src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecEw_VdwCSTab_GeomW4W4_avx_256_double.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013, 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 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "vec.h"
  47 #include "nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_256_double.h"
  50 #include "kernelutil_x86_avx_256_double.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW4W4_VF_avx_256_double
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            CubicSplineTable
  56  * Geometry:                   Water4-Water4
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEw_VdwCSTab_GeomW4W4_VF_avx_256_double
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB,jnrC,jnrD;
  77     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  78     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  79     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  80     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  81     real             rcutoff_scalar;
  82     real             *shiftvec,*fshift,*x,*f;
  83     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  84     real             scratch[4*DIM];
  85     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  86     real *           vdwioffsetptr0;
  87     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  88     real *           vdwioffsetptr1;
  89     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  90     real *           vdwioffsetptr2;
  91     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  92     real *           vdwioffsetptr3;
  93     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  94     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  95     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  96     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
  97     __m256d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
  98     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
  99     __m256d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
 100     int              vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
 101     __m256d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
 102     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 103     __m256d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
 104     __m256d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
 105     __m256d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
 106     __m256d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 107     __m256d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 108     __m256d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
 109     __m256d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
 110     __m256d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
 111     __m256d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
 112     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 113     real             *charge;
 114     int              nvdwtype;
 115     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 116     int              *vdwtype;
 117     real             *vdwparam;
 118     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
 119     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
 120     __m128i          vfitab;
 121     __m128i          ifour       = _mm_set1_epi32(4);
 122     __m256d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 123     real             *vftab;
 124     __m128i          ewitab;
 125     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 126     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 127     real             *ewtab;
 128     __m256d          dummy_mask,cutoff_mask;
 129     __m128           tmpmask0,tmpmask1;
 130     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 131     __m256d          one     = _mm256_set1_pd(1.0);
 132     __m256d          two     = _mm256_set1_pd(2.0);
 133     x                = xx[0];
 134     f                = ff[0];
 135
 136     nri              = nlist->nri;
 137     iinr             = nlist->iinr;
 138     jindex           = nlist->jindex;
 139     jjnr             = nlist->jjnr;
 140     shiftidx         = nlist->shift;
 141     gid              = nlist->gid;
 142     shiftvec         = fr->shift_vec[0];
 143     fshift           = fr->fshift[0];
 144     facel            = _mm256_set1_pd(fr->epsfac);
 145     charge           = mdatoms->chargeA;
 146     nvdwtype         = fr->ntype;
 147     vdwparam         = fr->nbfp;
 148     vdwtype          = mdatoms->typeA;
 149
 150     vftab            = kernel_data->table_vdw->data;
 151     vftabscale       = _mm256_set1_pd(kernel_data->table_vdw->scale);
 152
 153     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 154     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
 155     beta2            = _mm256_mul_pd(beta,beta);
 156     beta3            = _mm256_mul_pd(beta,beta2);
 157
 158     ewtab            = fr->ic->tabq_coul_FDV0;
 159     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 160     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 161
 162     /* Setup water-specific parameters */
 163     inr              = nlist->iinr[0];
 164     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 165     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 166     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
 167     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 168
 169     jq1              = _mm256_set1_pd(charge[inr+1]);
 170     jq2              = _mm256_set1_pd(charge[inr+2]);
 171     jq3              = _mm256_set1_pd(charge[inr+3]);
 172     vdwjidx0A        = 2*vdwtype[inr+0];
 173     c6_00            = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
 174     c12_00           = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
 175     qq11             = _mm256_mul_pd(iq1,jq1);
 176     qq12             = _mm256_mul_pd(iq1,jq2);
 177     qq13             = _mm256_mul_pd(iq1,jq3);
 178     qq21             = _mm256_mul_pd(iq2,jq1);
 179     qq22             = _mm256_mul_pd(iq2,jq2);
 180     qq23             = _mm256_mul_pd(iq2,jq3);
 181     qq31             = _mm256_mul_pd(iq3,jq1);
 182     qq32             = _mm256_mul_pd(iq3,jq2);
 183     qq33             = _mm256_mul_pd(iq3,jq3);
 184
 185     /* Avoid stupid compiler warnings */
 186     jnrA = jnrB = jnrC = jnrD = 0;
 187     j_coord_offsetA = 0;
 188     j_coord_offsetB = 0;
 189     j_coord_offsetC = 0;
 190     j_coord_offsetD = 0;
 191
 192     outeriter        = 0;
 193     inneriter        = 0;
 194
 195     for(iidx=0;iidx<4*DIM;iidx++)
 196     {
 197         scratch[iidx] = 0.0;
 198     }
 199
 200     /* Start outer loop over neighborlists */
 201     for(iidx=0; iidx<nri; iidx++)
 202     {
 203         /* Load shift vector for this list */
 204         i_shift_offset   = DIM*shiftidx[iidx];
 205
 206         /* Load limits for loop over neighbors */
 207         j_index_start    = jindex[iidx];
 208         j_index_end      = jindex[iidx+1];
 209
 210         /* Get outer coordinate index */
 211         inr              = iinr[iidx];
 212         i_coord_offset   = DIM*inr;
 213
 214         /* Load i particle coords and add shift vector */
 215         gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 216                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 217
 218         fix0             = _mm256_setzero_pd();
 219         fiy0             = _mm256_setzero_pd();
 220         fiz0             = _mm256_setzero_pd();
 221         fix1             = _mm256_setzero_pd();
 222         fiy1             = _mm256_setzero_pd();
 223         fiz1             = _mm256_setzero_pd();
 224         fix2             = _mm256_setzero_pd();
 225         fiy2             = _mm256_setzero_pd();
 226         fiz2             = _mm256_setzero_pd();
 227         fix3             = _mm256_setzero_pd();
 228         fiy3             = _mm256_setzero_pd();
 229         fiz3             = _mm256_setzero_pd();
 230
 231         /* Reset potential sums */
 232         velecsum         = _mm256_setzero_pd();
 233         vvdwsum          = _mm256_setzero_pd();
 234
 235         /* Start inner kernel loop */
 236         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 237         {
 238
 239             /* Get j neighbor index, and coordinate index */
 240             jnrA             = jjnr[jidx];
 241             jnrB             = jjnr[jidx+1];
 242             jnrC             = jjnr[jidx+2];
 243             jnrD             = jjnr[jidx+3];
 244             j_coord_offsetA  = DIM*jnrA;
 245             j_coord_offsetB  = DIM*jnrB;
 246             j_coord_offsetC  = DIM*jnrC;
 247             j_coord_offsetD  = DIM*jnrD;
 248
 249             /* load j atom coordinates */
 250             gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 251                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 252                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 253                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
 254
 255             /* Calculate displacement vector */
 256             dx00             = _mm256_sub_pd(ix0,jx0);
 257             dy00             = _mm256_sub_pd(iy0,jy0);
 258             dz00             = _mm256_sub_pd(iz0,jz0);
 259             dx11             = _mm256_sub_pd(ix1,jx1);
 260             dy11             = _mm256_sub_pd(iy1,jy1);
 261             dz11             = _mm256_sub_pd(iz1,jz1);
 262             dx12             = _mm256_sub_pd(ix1,jx2);
 263             dy12             = _mm256_sub_pd(iy1,jy2);
 264             dz12             = _mm256_sub_pd(iz1,jz2);
 265             dx13             = _mm256_sub_pd(ix1,jx3);
 266             dy13             = _mm256_sub_pd(iy1,jy3);
 267             dz13             = _mm256_sub_pd(iz1,jz3);
 268             dx21             = _mm256_sub_pd(ix2,jx1);
 269             dy21             = _mm256_sub_pd(iy2,jy1);
 270             dz21             = _mm256_sub_pd(iz2,jz1);
 271             dx22             = _mm256_sub_pd(ix2,jx2);
 272             dy22             = _mm256_sub_pd(iy2,jy2);
 273             dz22             = _mm256_sub_pd(iz2,jz2);
 274             dx23             = _mm256_sub_pd(ix2,jx3);
 275             dy23             = _mm256_sub_pd(iy2,jy3);
 276             dz23             = _mm256_sub_pd(iz2,jz3);
 277             dx31             = _mm256_sub_pd(ix3,jx1);
 278             dy31             = _mm256_sub_pd(iy3,jy1);
 279             dz31             = _mm256_sub_pd(iz3,jz1);
 280             dx32             = _mm256_sub_pd(ix3,jx2);
 281             dy32             = _mm256_sub_pd(iy3,jy2);
 282             dz32             = _mm256_sub_pd(iz3,jz2);
 283             dx33             = _mm256_sub_pd(ix3,jx3);
 284             dy33             = _mm256_sub_pd(iy3,jy3);
 285             dz33             = _mm256_sub_pd(iz3,jz3);
 286
 287             /* Calculate squared distance and things based on it */
 288             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 289             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
 290             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
 291             rsq13            = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
 292             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
 293             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
 294             rsq23            = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
 295             rsq31            = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
 296             rsq32            = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
 297             rsq33            = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
 298
 299             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 300             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
 301             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
 302             rinv13           = gmx_mm256_invsqrt_pd(rsq13);
 303             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
 304             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
 305             rinv23           = gmx_mm256_invsqrt_pd(rsq23);
 306             rinv31           = gmx_mm256_invsqrt_pd(rsq31);
