pd(rsq10,rinv10);
 554             r10              = _mm256_andnot_pd(dummy_mask,r10);
 555
 556             /* Compute parameters for interactions between i and j atoms */
 557             qq10             = _mm256_mul_pd(iq1,jq0);
 558
 559             /* Calculate table index by multiplying r with table scale and truncate to integer */
 560             rt               = _mm256_mul_pd(r10,vftabscale);
 561             vfitab           = _mm256_cvttpd_epi32(rt);
 562             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 563             vfitab           = _mm_slli_epi32(vfitab,2);
 564
 565             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 566             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 567             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 568             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 569             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 570             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 571             Heps             = _mm256_mul_pd(vfeps,H);
 572             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 573             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 574             velec            = _mm256_mul_pd(qq10,VV);
 575             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 576             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq10,FF),_mm256_mul_pd(vftabscale,rinv10)));
 577
 578             /* Update potential sum for this i atom from the interaction with this j atom. */
 579             velec            = _mm256_andnot_pd(dummy_mask,velec);
 580             velecsum         = _mm256_add_pd(velecsum,velec);
 581
 582             fscal            = felec;
 583
 584             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 585
 586             /* Calculate temporary vectorial force */
 587             tx               = _mm256_mul_pd(fscal,dx10);
 588             ty               = _mm256_mul_pd(fscal,dy10);
 589             tz               = _mm256_mul_pd(fscal,dz10);
 590
 591             /* Update vectorial force */
 592             fix1             = _mm256_add_pd(fix1,tx);
 593             fiy1             = _mm256_add_pd(fiy1,ty);
 594             fiz1             = _mm256_add_pd(fiz1,tz);
 595
 596             fjx0             = _mm256_add_pd(fjx0,tx);
 597             fjy0             = _mm256_add_pd(fjy0,ty);
 598             fjz0             = _mm256_add_pd(fjz0,tz);
 599
 600             /**************************
 601              * CALCULATE INTERACTIONS *
 602              **************************/
 603
 604             r20              = _mm256_mul_pd(rsq20,rinv20);
 605             r20              = _mm256_andnot_pd(dummy_mask,r20);
 606
 607             /* Compute parameters for interactions between i and j atoms */
 608             qq20             = _mm256_mul_pd(iq2,jq0);
 609
 610             /* Calculate table index by multiplying r with table scale and truncate to integer */
 611             rt               = _mm256_mul_pd(r20,vftabscale);
 612             vfitab           = _mm256_cvttpd_epi32(rt);
 613             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 614             vfitab           = _mm_slli_epi32(vfitab,2);
 615
 616             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 617             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 618             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 619             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 620             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 621             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 622             Heps             = _mm256_mul_pd(vfeps,H);
 623             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 624             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 625             velec            = _mm256_mul_pd(qq20,VV);
 626             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 627             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq20,FF),_mm256_mul_pd(vftabscale,rinv20)));
 628
 629             /* Update potential sum for this i atom from the interaction with this j atom. */
 630             velec            = _mm256_andnot_pd(dummy_mask,velec);
 631             velecsum         = _mm256_add_pd(velecsum,velec);
 632
 633             fscal            = felec;
 634
 635             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 636
 637             /* Calculate temporary vectorial force */
 638             tx               = _mm256_mul_pd(fscal,dx20);
 639             ty               = _mm256_mul_pd(fscal,dy20);
 640             tz               = _mm256_mul_pd(fscal,dz20);
 641
 642             /* Update vectorial force */
 643             fix2             = _mm256_add_pd(fix2,tx);
 644             fiy2             = _mm256_add_pd(fiy2,ty);
 645             fiz2             = _mm256_add_pd(fiz2,tz);
 646
 647             fjx0             = _mm256_add_pd(fjx0,tx);
 648             fjy0             = _mm256_add_pd(fjy0,ty);
 649             fjz0             = _mm256_add_pd(fjz0,tz);