 307             rinv32           = gmx_mm256_invsqrt_pd(rsq32);
 308             rinv33           = gmx_mm256_invsqrt_pd(rsq33);
 309
 310             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
 311             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
 312             rinvsq13         = _mm256_mul_pd(rinv13,rinv13);
 313             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
 314             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
 315             rinvsq23         = _mm256_mul_pd(rinv23,rinv23);
 316             rinvsq31         = _mm256_mul_pd(rinv31,rinv31);
 317             rinvsq32         = _mm256_mul_pd(rinv32,rinv32);
 318             rinvsq33         = _mm256_mul_pd(rinv33,rinv33);
 319
 320             fjx0             = _mm256_setzero_pd();
 321             fjy0             = _mm256_setzero_pd();
 322             fjz0             = _mm256_setzero_pd();
 323             fjx1             = _mm256_setzero_pd();
 324             fjy1             = _mm256_setzero_pd();
 325             fjz1             = _mm256_setzero_pd();
 326             fjx2             = _mm256_setzero_pd();
 327             fjy2             = _mm256_setzero_pd();
 328             fjz2             = _mm256_setzero_pd();
 329             fjx3             = _mm256_setzero_pd();
 330             fjy3             = _mm256_setzero_pd();
 331             fjz3             = _mm256_setzero_pd();
 332
 333             /**************************
 334              * CALCULATE INTERACTIONS *
 335              **************************/
 336
 337             r00              = _mm256_mul_pd(rsq00,rinv00);
 338
 339             /* Calculate table index by multiplying r with table scale and truncate to integer */
 340             rt               = _mm256_mul_pd(r00,vftabscale);
 341             vfitab           = _mm256_cvttpd_epi32(rt);
 342             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 343             vfitab           = _mm_slli_epi32(vfitab,3);
 344
 345             /* CUBIC SPLINE TABLE DISPERSION */
 346             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 347             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 348             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 349             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 350             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 351             Heps             = _mm256_mul_pd(vfeps,H);
 352             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 353             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 354             vvdw6            = _mm256_mul_pd(c6_00,VV);
 355             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 356             fvdw6            = _mm256_mul_pd(c6_00,FF);
 357
 358             /* CUBIC SPLINE TABLE REPULSION */
 359             vfitab           = _mm_add_epi32(vfitab,ifour);
 360             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 361             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 362             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 363             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 364             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 365             Heps             = _mm256_mul_pd(vfeps,H);
 366             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 367             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 368             vvdw12           = _mm256_mul_pd(c12_00,VV);
 369             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 370             fvdw12           = _mm256_mul_pd(c12_00,FF);
 371             vvdw             = _mm256_add_pd(vvdw12,vvdw6);
 372             fvdw             = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_add_pd(fvdw6,fvdw12),_mm256_mul_pd(vftabscale,rinv00)));
 373
 374             /* Update potential sum for this i atom from the interaction with this j atom. */
 375             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 376
 377             fscal            = fvdw;
 378
 379             /* Calculate temporary vectorial force */
 380             tx               = _mm256_mul_pd(fscal,dx00);
 381             ty               = _mm256_mul_pd(fscal,dy00);
 382             tz               = _mm256_mul_pd(fscal,dz00);
 383
 384             /* Update vectorial force */
 385             fix0             = _mm256_add_pd(fix0,tx);
 386             fiy0             = _mm256_add_pd(fiy0,ty);
 387             fiz0             = _mm256_add_pd(fiz0,tz);
 388
 389             fjx0             = _mm256_add_pd(fjx0,tx);
 390             fjy0             = _mm256_add_pd(fjy0,ty);
 391             fjz0             = _mm256_add_pd(fjz0,tz);
 392
 393             /**************************
 394              * CALCULATE INTERACTIONS *
 395              **************************/
 396
 397             r11              = _mm256_mul_pd(rsq11,rinv11);
 398
 399             /* EWALD ELECTROSTATICS */
 400
 401             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 402             ewrt             = _mm256_mul_pd(r11,ewtabscale);
 403             ewitab           = _mm256_cvttpd_epi32(ewrt);
 404             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 405             ewitab           = _mm_slli_epi32(ewitab,2);
 406             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 407             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 408             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 409             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 410             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 411             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 412             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 413             velec            = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
 414             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
 415
 416             /* Update potential sum for this i atom from the interaction with this j atom. */
 417             velecsum         = _mm256_add_pd(velecsum,velec);
 418
 419             fscal            = felec;
 420
 421             /* Calculate temporary vectorial force */
 422             tx               = _mm256_mul_pd(fscal,dx11);
 423             ty               = _mm256_mul_pd(fscal,dy11);
 424             tz               = _mm256_mul_pd(fscal,dz11);
 425
 426             /* Update vectorial force */
 427             fix1             = _mm256_add_pd(fix1,tx);
 428             fiy1             = _mm256_add_pd(fiy1,ty);
 429             fiz1             = _mm256_add_pd(fiz1,tz);
 430
 431             fjx1             = _mm256_add_pd(fjx1,tx);
 432             fjy1             = _mm256_add_pd(fjy1,ty);
 433             fjz1             = _mm256_add_pd(fjz1,tz);
 434
 435             /**************************
 436              * CALCULATE INTERACTIONS *
 437              **************************/
 438
 439             r12              = _mm256_mul_pd(rsq12,rinv12);
 440
 441             /* EWALD ELECTROSTATICS */
 442
 443             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 444             ewrt             = _mm256_mul_pd(r12,ewtabscale);
 445             ewitab           = _mm256_cvttpd_epi32(ewrt);
 446             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 447             ewitab           = _mm_slli_epi32(ewitab,2);
 448             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 449             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 450             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 451             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 452             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 453             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 454             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 455             velec            = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
 456             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
 457
 458             /* Update potential sum for this i atom from the interaction with this j atom. */
 459             velecsum         = _mm256_add_pd(velecsum,velec);
 460
 461             fscal            = felec;
 462
 463             /* Calculate temporary vectorial force */
 464             tx               = _mm256_mul_pd(fscal,dx12);
 465             ty               = _mm256_mul_pd(fscal,dy12);
 466             tz               = _mm256_mul_pd(fscal,dz12);
 467
 468             /* Update vectorial force */
 469             fix1             = _mm256_add_pd(fix1,tx);
 470             fiy1             = _mm256_add_pd(fiy1,ty);
 471             fiz1             = _mm256_add_pd(fiz1,tz);
 472
 473             fjx2             = _mm256_add_pd(fjx2,tx);
 474             fjy2             = _mm256_add_pd(fjy2,ty);
 475             fjz2             = _mm256_add_pd(fjz2,tz);
 476
 477             /**************************
 478              * CALCULATE INTERACTIONS *
 479              **************************/
 480
 481             r13              = _mm256_mul_pd(rsq13,rinv13);
 482
 483             /* EWALD ELECTROSTATICS */
 484
 485             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 486             ewrt             = _mm256_mul_pd(r13,ewtabscale);
 487             ewitab           = _mm256_cvttpd_epi32(ewrt);
 488             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 489             ewitab           = _mm_slli_epi32(ewitab,2);
 490             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 491             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 492             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 493             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 494             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 495             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 496             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 497             velec            = _mm256_mul_pd(qq13,_mm256_sub_pd(rinv13,velec));
 498             felec            = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
 499
 500             /* Update potential sum for this i atom from the interaction with this j atom. */
 501             velecsum         = _mm256_add_pd(velecsum,velec);
 502
 503             fscal            = felec;
 504
 505             /* Calculate temporary vectorial force */
 506             tx               = _mm256_mul_pd(fscal,dx13);
 507             ty               = _mm256_mul_pd(fscal,dy13);
 508             tz               = _mm256_mul_pd(fscal,dz13);
 509
 510             /* Update vectorial force */
 511             fix1             = _mm256_add_pd(fix1,tx);
 512             fiy1             = _mm256_add_pd(fiy1,ty);
 513             fiz1             = _mm256_add_pd(fiz1,tz);
 514
 515             fjx3             = _mm256_add_pd(fjx3,tx);
 516             fjy3             = _mm256_add_pd(fjy3,ty);
 517             fjz3             = _mm256_add_pd(fjz3,tz);
 518
 519             /**************************
 520              * CALCULATE INTERACTIONS *
 521              **************************/
 522
 523             r21              = _mm256_mul_pd(rsq21,rinv21);
 524
 525             /* EWALD ELECTROSTATICS */
 526
 527             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 528             ewrt             = _mm256_mul_pd(r21,ewtabscale);
 529             ewitab           = _mm256_cvttpd_epi32(ewrt);
 530             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 531             ewitab           = _mm_slli_epi32(ewitab,2);
 532             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 533             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 534             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 535             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 536             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 537             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 538             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 539             velec            = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
 540             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
 541
 542             /* Update potential sum for this i atom from the interaction with this j atom. */