 650
 651             /**************************
 652              * CALCULATE INTERACTIONS *
 653              **************************/
 654
 655             r30              = _mm256_mul_pd(rsq30,rinv30);
 656             r30              = _mm256_andnot_pd(dummy_mask,r30);
 657
 658             /* Compute parameters for interactions between i and j atoms */
 659             qq30             = _mm256_mul_pd(iq3,jq0);
 660
 661             /* Calculate table index by multiplying r with table scale and truncate to integer */
 662             rt               = _mm256_mul_pd(r30,vftabscale);
 663             vfitab           = _mm256_cvttpd_epi32(rt);
 664             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 665             vfitab           = _mm_slli_epi32(vfitab,2);
 666
 667             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 668             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 669             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 670             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 671             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 672             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 673             Heps             = _mm256_mul_pd(vfeps,H);
 674             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 675             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 676             velec            = _mm256_mul_pd(qq30,VV);
 677             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 678             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq30,FF),_mm256_mul_pd(vftabscale,rinv30)));
 679
 680             /* Update potential sum for this i atom from the interaction with this j atom. */
 681             velec            = _mm256_andnot_pd(dummy_mask,velec);
 682             velecsum         = _mm256_add_pd(velecsum,velec);
 683
 684             fscal            = felec;
 685
 686             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 687
 688             /* Calculate temporary vectorial force */
 689             tx               = _mm256_mul_pd(fscal,dx30);
 690             ty               = _mm256_mul_pd(fscal,dy30);
 691             tz               = _mm256_mul_pd(fscal,dz30);
 692
 693             /* Update vectorial force */
 694             fix3             = _mm256_add_pd(fix3,tx);
 695             fiy3             = _mm256_add_pd(fiy3,ty);
 696             fiz3             = _mm256_add_pd(fiz3,tz);
 697
 698             fjx0             = _mm256_add_pd(fjx0,tx);
 699             fjy0             = _mm256_add_pd(fjy0,ty);
 700             fjz0             = _mm256_add_pd(fjz0,tz);
 701
 702             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 703             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 704             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 705             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 706
 707             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
 708
 709             /* Inner loop uses 167 flops */
 710         }
 711
 712         /* End of innermost loop */
 713
 714         gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 715                                                  f+i_coord_offset,fshift+i_shift_offset);
 716
 717         ggid                        = gid[iidx];
 718         /* Update potential energies */
 719         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 720         gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
 721
 722         /* Increment number of inner iterations */
 723         inneriter                  += j_index_end - j_index_start;
 724
 725         /* Outer loop uses 26 flops */
 726     }
 727
 728     /* Increment number of outer iterations */
 729     outeriter        += nri;
 730
 731     /* Update outer/inner flops */
 732
 733     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*167);
 734 }
 735 /*
 736  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwLJ_GeomW4P1_F_avx_256_double
 737  * Electrostatics interaction: CubicSplineTable
 738  * VdW interaction:            LennardJones
 739  * Geometry:                   Water4-Particle
 740  * Calculate force/pot:        Force
 741  */
 742 void
 743 nb_kernel_ElecCSTab_VdwLJ_GeomW4P1_F_avx_256_double
 744                     (t_nblist                    * gmx_restrict       nlist,
 745                      rvec                        * gmx_restrict          xx,
 746                      rvec                        * gmx_restrict          ff,
 747                      t_forcerec                  * gmx_restrict          fr,
 748                      t_mdatoms                   * gmx_restrict     mdatoms,
 749                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 750                      t_nrnb                      * gmx_restrict        nrnb)
 751 {
 752     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 753      * just 0 for non-waters.
 754      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
 755      * jnr indices corresponding to data put in the four positions in the SIMD register.