 543             velecsum         = _mm256_add_pd(velecsum,velec);
 544
 545             fscal            = felec;
 546
 547             /* Calculate temporary vectorial force */
 548             tx               = _mm256_mul_pd(fscal,dx21);
 549             ty               = _mm256_mul_pd(fscal,dy21);
 550             tz               = _mm256_mul_pd(fscal,dz21);
 551
 552             /* Update vectorial force */
 553             fix2             = _mm256_add_pd(fix2,tx);
 554             fiy2             = _mm256_add_pd(fiy2,ty);
 555             fiz2             = _mm256_add_pd(fiz2,tz);
 556
 557             fjx1             = _mm256_add_pd(fjx1,tx);
 558             fjy1             = _mm256_add_pd(fjy1,ty);
 559             fjz1             = _mm256_add_pd(fjz1,tz);
 560
 561             /**************************
 562              * CALCULATE INTERACTIONS *
 563              **************************/
 564
 565             r22              = _mm256_mul_pd(rsq22,rinv22);
 566
 567             /* EWALD ELECTROSTATICS */
 568
 569             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 570             ewrt             = _mm256_mul_pd(r22,ewtabscale);
 571             ewitab           = _mm256_cvttpd_epi32(ewrt);
 572             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 573             ewitab           = _mm_slli_epi32(ewitab,2);
 574             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 575             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 576             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 577             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 578             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 579             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 580             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 581             velec            = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
 582             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
 583
 584             /* Update potential sum for this i atom from the interaction with this j atom. */
 585             velecsum         = _mm256_add_pd(velecsum,velec);
 586
 587             fscal            = felec;
 588
 589             /* Calculate temporary vectorial force */
 590             tx               = _mm256_mul_pd(fscal,dx22);
 591             ty               = _mm256_mul_pd(fscal,dy22);
 592             tz               = _mm256_mul_pd(fscal,dz22);
 593
 594             /* Update vectorial force */
 595             fix2             = _mm256_add_pd(fix2,tx);
 596             fiy2             = _mm256_add_pd(fiy2,ty);
 597             fiz2             = _mm256_add_pd(fiz2,tz);
 598
 599             fjx2             = _mm256_add_pd(fjx2,tx);
 600             fjy2             = _mm256_add_pd(fjy2,ty);
 601             fjz2             = _mm256_add_pd(fjz2,tz);
 602
 603             /**************************
 604              * CALCULATE INTERACTIONS *
 605              **************************/
 606
 607             r23              = _mm256_mul_pd(rsq23,rinv23);
 608
 609             /* EWALD ELECTROSTATICS */
 610
 611             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 612             ewrt             = _mm256_mul_pd(r23,ewtabscale);
 613             ewitab           = _mm256_cvttpd_epi32(ewrt);
 614             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 615             ewitab           = _mm_slli_epi32(ewitab,2);
 616             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 617             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 618             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 619             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 620             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 621             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 622             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 623             velec            = _mm256_mul_pd(qq23,_mm256_sub_pd(rinv23,velec));
 624             felec            = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
 625
 626             /* Update potential sum for this i atom from the interaction with this j atom. */
 627             velecsum         = _mm256_add_pd(velecsum,velec);
 628
 629             fscal            = felec;
 630
 631             /* Calculate temporary vectorial force */
 632             tx               = _mm256_mul_pd(fscal,dx23);
 633             ty               = _mm256_mul_pd(fscal,dy23);
 634             tz               = _mm256_mul_pd(fscal,dz23);
 635
 636             /* Update vectorial force */
 637             fix2             = _mm256_add_pd(fix2,tx);
 638             fiy2             = _mm256_add_pd(fiy2,ty);
 639             fiz2             = _mm256_add_pd(fiz2,tz);
 640
 641             fjx3             = _mm256_add_pd(fjx3,tx);
 642             fjy3             = _mm256_add_pd(fjy3,ty);
 643             fjz3             = _mm256_add_pd(fjz3,tz);
 644
 645             /**************************
 646              * CALCULATE INTERACTIONS *
 647              **************************/
 648
 649             r31              = _mm256_mul_pd(rsq31,rinv31);
 650
 651             /* EWALD ELECTROSTATICS */
 652
 653             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 654             ewrt             = _mm256_mul_pd(r31,ewtabscale);
 655             ewitab           = _mm256_cvttpd_epi32(ewrt);
 656             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 657             ewitab           = _mm_slli_epi32(ewitab,2);
 658             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 659             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 660             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 661             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 662             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 663             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 664             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 665             velec            = _mm256_mul_pd(qq31,_mm256_sub_pd(rinv31,velec));
 666             felec            = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
 667
 668             /* Update potential sum for this i atom from the interaction with this j atom. */
 669             velecsum         = _mm256_add_pd(velecsum,velec);
 670
 671             fscal            = felec;
 672
 673             /* Calculate temporary vectorial force */
 674             tx               = _mm256_mul_pd(fscal,dx31);
 675             ty               = _mm256_mul_pd(fscal,dy31);
 676             tz               = _mm256_mul_pd(fscal,dz31);
 677
 678             /* Update vectorial force */
 679             fix3             = _mm256_add_pd(fix3,tx);
 680             fiy3             = _mm256_add_pd(fiy3,ty);
 681             fiz3             = _mm256_add_pd(fiz3,tz);
 682
 683             fjx1             = _mm256_add_pd(fjx1,tx);
 684             fjy1             = _mm256_add_pd(fjy1,ty);
 685             fjz1             = _mm256_add_pd(fjz1,tz);
 686
 687             /**************************
 688              * CALCULATE INTERACTIONS *
 689              **************************/
 690
 691             r32              = _mm256_mul_pd(rsq32,rinv32);
 692
 693             /* EWALD ELECTROSTATICS */
 694
 695             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 696             ewrt             = _mm256_mul_pd(r32,ewtabscale);
 697             ewitab           = _mm256_cvttpd_epi32(ewrt);
 698             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 699             ewitab           = _mm_slli_epi32(ewitab,2);
 700             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 701             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 702             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 703             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 704             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 705             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 706             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 707             velec            = _mm256_mul_pd(qq32,_mm256_sub_pd(rinv32,velec));
 708             felec            = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
 709
 710             /* Update potential sum for this i atom from the interaction with this j atom. */
 711             velecsum         = _mm256_add_pd(velecsum,velec);
 712
 713             fscal            = felec;
 714
 715             /* Calculate temporary vectorial force */
 716             tx               = _mm256_mul_pd(fscal,dx32);
 717             ty               = _mm256_mul_pd(fscal,dy32);
 718             tz               = _mm256_mul_pd(fscal,dz32);
 719
 720             /* Update vectorial force */
 721             fix3             = _mm256_add_pd(fix3,tx);
 722             fiy3             = _mm256_add_pd(fiy3,ty);
 723             fiz3             = _mm256_add_pd(fiz3,tz);
 724
 725             fjx2             = _mm256_add_pd(fjx2,tx);
 726             fjy2             = _mm256_add_pd(fjy2,ty);
 727             fjz2             = _mm256_add_pd(fjz2,tz);
 728
 729             /**************************
 730              * CALCULATE INTERACTIONS *
 731              **************************/
 732
 733             r33              = _mm256_mul_pd(rsq33,rinv33);
 734
 735             /* EWALD ELECTROSTATICS */
 736
 737             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 738             ewrt             = _mm256_mul_pd(r33,ewtabscale);
 739             ewitab           = _mm256_cvttpd_epi32(ewrt);
 740             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 741             ewitab           = _mm_slli_epi32(ewitab,2);
 742             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 743             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 744             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 745             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 746             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 747             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 748             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 749             velec            = _mm256_mul_pd(qq33,_mm256_sub_pd(rinv33,velec));
 750             felec            = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
 751
 752             /* Update potential sum for this i atom from the interaction with this j atom. */
 753             velecsum         = _mm256_add_pd(velecsum,velec);
 754
 755             fscal            = felec;
 756
 757             /* Calculate temporary vectorial force */
 758             tx               = _mm256_mul_pd(fscal,dx33);
 759             ty               = _mm256_mul_pd(fscal,dy33);
 760             tz               = _mm256_mul_pd(fscal,dz33);
 761
 762             /* Update vectorial force */
 763             fix3             = _mm256_add_pd(fix3,tx);
 764             fiy3             = _mm256_add_pd(fiy3,ty);
 765             fiz3             = _mm256_add_pd(fiz3,tz);
 766
 767             fjx3             = _mm256_add_pd(fjx3,tx);
 768             fjy3             = _mm256_add_pd(fjy3,ty);
 769             fjz3             = _mm256_add_pd(fjz3,tz);
 770
 771             fjptrA             = f+j_coord_offsetA;
 772             fjptrB             = f+j_coord_offsetB;
 773             fjptrC             = f+j_coord_offsetC;
 774             fjptrD             = f+j_coord_offsetD;
 775
 776             gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
 777                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
 778                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 779
 780             /* Inner loop uses 428 flops */
 781         }
 782
 783         if(jidx<j_index_end)
 784         {
 785
 786             /* Get j neighbor index, and coordinate index */
 787             jnrlistA         = jjnr[jidx];
 788             jnrlistB         = jjnr[jidx+1];
 789             jnrlistC         = jjnr[jidx+2];
 790             jnrlistD         = jjnr[jidx+3];
 791             /* Sign of each element will be negative for non-real atoms.