 756      */
 757     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 758     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 759     int              jnrA,jnrB,jnrC,jnrD;
 760     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 761     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 762     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 763     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 764     real             rcutoff_scalar;
 765     real             *shiftvec,*fshift,*x,*f;
 766     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
 767     real             scratch[4*DIM];
 768     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 769     real *           vdwioffsetptr0;
 770     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 771     real *           vdwioffsetptr1;
 772     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 773     real *           vdwioffsetptr2;
 774     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 775     real *           vdwioffsetptr3;
 776     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 777     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
 778     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 779     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 780     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 781     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 782     __m256d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 783     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 784     real             *charge;
 785     int              nvdwtype;
 786     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 787     int              *vdwtype;
 788     real             *vdwparam;
 789     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
 790     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
 791     __m128i          vfitab;
 792     __m128i          ifour       = _mm_set1_epi32(4);
 793     __m256d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 794     real             *vftab;
 795     __m256d          dummy_mask,cutoff_mask;
 796     __m128           tmpmask0,tmpmask1;
 797     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 798     __m256d          one     = _mm256_set1_pd(1.0);
 799     __m256d          two     = _mm256_set1_pd(2.0);
 800     x                = xx[0];
 801     f                = ff[0];
 802
 803     nri              = nlist->nri;
 804     iinr             = nlist->iinr;
 805     jindex           = nlist->jindex;
 806     jjnr             = nlist->jjnr;
 807     shiftidx         = nlist->shift;
 808     gid              = nlist->gid;
 809     shiftvec         = fr->shift_vec[0];
 810     fshift           = fr->fshift[0];
 811     facel            = _mm256_set1_pd(fr->epsfac);
 812     charge           = mdatoms->chargeA;
 813     nvdwtype         = fr->ntype;
 814     vdwparam         = fr->nbfp;
 815     vdwtype          = mdatoms->typeA;
 816
 817     vftab            = kernel_data->table_elec->data;
 818     vftabscale       = _mm256_set1_pd(kernel_data->table_elec->scale);
 819
 820     /* Setup water-specific parameters */
 821     inr              = nlist->iinr[0];
 822     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 823     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 824     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 825     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 826
 827     /* Avoid stupid compiler warnings */
 828     jnrA = jnrB = jnrC = jnrD = 0;
 829     j_coord_offsetA = 0;
 830     j_coord_offsetB = 0;
 831     j_coord_offsetC = 0;
 832     j_coord_offsetD = 0;
 833
 834     outeriter        = 0;
 835     inneriter        = 0;
 836
 837     for(iidx=0;iidx<4*DIM;iidx++)
 838     {
 839         scratch[iidx] = 0.0;
 840     }
 841
 842     /* Start outer loop over neighborlists */
 843     for(iidx=0; iidx<nri; iidx++)
 844     {
 845         /* Load shift vector for this list */
 846         i_shift_offset   = DIM*shiftidx[iidx];
 847
 848         /* Load limits for loop over neighbors */
 849         j_index_start    = jindex[iidx];
 850         j_index_end      = jindex[iidx+1];
 851
 852         /* Get outer coordinate index */
 853         inr              = iinr[iidx];
 854         i_coord_offset   = DIM*inr;
 855
 856         /* Load i particle coords and add shift vector */
 857         gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 858                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 859
 860         fix0             = _mm256_setzero_pd();
 861         fiy0             = _mm256_setzero_pd();
 862         fiz0             = _mm256_setzero_pd();
 863         fix1             = _mm256_setzero_pd();
 864         fiy1             = _mm256_setzero_pd();
 865         fiz1             = _mm256_setzero_pd();
 866         fix2             = _mm256_setzero_pd();
 867         fiy2             = _mm256_setzero_pd();
 868         fiz2             = _mm256_setzero_pd();
 869         fix3             = _mm256_setzero_pd();
 870         fiy3             = _mm256_setzero_pd();
 871         fiz3             = _mm256_setzero_pd();
 872
 873         /* Start inner kernel loop */
 874         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 875         {
 876
 877             /* Get j neighbor index, and coordinate index */
 878             jnrA             = jjnr[jidx];
 879             jnrB             = jjnr[jidx+1];
 880             jnrC             = jjnr[jidx+2];
 881             jnrD             = jjnr[jidx+3];
 882             j_coord_offsetA  = DIM*jnrA;
 883             j_coord_offsetB  = DIM*jnrB;
 884             j_coord_offsetC  = DIM*jnrC;
 885             j_coord_offsetD  = DIM*jnrD;
 886
 887             /* load j atom coordinates */
 888             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 889                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 890                                                  &jx0,&jy0,&jz0);
 891
 892             /* Calculate displacement vector */
 893             dx00             = _mm256_sub_pd(ix0,jx0);
 894             dy00             = _mm256_sub_pd(iy0,jy0);
 895             dz00             = _mm256_sub_pd(iz0,jz0);
 896             dx10             = _mm256_sub_pd(ix1,jx0);
 897             dy10             = _mm256_sub_pd(iy1,jy0);
 898             dz10             = _mm256_sub_pd(iz1,jz0);
 899             dx20             = _mm256_sub_pd(ix2,jx0);
 900             dy20             = _mm256_sub_pd(iy2,jy0);
 901             dz20             = _mm256_sub_pd(iz2,jz0);
 902             dx30             = _mm256_sub_pd(ix3,jx0);
 903             dy30             = _mm256_sub_pd(iy3,jy0);
 904             dz30             = _mm256_sub_pd(iz3,jz0);
 905
 906             /* Calculate squared distance and things based on it */
 907             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 908             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 909             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 910             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
 911
 912             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 913             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 914             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
 915
 916             rinvsq00         = gmx_mm256_inv_pd(rsq00);
 917
 918             /* Load parameters for j particles */
 919             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
 920                                                                  charge+jnrC+0,charge+jnrD+0);
 921             vdwjidx0A        = 2*vdwtype[jnrA+0];
 922             vdwjidx0B        = 2*vdwtype[jnrB+0];
 923             vdwjidx0C        = 2*vdwtype[jnrC+0];
 924             vdwjidx0D        = 2*vdwtype[jnrD+0];
 925
 926             fjx0             = _mm256_setzero_pd();
 927             fjy0             = _mm256_setzero_pd();
 928             fjz0             = _mm256_setzero_pd();
 929
 930             /**************************
 931              * CALCULATE INTERACTIONS *
 932              **************************/
 933
 934             /* Compute parameters for interactions between i and j atoms */
 935             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
 936                                             vdwioffsetptr0+vdwjidx0B,
 937                                             vdwioffsetptr0+vdwjidx0C,
 938                                             vdwioffsetptr0+vdwjidx0D,
 939                                             &c6_00,&c12_00);
 940
 941             /* LENNARD-JONES DISPERSION/REPULSION */
 942
 943             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 944             fvdw             = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
 945
 946             fscal            = fvdw;
 947
 948             /* Calculate temporary vectorial force */
 949             tx               = _mm256_mul_pd(fscal,dx00);
 950             ty               = _mm256_mul_pd(fscal,dy00);
 951             tz               = _mm256_mul_pd(fscal,dz00);
 952
 953             /* Update vectorial force */
 954             fix0             = _mm256_add_pd(fix0,tx);
 955             fiy0             = _mm256_add_pd(fiy0,ty);
 956             fiz0             = _mm256_add_pd(fiz0,tz);
 957
 958             fjx0             = _mm256_add_pd(fjx0,tx);
 959             fjy0             = _mm256_add_pd(fjy0,ty);
 960             fjz0             = _mm256_add_pd(fjz0,tz);
 961
 962             /**************************
 963              * CALCULATE INTERACTIONS *
 964              **************************/
 965
 966             r10              = _mm256_mul_pd(rsq10,rinv10);
 967
 968             /* Compute parameters for interactions between i and j atoms */
 969             qq10             = _mm256_mul_pd(iq1,jq0);
 970
 971             /* Calculate table index by multiplying r with table scale and truncate to integer */
 972             rt               = _mm256_mul_pd(r10,vftabscale);
 973             vfitab           = _mm256_cvttpd_epi32(rt);
 974             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 975             vfitab           = _mm_slli_epi32(vfitab,2);
 976
 977             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 978             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 979             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 980             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 981             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 982             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 983             Heps             = _mm256_mul_pd(vfeps,H);
 984             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 985             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 986             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq10,FF),_mm256_mul_pd(vftabscale,rinv10)));
 987
 988             fscal            = felec;
 989
 990             /* Calculate temporary vectorial force */
 991             tx               = _mm256_mul_pd(fscal,dx10);
 992             ty               = _mm256_mul_pd(fscal,dy10);