 792              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 793              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 794              */
 795             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 796
 797             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 798             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 799             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 800
 801             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 802             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 803             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 804             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 805             j_coord_offsetA  = DIM*jnrA;
 806             j_coord_offsetB  = DIM*jnrB;
 807             j_coord_offsetC  = DIM*jnrC;
 808             j_coord_offsetD  = DIM*jnrD;
 809
 810             /* load j atom coordinates */
 811             gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 812                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 813                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 814                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
 815
 816             /* Calculate displacement vector */
 817             dx00             = _mm256_sub_pd(ix0,jx0);
 818             dy00             = _mm256_sub_pd(iy0,jy0);
 819             dz00             = _mm256_sub_pd(iz0,jz0);
 820             dx11             = _mm256_sub_pd(ix1,jx1);
 821             dy11             = _mm256_sub_pd(iy1,jy1);
 822             dz11             = _mm256_sub_pd(iz1,jz1);
 823             dx12             = _mm256_sub_pd(ix1,jx2);
 824             dy12             = _mm256_sub_pd(iy1,jy2);
 825             dz12             = _mm256_sub_pd(iz1,jz2);
 826             dx13             = _mm256_sub_pd(ix1,jx3);
 827             dy13             = _mm256_sub_pd(iy1,jy3);
 828             dz13             = _mm256_sub_pd(iz1,jz3);
 829             dx21             = _mm256_sub_pd(ix2,jx1);
 830             dy21             = _mm256_sub_pd(iy2,jy1);
 831             dz21             = _mm256_sub_pd(iz2,jz1);
 832             dx22             = _mm256_sub_pd(ix2,jx2);
 833             dy22             = _mm256_sub_pd(iy2,jy2);
 834             dz22             = _mm256_sub_pd(iz2,jz2);
 835             dx23             = _mm256_sub_pd(ix2,jx3);
 836             dy23             = _mm256_sub_pd(iy2,jy3);
 837             dz23             = _mm256_sub_pd(iz2,jz3);
 838             dx31             = _mm256_sub_pd(ix3,jx1);
 839             dy31             = _mm256_sub_pd(iy3,jy1);
 840             dz31             = _mm256_sub_pd(iz3,jz1);
 841             dx32             = _mm256_sub_pd(ix3,jx2);
 842             dy32             = _mm256_sub_pd(iy3,jy2);
 843             dz32             = _mm256_sub_pd(iz3,jz2);
 844             dx33             = _mm256_sub_pd(ix3,jx3);
 845             dy33             = _mm256_sub_pd(iy3,jy3);
 846             dz33             = _mm256_sub_pd(iz3,jz3);
 847
 848             /* Calculate squared distance and things based on it */
 849             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 850             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
 851             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
 852             rsq13            = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
 853             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
 854             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
 855             rsq23            = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
 856             rsq31            = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
 857             rsq32            = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
 858             rsq33            = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
 859
 860             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 861             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
 862             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
 863             rinv13           = gmx_mm256_invsqrt_pd(rsq13);
 864             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
 865             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
 866             rinv23           = gmx_mm256_invsqrt_pd(rsq23);
 867             rinv31           = gmx_mm256_invsqrt_pd(rsq31);
 868             rinv32           = gmx_mm256_invsqrt_pd(rsq32);
 869             rinv33           = gmx_mm256_invsqrt_pd(rsq33);
 870
 871             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
 872             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
 873             rinvsq13         = _mm256_mul_pd(rinv13,rinv13);
 874             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
 875             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
 876             rinvsq23         = _mm256_mul_pd(rinv23,rinv23);
 877             rinvsq31         = _mm256_mul_pd(rinv31,rinv31);
 878             rinvsq32         = _mm256_mul_pd(rinv32,rinv32);
 879             rinvsq33         = _mm256_mul_pd(rinv33,rinv33);
 880
 881             fjx0             = _mm256_setzero_pd();
 882             fjy0             = _mm256_setzero_pd();
 883             fjz0             = _mm256_setzero_pd();
 884             fjx1             = _mm256_setzero_pd();
 885             fjy1             = _mm256_setzero_pd();
 886             fjz1             = _mm256_setzero_pd();
 887             fjx2             = _mm256_setzero_pd();
 888             fjy2             = _mm256_setzero_pd();
 889             fjz2             = _mm256_setzero_pd();
 890             fjx3             = _mm256_setzero_pd();
 891             fjy3             = _mm256_setzero_pd();
 892             fjz3             = _mm256_setzero_pd();
 893
 894             /**************************
 895              * CALCULATE INTERACTIONS *
 896              **************************/
 897
 898             r00              = _mm256_mul_pd(rsq00,rinv00);
 899             r00              = _mm256_andnot_pd(dummy_mask,r00);
 900
 901             /* Calculate table index by multiplying r with table scale and truncate to integer */
 902             rt               = _mm256_mul_pd(r00,vftabscale);
 903             vfitab           = _mm256_cvttpd_epi32(rt);
 904             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
 905             vfitab           = _mm_slli_epi32(vfitab,3);
 906
 907             /* CUBIC SPLINE TABLE DISPERSION */
 908             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 909             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 910             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 911             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 912             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 913             Heps             = _mm256_mul_pd(vfeps,H);
 914             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 915             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 916             vvdw6            = _mm256_mul_pd(c6_00,VV);
 917             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 918             fvdw6            = _mm256_mul_pd(c6_00,FF);
 919
 920             /* CUBIC SPLINE TABLE REPULSION */
 921             vfitab           = _mm_add_epi32(vfitab,ifour);
 922             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 923             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 924             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
 925             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
 926             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
 927             Heps             = _mm256_mul_pd(vfeps,H);
 928             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
 929             VV               = _mm256_add_pd(Y,_mm256_mul_pd(vfeps,Fp));
 930             vvdw12           = _mm256_mul_pd(c12_00,VV);
 931             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
 932             fvdw12           = _mm256_mul_pd(c12_00,FF);
 933             vvdw             = _mm256_add_pd(vvdw12,vvdw6);
 934             fvdw             = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_add_pd(fvdw6,fvdw12),_mm256_mul_pd(vftabscale,rinv00)));
 935
 936             /* Update potential sum for this i atom from the interaction with this j atom. */
 937             vvdw             = _mm256_andnot_pd(dummy_mask,vvdw);
 938             vvdwsum          = _mm256_add_pd(vvdwsum,vvdw);
 939
 940             fscal            = fvdw;
 941
 942             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 943
 944             /* Calculate temporary vectorial force */
 945             tx               = _mm256_mul_pd(fscal,dx00);
 946             ty               = _mm256_mul_pd(fscal,dy00);
 947             tz               = _mm256_mul_pd(fscal,dz00);
 948
 949             /* Update vectorial force */
 950             fix0             = _mm256_add_pd(fix0,tx);
 951             fiy0             = _mm256_add_pd(fiy0,ty);
 952             fiz0             = _mm256_add_pd(fiz0,tz);
 953
 954             fjx0             = _mm256_add_pd(fjx0,tx);
 955             fjy0             = _mm256_add_pd(fjy0,ty);
 956             fjz0             = _mm256_add_pd(fjz0,tz);
 957
 958             /**************************
 959              * CALCULATE INTERACTIONS *
 960              **************************/
 961
 962             r11              = _mm256_mul_pd(rsq11,rinv11);
 963             r11              = _mm256_andnot_pd(dummy_mask,r11);
 964
 965             /* EWALD ELECTROSTATICS */
 966
 967             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 968             ewrt             = _mm256_mul_pd(r11,ewtabscale);
 969             ewitab           = _mm256_cvttpd_epi32(ewrt);
 970             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 971             ewitab           = _mm_slli_epi32(ewitab,2);
 972             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 973             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 974             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 975             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 976             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 977             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 978             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 979             velec            = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
 980             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
 981
 982             /* Update potential sum for this i atom from the interaction with this j atom. */
 983             velec            = _mm256_andnot_pd(dummy_mask,velec);
 984             velecsum         = _mm256_add_pd(velecsum,velec);
 985
 986             fscal            = felec;
 987
 988             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 989
 990             /* Calculate temporary vectorial force */
 991             tx               = _mm256_mul_pd(fscal,dx11);
 992             ty               = _mm256_mul_pd(fscal,dy11);
 993             tz               = _mm256_mul_pd(fscal,dz11);
 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             fjx1             = _mm256_add_pd(fjx1,tx);
1001             fjy1             = _mm256_add_pd(fjy1,ty);
1002             fjz1             = _mm256_add_pd(fjz1,tz);
1003
1004             /**************************
1005              * CALCULATE INTERACTIONS *
1006              **************************/
1007
1008             r12              = _mm256_mul_pd(rsq12,rinv12);
1009             r12              = _mm256_andnot_pd(dummy_mask,r12);
1010
1011             /* EWALD ELECTROSTATICS */
1012
1013             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1014             ewrt             = _mm256_mul_pd(r12,ewtabscale);
1015             ewitab           = _mm256_cvttpd_epi32(ewrt);
1016             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1017             ewitab           = _mm_slli_epi32(ewitab,2);
1018             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1019             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1020             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1021             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1022             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1023             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1024             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1025             velec            = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
1026             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
1027
1028             /* Update potential sum for this i atom from the interaction with this j atom. */
1029             velec            = _mm256_andnot_pd(dummy_mask,velec);
1030             velecsum         = _mm256_add_pd(velecsum,velec);
1031
1032             fscal            = felec;
1033
1034             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1035
1036             /* Calculate temporary vectorial force */
1037             tx               = _mm256_mul_pd(fscal,dx12);
1038             ty               = _mm256_mul_pd(fscal,dy12);
1039             tz               = _mm256_mul_pd(fscal,dz12);
1040
1041             /* Update vectorial force */
1042             fix1             = _mm256_add_pd(fix1,tx);
1043             fiy1             = _mm256_add_pd(fiy1,ty);
1044             fiz1             = _mm256_add_pd(fiz1,tz);
1045
1046             fjx2             = _mm256_add_pd(fjx2,tx);
1047             fjy2             = _mm256_add_pd(fjy2,ty);
1048             fjz2             = _mm256_add_pd(fjz2,tz);
1049
1050             /**************************
1051              * CALCULATE INTERACTIONS *
1052              **************************/
1053
1054             r13              = _mm256_mul_pd(rsq13,rinv13);
1055             r13              = _mm256_andnot_pd(dummy_mask,r13);
1056
1057             /* EWALD ELECTROSTATICS */
1058
1059             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1060             ewrt             = _mm256_mul_pd(r13,ewtabscale);
1061             ewitab           = _mm256_cvttpd_epi32(ewrt);
1062             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1063             ewitab           = _mm_slli_epi32(ewitab,2);
1064             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1065             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1066             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1067             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1068             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1069             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1070             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1071             velec            = _mm256_mul_pd(qq13,_mm256_sub_pd(rinv13,velec));
1072             felec            = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
1073
1074             /* Update potential sum for this i atom from the interaction with this j atom. */
1075             velec            = _mm256_andnot_pd(dummy_mask,velec);
1076             velecsum         = _mm256_add_pd(velecsum,velec);
1077
1078             fscal            = felec;
1079
1080             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1081
1082             /* Calculate temporary vectorial force */
1083             tx               = _mm256_mul_pd(fscal,dx13);
1084             ty               = _mm256_mul_pd(fscal,dy13);
1085             tz               = _mm256_mul_pd(fscal,dz13);
1086
1087             /* Update vectorial force */
1088             fix1             = _mm256_add_pd(fix1,tx);
1089             fiy1             = _mm256_add_pd(fiy1,ty);
1090             fiz1             = _mm256_add_pd(fiz1,tz);
1091
1092             fjx3             = _mm256_add_pd(fjx3,tx);
1093             fjy3             = _mm256_add_pd(fjy3,ty);
1094             fjz3             = _mm256_add_pd(fjz3,tz);
1095
1096             /**************************
1097              * CALCULATE INTERACTIONS *