 993             tz               = _mm256_mul_pd(fscal,dz10);
 994
 995             /* Update vectorial force */
 996             fix1             = _mm256_add_pd(fix1,tx);
 997             fiy1             = _mm256_add_pd(fiy1,ty);
 998             fiz1             = _mm256_add_pd(fiz1,tz);
 999
1000             fjx0             = _mm256_add_pd(fjx0,tx);
1001             fjy0             = _mm256_add_pd(fjy0,ty);
1002             fjz0             = _mm256_add_pd(fjz0,tz);
1003
1004             /**************************
1005              * CALCULATE INTERACTIONS *
1006              **************************/
1007
1008             r20              = _mm256_mul_pd(rsq20,rinv20);
1009
1010             /* Compute parameters for interactions between i and j atoms */
1011             qq20             = _mm256_mul_pd(iq2,jq0);
1012
1013             /* Calculate table index by multiplying r with table scale and truncate to integer */
1014             rt               = _mm256_mul_pd(r20,vftabscale);
1015             vfitab           = _mm256_cvttpd_epi32(rt);
1016             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1017             vfitab           = _mm_slli_epi32(vfitab,2);
1018
1019             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1020             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1021             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1022             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1023             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1024             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1025             Heps             = _mm256_mul_pd(vfeps,H);
1026             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1027             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1028             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq20,FF),_mm256_mul_pd(vftabscale,rinv20)));
1029
1030             fscal            = felec;
1031
1032             /* Calculate temporary vectorial force */
1033             tx               = _mm256_mul_pd(fscal,dx20);
1034             ty               = _mm256_mul_pd(fscal,dy20);
1035             tz               = _mm256_mul_pd(fscal,dz20);
1036
1037             /* Update vectorial force */
1038             fix2             = _mm256_add_pd(fix2,tx);
1039             fiy2             = _mm256_add_pd(fiy2,ty);
1040             fiz2             = _mm256_add_pd(fiz2,tz);
1041
1042             fjx0             = _mm256_add_pd(fjx0,tx);
1043             fjy0             = _mm256_add_pd(fjy0,ty);
1044             fjz0             = _mm256_add_pd(fjz0,tz);
1045
1046             /**************************
1047              * CALCULATE INTERACTIONS *
1048              **************************/
1049
1050             r30              = _mm256_mul_pd(rsq30,rinv30);
1051
1052             /* Compute parameters for interactions between i and j atoms */
1053             qq30             = _mm256_mul_pd(iq3,jq0);
1054
1055             /* Calculate table index by multiplying r with table scale and truncate to integer */
1056             rt               = _mm256_mul_pd(r30,vftabscale);
1057             vfitab           = _mm256_cvttpd_epi32(rt);
1058             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1059             vfitab           = _mm_slli_epi32(vfitab,2);
1060
1061             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1062             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1063             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1064             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1065             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1066             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1067             Heps             = _mm256_mul_pd(vfeps,H);
1068             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1069             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1070             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq30,FF),_mm256_mul_pd(vftabscale,rinv30)));
1071
1072             fscal            = felec;
1073
1074             /* Calculate temporary vectorial force */
1075             tx               = _mm256_mul_pd(fscal,dx30);
1076             ty               = _mm256_mul_pd(fscal,dy30);
1077             tz               = _mm256_mul_pd(fscal,dz30);
1078
1079             /* Update vectorial force */
1080             fix3             = _mm256_add_pd(fix3,tx);
1081             fiy3             = _mm256_add_pd(fiy3,ty);
1082             fiz3             = _mm256_add_pd(fiz3,tz);
1083
1084             fjx0             = _mm256_add_pd(fjx0,tx);
1085             fjy0             = _mm256_add_pd(fjy0,ty);
1086             fjz0             = _mm256_add_pd(fjz0,tz);
1087
1088             fjptrA             = f+j_coord_offsetA;
1089             fjptrB             = f+j_coord_offsetB;
1090             fjptrC             = f+j_coord_offsetC;
1091             fjptrD             = f+j_coord_offsetD;
1092
1093             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1094
1095             /* Inner loop uses 147 flops */
1096         }
1097
1098         if(jidx<j_index_end)
1099         {
1100
1101             /* Get j neighbor index, and coordinate index */
1102             jnrlistA         = jjnr[jidx];
1103             jnrlistB         = jjnr[jidx+1];
1104             jnrlistC         = jjnr[jidx+2];
1105             jnrlistD         = jjnr[jidx+3];
1106             /* Sign of each element will be negative for non-real atoms.