1098              **************************/
1099
1100             r21              = _mm256_mul_pd(rsq21,rinv21);
1101             r21              = _mm256_andnot_pd(dummy_mask,r21);
1102
1103             /* EWALD ELECTROSTATICS */
1104
1105             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1106             ewrt             = _mm256_mul_pd(r21,ewtabscale);
1107             ewitab           = _mm256_cvttpd_epi32(ewrt);
1108             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1109             ewitab           = _mm_slli_epi32(ewitab,2);
1110             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1111             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1112             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1113             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1114             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1115             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1116             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1117             velec            = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
1118             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1119
1120             /* Update potential sum for this i atom from the interaction with this j atom. */
1121             velec            = _mm256_andnot_pd(dummy_mask,velec);
1122             velecsum         = _mm256_add_pd(velecsum,velec);
1123
1124             fscal            = felec;
1125
1126             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1127
1128             /* Calculate temporary vectorial force */
1129             tx               = _mm256_mul_pd(fscal,dx21);
1130             ty               = _mm256_mul_pd(fscal,dy21);
1131             tz               = _mm256_mul_pd(fscal,dz21);
1132
1133             /* Update vectorial force */
1134             fix2             = _mm256_add_pd(fix2,tx);
1135             fiy2             = _mm256_add_pd(fiy2,ty);
1136             fiz2             = _mm256_add_pd(fiz2,tz);
1137
1138             fjx1             = _mm256_add_pd(fjx1,tx);
1139             fjy1             = _mm256_add_pd(fjy1,ty);
1140             fjz1             = _mm256_add_pd(fjz1,tz);
1141
1142             /**************************
1143              * CALCULATE INTERACTIONS *
1144              **************************/
1145
1146             r22              = _mm256_mul_pd(rsq22,rinv22);
1147             r22              = _mm256_andnot_pd(dummy_mask,r22);
1148
1149             /* EWALD ELECTROSTATICS */
1150
1151             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1152             ewrt             = _mm256_mul_pd(r22,ewtabscale);
1153             ewitab           = _mm256_cvttpd_epi32(ewrt);
1154             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1155             ewitab           = _mm_slli_epi32(ewitab,2);
1156             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1157             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1158             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1159             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1160             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1161             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1162             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1163             velec            = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
1164             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1165
1166             /* Update potential sum for this i atom from the interaction with this j atom. */
1167             velec            = _mm256_andnot_pd(dummy_mask,velec);
1168             velecsum         = _mm256_add_pd(velecsum,velec);
1169
1170             fscal            = felec;
1171
1172             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1173
1174             /* Calculate temporary vectorial force */
1175             tx               = _mm256_mul_pd(fscal,dx22);
1176             ty               = _mm256_mul_pd(fscal,dy22);
1177             tz               = _mm256_mul_pd(fscal,dz22);
1178
1179             /* Update vectorial force */
1180             fix2             = _mm256_add_pd(fix2,tx);
1181             fiy2             = _mm256_add_pd(fiy2,ty);
1182             fiz2             = _mm256_add_pd(fiz2,tz);
1183
1184             fjx2             = _mm256_add_pd(fjx2,tx);
1185             fjy2             = _mm256_add_pd(fjy2,ty);
1186             fjz2             = _mm256_add_pd(fjz2,tz);
1187
1188             /**************************
1189              * CALCULATE INTERACTIONS *
1190              **************************/
1191
1192             r23              = _mm256_mul_pd(rsq23,rinv23);
1193             r23              = _mm256_andnot_pd(dummy_mask,r23);
1194
1195             /* EWALD ELECTROSTATICS */
1196
1197             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1198             ewrt             = _mm256_mul_pd(r23,ewtabscale);
1199             ewitab           = _mm256_cvttpd_epi32(ewrt);
1200             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1201             ewitab           = _mm_slli_epi32(ewitab,2);
1202             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1203             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1204             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1205             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1206             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1207             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1208             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1209             velec            = _mm256_mul_pd(qq23,_mm256_sub_pd(rinv23,velec));
1210             felec            = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
1211
1212             /* Update potential sum for this i atom from the interaction with this j atom. */
1213             velec            = _mm256_andnot_pd(dummy_mask,velec);
1214             velecsum         = _mm256_add_pd(velecsum,velec);
1215
1216             fscal            = felec;
1217
1218             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1219
1220             /* Calculate temporary vectorial force */
1221             tx               = _mm256_mul_pd(fscal,dx23);
1222             ty               = _mm256_mul_pd(fscal,dy23);
1223             tz               = _mm256_mul_pd(fscal,dz23);
1224
1225             /* Update vectorial force */
1226             fix2             = _mm256_add_pd(fix2,tx);
1227             fiy2             = _mm256_add_pd(fiy2,ty);
1228             fiz2             = _mm256_add_pd(fiz2,tz);
1229
1230             fjx3             = _mm256_add_pd(fjx3,tx);
1231             fjy3             = _mm256_add_pd(fjy3,ty);
1232             fjz3             = _mm256_add_pd(fjz3,tz);
1233
1234             /**************************
1235              * CALCULATE INTERACTIONS *
1236              **************************/
1237
1238             r31              = _mm256_mul_pd(rsq31,rinv31);
1239             r31              = _mm256_andnot_pd(dummy_mask,r31);
1240
1241             /* EWALD ELECTROSTATICS */
1242
1243             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1244             ewrt             = _mm256_mul_pd(r31,ewtabscale);
1245             ewitab           = _mm256_cvttpd_epi32(ewrt);
1246             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1247             ewitab           = _mm_slli_epi32(ewitab,2);
1248             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1249             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1250             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1251             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1252             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1253             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1254             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1255             velec            = _mm256_mul_pd(qq31,_mm256_sub_pd(rinv31,velec));
1256             felec            = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
1257
1258             /* Update potential sum for this i atom from the interaction with this j atom. */
1259             velec            = _mm256_andnot_pd(dummy_mask,velec);
1260             velecsum         = _mm256_add_pd(velecsum,velec);
1261
1262             fscal            = felec;
1263
1264             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1265
1266             /* Calculate temporary vectorial force */
1267             tx               = _mm256_mul_pd(fscal,dx31);
1268             ty               = _mm256_mul_pd(fscal,dy31);
1269             tz               = _mm256_mul_pd(fscal,dz31);
1270
1271             /* Update vectorial force */
1272             fix3             = _mm256_add_pd(fix3,tx);
1273             fiy3             = _mm256_add_pd(fiy3,ty);
1274             fiz3             = _mm256_add_pd(fiz3,tz);
1275
1276             fjx1             = _mm256_add_pd(fjx1,tx);
1277             fjy1             = _mm256_add_pd(fjy1,ty);
1278             fjz1             = _mm256_add_pd(fjz1,tz);
1279
1280             /**************************
1281              * CALCULATE INTERACTIONS *
1282              **************************/
1283
1284             r32              = _mm256_mul_pd(rsq32,rinv32);
1285             r32              = _mm256_andnot_pd(dummy_mask,r32);
1286
1287             /* EWALD ELECTROSTATICS */
1288
1289             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1290             ewrt             = _mm256_mul_pd(r32,ewtabscale);
1291             ewitab           = _mm256_cvttpd_epi32(ewrt);
1292             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1293             ewitab           = _mm_slli_epi32(ewitab,2);
1294             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1295             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1296             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1297             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1298             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1299             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1300             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1301             velec            = _mm256_mul_pd(qq32,_mm256_sub_pd(rinv32,velec));
1302             felec            = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
1303
1304             /* Update potential sum for this i atom from the interaction with this j atom. */
1305             velec            = _mm256_andnot_pd(dummy_mask,velec);
1306             velecsum         = _mm256_add_pd(velecsum,velec);
1307
1308             fscal            = felec;
1309
1310             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1311
1312             /* Calculate temporary vectorial force */
1313             tx               = _mm256_mul_pd(fscal,dx32);
1314             ty               = _mm256_mul_pd(fscal,dy32);
1315             tz               = _mm256_mul_pd(fscal,dz32);
1316
1317             /* Update vectorial force */
1318             fix3             = _mm256_add_pd(fix3,tx);
1319             fiy3             = _mm256_add_pd(fiy3,ty);
1320             fiz3             = _mm256_add_pd(fiz3,tz);
1321
1322             fjx2             = _mm256_add_pd(fjx2,tx);
1323             fjy2             = _mm256_add_pd(fjy2,ty);
1324             fjz2             = _mm256_add_pd(fjz2,tz);
1325
1326             /**************************
1327              * CALCULATE INTERACTIONS *
1328              **************************/
1329
1330             r33              = _mm256_mul_pd(rsq33,rinv33);
1331             r33              = _mm256_andnot_pd(dummy_mask,r33);
1332
1333             /* EWALD ELECTROSTATICS */
1334
1335             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1336             ewrt             = _mm256_mul_pd(r33,ewtabscale);
1337             ewitab           = _mm256_cvttpd_epi32(ewrt);
1338             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1339             ewitab           = _mm_slli_epi32(ewitab,2);
1340             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1341             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1342             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1343             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1344             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1345             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1346             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1347             velec            = _mm256_mul_pd(qq33,_mm256_sub_pd(rinv33,velec));
1348             felec            = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
1349
1350             /* Update potential sum for this i atom from the interaction with this j atom. */
1351             velec            = _mm256_andnot_pd(dummy_mask,velec);
1352             velecsum         = _mm256_add_pd(velecsum,velec);
1353
1354             fscal            = felec;
1355
1356             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1357
1358             /* Calculate temporary vectorial force */
1359             tx               = _mm256_mul_pd(fscal,dx33);
1360             ty               = _mm256_mul_pd(fscal,dy33);
1361             tz               = _mm256_mul_pd(fscal,dz33);
1362
1363             /* Update vectorial force */
1364             fix3             = _mm256_add_pd(fix3,tx);
1365             fiy3             = _mm256_add_pd(fiy3,ty);
1366             fiz3             = _mm256_add_pd(fiz3,tz);
1367
1368             fjx3             = _mm256_add_pd(fjx3,tx);
1369             fjy3             = _mm256_add_pd(fjy3,ty);
1370             fjz3             = _mm256_add_pd(fjz3,tz);
1371
1372             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1373             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1374             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1375             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1376
1377             gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1378                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1379                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1380
1381             /* Inner loop uses 438 flops */
1382         }
1383
1384         /* End of innermost loop */
1385
1386         gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1387                                                  f+i_coord_offset,fshift+i_shift_offset);
1388
1389         ggid                        = gid[iidx];
1390         /* Update potential energies */
1391         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1392         gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1393
1394         /* Increment number of inner iterations */
1395         inneriter                  += j_index_end - j_index_start;
1396
1397         /* Outer loop uses 26 flops */
1398     }
1399
1400     /* Increment number of outer iterations */
1401     outeriter        += nri;
1402
1403     /* Update outer/inner flops */
1404
1405     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*438);
1406 }
1407 /*
1408  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomW4W4_F_avx_256_double
1409  * Electrostatics interaction: Ewald
1410  * VdW interaction:            CubicSplineTable
1411  * Geometry:                   Water4-Water4
1412  * Calculate force/pot:        Force
1413  */
1414 void
1415 nb_kernel_ElecEw_VdwCSTab_GeomW4W4_F_avx_256_double
1416                     (t_nblist                    * gmx_restrict       nlist,
1417                      rvec                        * gmx_restrict          xx,
1418                      rvec                        * gmx_restrict          ff,
1419                      t_forcerec                  * gmx_restrict          fr,
1420                      t_mdatoms                   * gmx_restrict     mdatoms,
1421                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1422                      t_nrnb                      * gmx_restrict        nrnb)
1423 {
1424     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1425      * just 0 for non-waters.
1426      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
1427      * jnr indices corresponding to data put in the four positions in the SIMD register.
1428      */
1429     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1430     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1431     int              jnrA,jnrB,jnrC,jnrD;
1432     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1433     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1434     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1435     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1436     real             rcutoff_scalar;
1437     real             *shiftvec,*fshift,*x,*f;
1438     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1439     real             scratch[4*DIM];
1440     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1441     real *           vdwioffsetptr0;
1442     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1443     real *           vdwioffsetptr1;
1444     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1445     real *           vdwioffsetptr2;
1446     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1447     real *           vdwioffsetptr3;
1448     __m256d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1449     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1450     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1451     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1452     __m256d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1453     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1454     __m256d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1455     int              vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
1456     __m256d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1457     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1458     __m256d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1459     __m256d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1460     __m256d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1461     __m256d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1462     __m256d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1463     __m256d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1464     __m256d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1465     __m256d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1466     __m256d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1467     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
1468     real             *charge;
1469     int              nvdwtype;
1470     __m256d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1471     int              *vdwtype;
1472     real             *vdwparam;
1473     __m256d          one_sixth   = _mm256_set1_pd(1.0/6.0);
1474     __m256d          one_twelfth = _mm256_set1_pd(1.0/12.0);
1475     __m128i          vfitab;
1476     __m128i          ifour       = _mm_set1_epi32(4);
1477     __m256d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
1478     real             *vftab;
1479     __m128i          ewitab;
1480     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1481     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
1482     real             *ewtab;
1483     __m256d          dummy_mask,cutoff_mask;
1484     __m128           tmpmask0,tmpmask1;
1485     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
1486     __m256d          one     = _mm256_set1_pd(1.0);
1487     __m256d          two     = _mm256_set1_pd(2.0);
1488     x                = xx[0];
1489     f                = ff[0];
1490
1491     nri              = nlist->nri;
1492     iinr             = nlist->iinr;
1493     jindex           = nlist->jindex;
1494     jjnr             = nlist->jjnr;
1495     shiftidx         = nlist->shift;
1496     gid              = nlist->gid;
1497     shiftvec         = fr->shift_vec[0];
1498     fshift           = fr->fshift[0];
1499     facel            = _mm256_set1_pd(fr->epsfac);
1500     charge           = mdatoms->chargeA;
1501     nvdwtype         = fr->ntype;
1502     vdwparam         = fr->nbfp;
1503     vdwtype          = mdatoms->typeA;
1504
1505     vftab            = kernel_data->table_vdw->data;
1506     vftabscale       = _mm256_set1_pd(kernel_data->table_vdw->scale);
1507
1508     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
1509     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
1510     beta2            = _mm256_mul_pd(beta,beta);
1511     beta3            = _mm256_mul_pd(beta,beta2);
1512
1513     ewtab            = fr->ic->tabq_coul_F;
1514     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
1515     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
1516
1517     /* Setup water-specific parameters */
1518     inr              = nlist->iinr[0];
1519     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
1520     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
1521     iq3              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
1522     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
1523
1524     jq1              = _mm256_set1_pd(charge[inr+1]);
1525     jq2              = _mm256_set1_pd(charge[inr+2]);
1526     jq3              = _mm256_set1_pd(charge[inr+3]);
1527     vdwjidx0A        = 2*vdwtype[inr+0];
1528     c6_00            = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
1529     c12_00           = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
1530     qq11             = _mm256_mul_pd(iq1,jq1);
1531     qq12             = _mm256_mul_pd(iq1,jq2);
1532     qq13             = _mm256_mul_pd(iq1,jq3);
1533     qq21             = _mm256_mul_pd(iq2,jq1);
1534     qq22             = _mm256_mul_pd(iq2,jq2);
1535     qq23             = _mm256_mul_pd(iq2,jq3);
1536     qq31             = _mm256_mul_pd(iq3,jq1);
1537     qq32             = _mm256_mul_pd(iq3,jq2);
1538     qq33             = _mm256_mul_pd(iq3,jq3);
1539
1540     /* Avoid stupid compiler warnings */
1541     jnrA = jnrB = jnrC = jnrD = 0;
1542     j_coord_offsetA = 0;
1543     j_coord_offsetB = 0;
1544     j_coord_offsetC = 0;
1545     j_coord_offsetD = 0;
1546
1547     outeriter        = 0;
1548     inneriter        = 0;
1549
1550     for(iidx=0;iidx<4*DIM;iidx++)
1551     {
1552         scratch[iidx] = 0.0;
1553     }
1554
1555     /* Start outer loop over neighborlists */
1556     for(iidx=0; iidx<nri; iidx++)
1557     {
1558         /* Load shift vector for this list */
1559         i_shift_offset   = DIM*shiftidx[iidx];
1560
1561         /* Load limits for loop over neighbors */
1562         j_index_start    = jindex[iidx];
1563         j_index_end      = jindex[iidx+1];
1564
1565         /* Get outer coordinate index */
1566         inr              = iinr[iidx];
1567         i_coord_offset   = DIM*inr;
1568
1569         /* Load i particle coords and add shift vector */
1570         gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1571                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1572
1573         fix0             = _mm256_setzero_pd();
1574         fiy0             = _mm256_setzero_pd();
1575         fiz0             = _mm256_setzero_pd();
1576         fix1             = _mm256_setzero_pd();
1577         fiy1             = _mm256_setzero_pd();
1578         fiz1             = _mm256_setzero_pd();
1579         fix2             = _mm256_setzero_pd();
1580         fiy2             = _mm256_setzero_pd();
1581         fiz2             = _mm256_setzero_pd();
1582         fix3             = _mm256_setzero_pd();
1583         fiy3             = _mm256_setzero_pd();
1584         fiz3             = _mm256_setzero_pd();
1585
1586         /* Start inner kernel loop */
1587         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1588         {
1589
1590             /* Get j neighbor index, and coordinate index */
1591             jnrA             = jjnr[jidx];
1592             jnrB             = jjnr[jidx+1];
1593             jnrC             = jjnr[jidx+2];
1594             jnrD             = jjnr[jidx+3];
1595             j_coord_offsetA  = DIM*jnrA;
1596             j_coord_offsetB  = DIM*jnrB;
1597             j_coord_offsetC  = DIM*jnrC;
1598             j_coord_offsetD  = DIM*jnrD;
1599
1600             /* load j atom coordinates */
1601             gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1602                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1603                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1604                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
1605
1606             /* Calculate displacement vector */
1607             dx00             = _mm256_sub_pd(ix0,jx0);
1608             dy00             = _mm256_sub_pd(iy0,jy0);
1609             dz00             = _mm256_sub_pd(iz0,jz0);
1610             dx11             = _mm256_sub_pd(ix1,jx1);
1611             dy11             = _mm256_sub_pd(iy1,jy1);
1612             dz11             = _mm256_sub_pd(iz1,jz1);
1613             dx12             = _mm256_sub_pd(ix1,jx2);
1614             dy12             = _mm256_sub_pd(iy1,jy2);
1615             dz12             = _mm256_sub_pd(iz1,jz2);
1616             dx13             = _mm256_sub_pd(ix1,jx3);
1617             dy13             = _mm256_sub_pd(iy1,jy3);
1618             dz13             = _mm256_sub_pd(iz1,jz3);
1619             dx21             = _mm256_sub_pd(ix2,jx1);
1620             dy21             = _mm256_sub_pd(iy2,jy1);
1621             dz21             = _mm256_sub_pd(iz2,jz1);
1622             dx22             = _mm256_sub_pd(ix2,jx2);
1623             dy22             = _mm256_sub_pd(iy2,jy2);
1624             dz22             = _mm256_sub_pd(iz2,jz2);
1625             dx23             = _mm256_sub_pd(ix2,jx3);
1626             dy23             = _mm256_sub_pd(iy2,jy3);
1627             dz23             = _mm256_sub_pd(iz2,jz3);
1628             dx31             = _mm256_sub_pd(ix3,jx1);
1629             dy31             = _mm256_sub_pd(iy3,jy1);
1630             dz31             = _mm256_sub_pd(iz3,jz1);
1631             dx32             = _mm256_sub_pd(ix3,jx2);
1632             dy32             = _mm256_sub_pd(iy3,jy2);
1633             dz32             = _mm256_sub_pd(iz3,jz2);
1634             dx33             = _mm256_sub_pd(ix3,jx3);
1635             dy33             = _mm256_sub_pd(iy3,jy3);
1636             dz33             = _mm256_sub_pd(iz3,jz3);
1637
1638             /* Calculate squared distance and things based on it */
1639             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1640             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1641             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1642             rsq13            = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
1643             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1644             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1645             rsq23            = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
1646             rsq31            = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
1647             rsq32            = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
1648             rsq33            = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
1649
1650             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
1651             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
1652             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
1653             rinv13           = gmx_mm256_invsqrt_pd(rsq13);
1654             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
1655             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
1656             rinv23           = gmx_mm256_invsqrt_pd(rsq23);
1657             rinv31           = gmx_mm256_invsqrt_pd(rsq31);
1658             rinv32           = gmx_mm256_invsqrt_pd(rsq32);
1659             rinv33           = gmx_mm256_invsqrt_pd(rsq33);
1660
1661             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
1662             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
1663             rinvsq13         = _mm256_mul_pd(rinv13,rinv13);
1664             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
1665             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