1107              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1108              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1109              */
1110             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1111
1112             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1113             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1114             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1115
1116             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1117             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1118             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1119             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1120             j_coord_offsetA  = DIM*jnrA;
1121             j_coord_offsetB  = DIM*jnrB;
1122             j_coord_offsetC  = DIM*jnrC;
1123             j_coord_offsetD  = DIM*jnrD;
1124
1125             /* load j atom coordinates */
1126             gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1127                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1128                                                  &jx0,&jy0,&jz0);
1129
1130             /* Calculate displacement vector */
1131             dx00             = _mm256_sub_pd(ix0,jx0);
1132             dy00             = _mm256_sub_pd(iy0,jy0);
1133             dz00             = _mm256_sub_pd(iz0,jz0);
1134             dx10             = _mm256_sub_pd(ix1,jx0);
1135             dy10             = _mm256_sub_pd(iy1,jy0);
1136             dz10             = _mm256_sub_pd(iz1,jz0);
1137             dx20             = _mm256_sub_pd(ix2,jx0);
1138             dy20             = _mm256_sub_pd(iy2,jy0);
1139             dz20             = _mm256_sub_pd(iz2,jz0);
1140             dx30             = _mm256_sub_pd(ix3,jx0);
1141             dy30             = _mm256_sub_pd(iy3,jy0);
1142             dz30             = _mm256_sub_pd(iz3,jz0);
1143
1144             /* Calculate squared distance and things based on it */
1145             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1146             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1147             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1148             rsq30            = gmx_mm256_calc_rsq_pd(dx30,dy30,dz30);
1149
1150             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
1151             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
1152             rinv30           = gmx_mm256_invsqrt_pd(rsq30);
1153
1154             rinvsq00         = gmx_mm256_inv_pd(rsq00);
1155
1156             /* Load parameters for j particles */
1157             jq0              = gmx_mm256_load_4real_swizzle_pd(charge+jnrA+0,charge+jnrB+0,
1158                                                                  charge+jnrC+0,charge+jnrD+0);
1159             vdwjidx0A        = 2*vdwtype[jnrA+0];
1160             vdwjidx0B        = 2*vdwtype[jnrB+0];
1161             vdwjidx0C        = 2*vdwtype[jnrC+0];
1162             vdwjidx0D        = 2*vdwtype[jnrD+0];
1163
1164             fjx0             = _mm256_setzero_pd();
1165             fjy0             = _mm256_setzero_pd();
1166             fjz0             = _mm256_setzero_pd();
1167
1168             /**************************
1169              * CALCULATE INTERACTIONS *
1170              **************************/
1171
1172             /* Compute parameters for interactions between i and j atoms */
1173             gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
1174                                             vdwioffsetptr0+vdwjidx0B,
1175                                             vdwioffsetptr0+vdwjidx0C,
1176                                             vdwioffsetptr0+vdwjidx0D,
1177                                             &c6_00,&c12_00);
1178
1179             /* LENNARD-JONES DISPERSION/REPULSION */
1180
1181             rinvsix          = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1182             fvdw             = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1183
1184             fscal            = fvdw;
1185
1186             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1187
1188             /* Calculate temporary vectorial force */
1189             tx               = _mm256_mul_pd(fscal,dx00);
1190             ty               = _mm256_mul_pd(fscal,dy00);
1191             tz               = _mm256_mul_pd(fscal,dz00);
1192
1193             /* Update vectorial force */
1194             fix0             = _mm256_add_pd(fix0,tx);
1195             fiy0             = _mm256_add_pd(fiy0,ty);
1196             fiz0             = _mm256_add_pd(fiz0,tz);
1197
1198             fjx0             = _mm256_add_pd(fjx0,tx);
1199             fjy0             = _mm256_add_pd(fjy0,ty);
1200             fjz0             = _mm256_add_pd(fjz0,tz);
1201
1202             /**************************
1203              * CALCULATE INTERACTIONS *
1204              **************************/
1205
1206             r10              = _mm256_mul_pd(rsq10,rinv10);
1207             r10              = _mm256_andnot_pd(dummy_mask,r10);
1208
1209             /* Compute parameters for interactions between i and j atoms */
1210             qq10             = _mm256_mul_pd(iq1,jq0);
1211
1212             /* Calculate table index by multiplying r with table scale and truncate to integer */
1213             rt               = _mm256_mul_pd(r10,vftabscale);
1214             vfitab           = _mm256_cvttpd_epi32(rt);
1215             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1216             vfitab           = _mm_slli_epi32(vfitab,2);
1217
1218             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1219             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1220             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1221             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1222             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1223             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1224             Heps             = _mm256_mul_pd(vfeps,H);
1225             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1226             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1227             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq10,FF),_mm256_mul_pd(vftabscale,rinv10)));
1228
1229             fscal            = felec;
1230