1666             rinvsq23         = _mm256_mul_pd(rinv23,rinv23);
1667             rinvsq31         = _mm256_mul_pd(rinv31,rinv31);
1668             rinvsq32         = _mm256_mul_pd(rinv32,rinv32);
1669             rinvsq33         = _mm256_mul_pd(rinv33,rinv33);
1670
1671             fjx0             = _mm256_setzero_pd();
1672             fjy0             = _mm256_setzero_pd();
1673             fjz0             = _mm256_setzero_pd();
1674             fjx1             = _mm256_setzero_pd();
1675             fjy1             = _mm256_setzero_pd();
1676             fjz1             = _mm256_setzero_pd();
1677             fjx2             = _mm256_setzero_pd();
1678             fjy2             = _mm256_setzero_pd();
1679             fjz2             = _mm256_setzero_pd();
1680             fjx3             = _mm256_setzero_pd();
1681             fjy3             = _mm256_setzero_pd();
1682             fjz3             = _mm256_setzero_pd();
1683
1684             /**************************
1685              * CALCULATE INTERACTIONS *
1686              **************************/
1687
1688             r00              = _mm256_mul_pd(rsq00,rinv00);
1689
1690             /* Calculate table index by multiplying r with table scale and truncate to integer */
1691             rt               = _mm256_mul_pd(r00,vftabscale);
1692             vfitab           = _mm256_cvttpd_epi32(rt);
1693             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
1694             vfitab           = _mm_slli_epi32(vfitab,3);
1695
1696             /* CUBIC SPLINE TABLE DISPERSION */
1697             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1698             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1699             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1700             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1701             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1702             Heps             = _mm256_mul_pd(vfeps,H);
1703             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1704             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1705             fvdw6            = _mm256_mul_pd(c6_00,FF);
1706
1707             /* CUBIC SPLINE TABLE REPULSION */
1708             vfitab           = _mm_add_epi32(vfitab,ifour);
1709             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
1710             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
1711             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
1712             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
1713             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
1714             Heps             = _mm256_mul_pd(vfeps,H);
1715             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
1716             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
1717             fvdw12           = _mm256_mul_pd(c12_00,FF);
1718             fvdw             = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_add_pd(fvdw6,fvdw12),_mm256_mul_pd(vftabscale,rinv00)));
1719
1720             fscal            = fvdw;
1721
1722             /* Calculate temporary vectorial force */
1723             tx               = _mm256_mul_pd(fscal,dx00);
1724             ty               = _mm256_mul_pd(fscal,dy00);
1725             tz               = _mm256_mul_pd(fscal,dz00);
1726
1727             /* Update vectorial force */
1728             fix0             = _mm256_add_pd(fix0,tx);
1729             fiy0             = _mm256_add_pd(fiy0,ty);
1730             fiz0             = _mm256_add_pd(fiz0,tz);
1731
1732             fjx0             = _mm256_add_pd(fjx0,tx);
1733             fjy0             = _mm256_add_pd(fjy0,ty);
1734             fjz0             = _mm256_add_pd(fjz0,tz);
1735
1736             /**************************
1737              * CALCULATE INTERACTIONS *
1738              **************************/
1739
1740             r11              = _mm256_mul_pd(rsq11,rinv11);
1741
1742             /* EWALD ELECTROSTATICS */
1743
1744             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1745             ewrt             = _mm256_mul_pd(r11,ewtabscale);
1746             ewitab           = _mm256_cvttpd_epi32(ewrt);
1747             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1748             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1749                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1750                                             &ewtabF,&ewtabFn);
1751             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1752             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
1753
1754             fscal            = felec;
1755
1756             /* Calculate temporary vectorial force */
1757             tx               = _mm256_mul_pd(fscal,dx11);
1758             ty               = _mm256_mul_pd(fscal,dy11);
1759             tz               = _mm256_mul_pd(fscal,dz11);
1760
1761             /* Update vectorial force */
1762             fix1             = _mm256_add_pd(fix1,tx);
1763             fiy1             = _mm256_add_pd(fiy1,ty);
1764             fiz1             = _mm256_add_pd(fiz1,tz);
1765
1766             fjx1             = _mm256_add_pd(fjx1,tx);
1767             fjy1             = _mm256_add_pd(fjy1,ty);
1768             fjz1             = _mm256_add_pd(fjz1,tz);
1769
1770             /**************************
1771              * CALCULATE INTERACTIONS *
1772              **************************/
1773
1774             r12              = _mm256_mul_pd(rsq12,rinv12);
1775
1776             /* EWALD ELECTROSTATICS */
1777
1778             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1779             ewrt             = _mm256_mul_pd(r12,ewtabscale);
1780             ewitab           = _mm256_cvttpd_epi32(ewrt);
1781             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1782             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1783                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1784                                             &ewtabF,&ewtabFn);
1785             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1786             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
1787
1788             fscal            = felec;
1789
1790             /* Calculate temporary vectorial force */
1791             tx               = _mm256_mul_pd(fscal,dx12);
1792             ty               = _mm256_mul_pd(fscal,dy12);
1793             tz               = _mm256_mul_pd(fscal,dz12);
1794
1795             /* Update vectorial force */
1796             fix1             = _mm256_add_pd(fix1,tx);
1797             fiy1             = _mm256_add_pd(fiy1,ty);
1798             fiz1             = _mm256_add_pd(fiz1,tz);
1799
1800             fjx2             = _mm256_add_pd(fjx2,tx);
1801             fjy2             = _mm256_add_pd(fjy2,ty);
1802             fjz2             = _mm256_add_pd(fjz2,tz);
1803
1804             /**************************
1805              * CALCULATE INTERACTIONS *
1806              **************************/
1807
1808             r13              = _mm256_mul_pd(rsq13,rinv13);
1809
1810             /* EWALD ELECTROSTATICS */
1811
1812             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1813             ewrt             = _mm256_mul_pd(r13,ewtabscale);
1814             ewitab           = _mm256_cvttpd_epi32(ewrt);
1815             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1816             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1817                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1818                                             &ewtabF,&ewtabFn);
1819             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1820             felec            = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
1821
1822             fscal            = felec;
1823
1824             /* Calculate temporary vectorial force */
1825             tx               = _mm256_mul_pd(fscal,dx13);
1826             ty               = _mm256_mul_pd(fscal,dy13);
1827             tz               = _mm256_mul_pd(fscal,dz13);
1828
1829             /* Update vectorial force */
1830             fix1             = _mm256_add_pd(fix1,tx);
1831             fiy1             = _mm256_add_pd(fiy1,ty);
1832             fiz1             = _mm256_add_pd(fiz1,tz);
1833
1834             fjx3             = _mm256_add_pd(fjx3,tx);
1835             fjy3             = _mm256_add_pd(fjy3,ty);
1836             fjz3             = _mm256_add_pd(fjz3,tz);
1837
1838             /**************************
1839              * CALCULATE INTERACTIONS *
1840              **************************/
1841
1842             r21              = _mm256_mul_pd(rsq21,rinv21);
1843
1844             /* EWALD ELECTROSTATICS */
1845
1846             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1847             ewrt             = _mm256_mul_pd(r21,ewtabscale);
1848             ewitab           = _mm256_cvttpd_epi32(ewrt);
1849             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1850             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1851                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1852                                             &ewtabF,&ewtabFn);
1853             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1854             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1855
1856             fscal            = felec;
1857
1858             /* Calculate temporary vectorial force */
1859             tx               = _mm256_mul_pd(fscal,dx21);
1860             ty               = _mm256_mul_pd(fscal,dy21);
1861             tz               = _mm256_mul_pd(fscal,dz21);
1862
1863             /* Update vectorial force */
1864             fix2             = _mm256_add_pd(fix2,tx);
1865             fiy2             = _mm256_add_pd(fiy2,ty);
1866             fiz2             = _mm256_add_pd(fiz2,tz);
1867
1868             fjx1             = _mm256_add_pd(fjx1,tx);
1869             fjy1             = _mm256_add_pd(fjy1,ty);
1870             fjz1             = _mm256_add_pd(fjz1,tz);
1871
1872             /**************************
1873              * CALCULATE INTERACTIONS *
1874              **************************/
1875
1876             r22              = _mm256_mul_pd(rsq22,rinv22);
1877
1878             /* EWALD ELECTROSTATICS */
1879
1880             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1881             ewrt             = _mm256_mul_pd(r22,ewtabscale);
1882             ewitab           = _mm256_cvttpd_epi32(ewrt);
1883             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1884             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1885                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1886                                             &ewtabF,&ewtabFn);
1887             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1888             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1889
1890             fscal            = felec;
1891
1892             /* Calculate temporary vectorial force */
1893             tx               = _mm256_mul_pd(fscal,dx22);
1894             ty               = _mm256_mul_pd(fscal,dy22);
1895             tz               = _mm256_mul_pd(fscal,dz22);
1896
1897             /* Update vectorial force */
1898             fix2             = _mm256_add_pd(fix2,tx);
1899             fiy2             = _mm256_add_pd(fiy2,ty);
1900             fiz2             = _mm256_add_pd(fiz2,tz);
1901
1902             fjx2             = _mm256_add_pd(fjx2,tx);
1903             fjy2             = _mm256_add_pd(fjy2,ty);
1904             fjz2             = _mm256_add_pd(fjz2,tz);
1905
1906             /**************************
1907              * CALCULATE INTERACTIONS *
1908              **************************/
1909
1910             r23              = _mm256_mul_pd(rsq23,rinv23);
1911
1912             /* EWALD ELECTROSTATICS */
1913
1914             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1915             ewrt             = _mm256_mul_pd(r23,ewtabscale);
1916             ewitab           = _mm256_cvttpd_epi32(ewrt);