1231             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1232
1233             /* Calculate temporary vectorial force */
1234             tx               = _mm256_mul_pd(fscal,dx10);
1235             ty               = _mm256_mul_pd(fscal,dy10);
1236             tz               = _mm256_mul_pd(fscal,dz10);
1237
1238             /* Update vectorial force */
1239             fix1             = _mm256_add_pd(fix1,tx);
1240             fiy1             = _mm256_add_pd(fiy1,ty);
1241             fiz1             = _mm256_add_pd(fiz1,tz);
1242
1243             fjx0             = _mm256_add_pd(fjx0,tx);
1244             fjy0             = _mm256_add_pd(fjy0,ty);
1245             fjz0             = _mm256_add_pd(fjz0,tz);
1246
1247             /**************************
1248              * CALCULATE INTERACTIONS *
1249              **************************/
1250
1251             r20              = _mm256_mul_pd(rsq20,rinv20);
1252             r20              = _mm256_andnot_pd(dummy_mask,r20);
1253
1254             /* Compute parameters for interactions between i and j atoms */
1255             qq20             = _mm256_mul_pd(iq2,jq0);
1256
1257             /* Calculate table index by multiplying r with table scale and truncate to integer */
1258             rt               = _mm256_mul_pd(r20,vftabscale);
1259             vfitab           = _mm256_cvttpd_epi32(rt);
1260             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1261             vfitab           = _mm_slli_epi32(vfitab,2);
1262
1263             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1264             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1265             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1266             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1267             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1268             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1269             Heps             = _mm256_mul_pd(vfeps,H);
1270             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1271             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1272             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq20,FF),_mm256_mul_pd(vftabscale,rinv20)));
1273
1274             fscal            = felec;
1275
1276             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1277
1278             /* Calculate temporary vectorial force */
1279             tx               = _mm256_mul_pd(fscal,dx20);
1280             ty               = _mm256_mul_pd(fscal,dy20);
1281             tz               = _mm256_mul_pd(fscal,dz20);
1282
1283             /* Update vectorial force */
1284             fix2             = _mm256_add_pd(fix2,tx);
1285             fiy2             = _mm256_add_pd(fiy2,ty);
1286             fiz2             = _mm256_add_pd(fiz2,tz);
1287
1288             fjx0             = _mm256_add_pd(fjx0,tx);
1289             fjy0             = _mm256_add_pd(fjy0,ty);
1290             fjz0             = _mm256_add_pd(fjz0,tz);
1291
1292             /**************************
1293              * CALCULATE INTERACTIONS *
1294              **************************/
1295
1296             r30              = _mm256_mul_pd(rsq30,rinv30);
1297             r30              = _mm256_andnot_pd(dummy_mask,r30);
1298
1299             /* Compute parameters for interactions between i and j atoms */
1300             qq30             = _mm256_mul_pd(iq3,jq0);
1301
1302             /* Calculate table index by multiplying r with table scale and truncate to integer */
1303             rt               = _mm256_mul_pd(r30,vftabscale);
1304             vfitab           = _mm256_cvttpd_epi32(rt);
1305             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1306             vfitab           = _mm_slli_epi32(vfitab,2);
1307
1308             /* CUBIC SPLINE TABLE ELECTROSTATICS */
1309             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1310             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1311             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1312             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1313             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1314             Heps             = _mm256_mul_pd(vfeps,H);
1315             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1316             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1317             felec            = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_mul_pd(qq30,FF),_mm256_mul_pd(vftabscale,rinv30)));
1318
1319             fscal            = felec;
1320
1321             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1322
1323             /* Calculate temporary vectorial force */
1324             tx               = _mm256_mul_pd(fscal,dx30);
1325             ty               = _mm256_mul_pd(fscal,dy30);
1326             tz               = _mm256_mul_pd(fscal,dz30);
1327
1328             /* Update vectorial force */
1329             fix3             = _mm256_add_pd(fix3,tx);
1330             fiy3             = _mm256_add_pd(fiy3,ty);
1331             fiz3             = _mm256_add_pd(fiz3,tz);
1332
1333             fjx0             = _mm256_add_pd(fjx0,tx);
1334             fjy0             = _mm256_add_pd(fjy0,ty);
1335             fjz0             = _mm256_add_pd(fjz0,tz);
1336
1337             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1338             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1339             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1340             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1341
1342             gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,fjx0,fjy0,fjz0);
1343
1344             /* Inner loop uses 150 flops */
1345         }
1346
1347         /* End of innermost loop */
1348
1349         gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1350                                                  f+i_coord_offset,fshift+i_shift_offset);
1351
1352         /* Increment number of inner iterations */
1353         inneriter                  += j_index_end - j_index_start;
1354
1355         /* Outer loop uses 24 flops */
1356     }
1357
1358     /* Increment number of outer iterations */
1359     outeriter        += nri;
1360
1361     /* Update outer/inner flops */
1362
1363     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*150);
1364 }