1917             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1918             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1919                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1920                                             &ewtabF,&ewtabFn);
1921             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1922             felec            = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
1923
1924             fscal            = felec;
1925
1926             /* Calculate temporary vectorial force */
1927             tx               = _mm256_mul_pd(fscal,dx23);
1928             ty               = _mm256_mul_pd(fscal,dy23);
1929             tz               = _mm256_mul_pd(fscal,dz23);
1930
1931             /* Update vectorial force */
1932             fix2             = _mm256_add_pd(fix2,tx);
1933             fiy2             = _mm256_add_pd(fiy2,ty);
1934             fiz2             = _mm256_add_pd(fiz2,tz);
1935
1936             fjx3             = _mm256_add_pd(fjx3,tx);
1937             fjy3             = _mm256_add_pd(fjy3,ty);
1938             fjz3             = _mm256_add_pd(fjz3,tz);
1939
1940             /**************************
1941              * CALCULATE INTERACTIONS *
1942              **************************/
1943
1944             r31              = _mm256_mul_pd(rsq31,rinv31);
1945
1946             /* EWALD ELECTROSTATICS */
1947
1948             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1949             ewrt             = _mm256_mul_pd(r31,ewtabscale);
1950             ewitab           = _mm256_cvttpd_epi32(ewrt);
1951             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1952             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1953                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1954                                             &ewtabF,&ewtabFn);
1955             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1956             felec            = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
1957
1958             fscal            = felec;
1959
1960             /* Calculate temporary vectorial force */
1961             tx               = _mm256_mul_pd(fscal,dx31);
1962             ty               = _mm256_mul_pd(fscal,dy31);
1963             tz               = _mm256_mul_pd(fscal,dz31);
1964
1965             /* Update vectorial force */
1966             fix3             = _mm256_add_pd(fix3,tx);
1967             fiy3             = _mm256_add_pd(fiy3,ty);
1968             fiz3             = _mm256_add_pd(fiz3,tz);
1969
1970             fjx1             = _mm256_add_pd(fjx1,tx);
1971             fjy1             = _mm256_add_pd(fjy1,ty);
1972             fjz1             = _mm256_add_pd(fjz1,tz);
1973
1974             /**************************
1975              * CALCULATE INTERACTIONS *
1976              **************************/
1977
1978             r32              = _mm256_mul_pd(rsq32,rinv32);
1979
1980             /* EWALD ELECTROSTATICS */
1981
1982             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1983             ewrt             = _mm256_mul_pd(r32,ewtabscale);
1984             ewitab           = _mm256_cvttpd_epi32(ewrt);
1985             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1986             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1987                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1988                                             &ewtabF,&ewtabFn);
1989             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1990             felec            = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
1991
1992             fscal            = felec;
1993
1994             /* Calculate temporary vectorial force */
1995             tx               = _mm256_mul_pd(fscal,dx32);
1996             ty               = _mm256_mul_pd(fscal,dy32);
1997             tz               = _mm256_mul_pd(fscal,dz32);
1998
1999             /* Update vectorial force */
2000             fix3             = _mm256_add_pd(fix3,tx);
2001             fiy3             = _mm256_add_pd(fiy3,ty);
2002             fiz3             = _mm256_add_pd(fiz3,tz);
2003
2004             fjx2             = _mm256_add_pd(fjx2,tx);
2005             fjy2             = _mm256_add_pd(fjy2,ty);
2006             fjz2             = _mm256_add_pd(fjz2,tz);
2007
2008             /**************************
2009              * CALCULATE INTERACTIONS *
2010              **************************/
2011
2012             r33              = _mm256_mul_pd(rsq33,rinv33);
2013
2014             /* EWALD ELECTROSTATICS */
2015
2016             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2017             ewrt             = _mm256_mul_pd(r33,ewtabscale);
2018             ewitab           = _mm256_cvttpd_epi32(ewrt);
2019             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2020             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2021                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2022                                             &ewtabF,&ewtabFn);
2023             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2024             felec            = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
2025
2026             fscal            = felec;
2027
2028             /* Calculate temporary vectorial force */
2029             tx               = _mm256_mul_pd(fscal,dx33);
2030             ty               = _mm256_mul_pd(fscal,dy33);
2031             tz               = _mm256_mul_pd(fscal,dz33);
2032
2033             /* Update vectorial force */
2034             fix3             = _mm256_add_pd(fix3,tx);
2035             fiy3             = _mm256_add_pd(fiy3,ty);
2036             fiz3             = _mm256_add_pd(fiz3,tz);
2037
2038             fjx3             = _mm256_add_pd(fjx3,tx);
2039             fjy3             = _mm256_add_pd(fjy3,ty);
2040             fjz3             = _mm256_add_pd(fjz3,tz);
2041
2042             fjptrA             = f+j_coord_offsetA;
2043             fjptrB             = f+j_coord_offsetB;
2044             fjptrC             = f+j_coord_offsetC;
2045             fjptrD             = f+j_coord_offsetD;
2046
2047             gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
2048                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
2049                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2050
2051             /* Inner loop uses 375 flops */
2052         }
2053
2054         if(jidx<j_index_end)
2055         {
2056
2057             /* Get j neighbor index, and coordinate index */
2058             jnrlistA         = jjnr[jidx];
2059             jnrlistB         = jjnr[jidx+1];
2060             jnrlistC         = jjnr[jidx+2];
2061             jnrlistD         = jjnr[jidx+3];
2062             /* Sign of each element will be negative for non-real atoms.
2063              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
2064              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
2065              */
2066             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
2067
2068             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
2069             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
2070             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
2071
2072             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
2073             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
2074             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
2075             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
2076             j_coord_offsetA  = DIM*jnrA;
2077             j_coord_offsetB  = DIM*jnrB;
2078             j_coord_offsetC  = DIM*jnrC;
2079             j_coord_offsetD  = DIM*jnrD;
2080
2081             /* load j atom coordinates */
2082             gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
2083                                                  x+j_coord_offsetC,x+j_coord_offsetD,
2084                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
2085                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
2086
2087             /* Calculate displacement vector */
2088             dx00             = _mm256_sub_pd(ix0,jx0);
2089             dy00             = _mm256_sub_pd(iy0,jy0);
2090             dz00             = _mm256_sub_pd(iz0,jz0);
2091             dx11             = _mm256_sub_pd(ix1,jx1);
2092             dy11             = _mm256_sub_pd(iy1,jy1);
2093             dz11             = _mm256_sub_pd(iz1,jz1);
2094             dx12             = _mm256_sub_pd(ix1,jx2);
2095             dy12             = _mm256_sub_pd(iy1,jy2);
2096             dz12             = _mm256_sub_pd(iz1,jz2);
2097             dx13             = _mm256_sub_pd(ix1,jx3);
2098             dy13             = _mm256_sub_pd(iy1,jy3);
2099             dz13             = _mm256_sub_pd(iz1,jz3);
2100             dx21             = _mm256_sub_pd(ix2,jx1);
2101             dy21             = _mm256_sub_pd(iy2,jy1);
2102             dz21             = _mm256_sub_pd(iz2,jz1);
2103             dx22             = _mm256_sub_pd(ix2,jx2);
2104             dy22             = _mm256_sub_pd(iy2,jy2);
2105             dz22             = _mm256_sub_pd(iz2,jz2);
2106             dx23             = _mm256_sub_pd(ix2,jx3);
2107             dy23             = _mm256_sub_pd(iy2,jy3);
2108             dz23             = _mm256_sub_pd(iz2,jz3);
2109             dx31             = _mm256_sub_pd(ix3,jx1);
2110             dy31             = _mm256_sub_pd(iy3,jy1);
2111             dz31             = _mm256_sub_pd(iz3,jz1);
2112             dx32             = _mm256_sub_pd(ix3,jx2);
2113             dy32             = _mm256_sub_pd(iy3,jy2);
2114             dz32             = _mm256_sub_pd(iz3,jz2);
2115             dx33             = _mm256_sub_pd(ix3,jx3);
2116             dy33             = _mm256_sub_pd(iy3,jy3);
2117             dz33             = _mm256_sub_pd(iz3,jz3);
2118
2119             /* Calculate squared distance and things based on it */
2120             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
2121             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
2122             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
2123             rsq13            = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
2124             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
2125             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
2126             rsq23            = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
2127             rsq31            = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
2128             rsq32            = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
2129             rsq33            = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
2130
2131             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
2132             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
2133             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
2134             rinv13           = gmx_mm256_invsqrt_pd(rsq13);
2135             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
2136             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
2137             rinv23           = gmx_mm256_invsqrt_pd(rsq23);
2138             rinv31           = gmx_mm256_invsqrt_pd(rsq31);
2139             rinv32           = gmx_mm256_invsqrt_pd(rsq32);
2140             rinv33           = gmx_mm256_invsqrt_pd(rsq33);
2141
2142             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
2143             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
2144             rinvsq13         = _mm256_mul_pd(rinv13,rinv13);
2145             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
2146             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
2147             rinvsq23         = _mm256_mul_pd(rinv23,rinv23);
2148             rinvsq31         = _mm256_mul_pd(rinv31,rinv31);
2149             rinvsq32         = _mm256_mul_pd(rinv32,rinv32);
2150             rinvsq33         = _mm256_mul_pd(rinv33,rinv33);
2151
2152             fjx0             = _mm256_setzero_pd();
2153             fjy0             = _mm256_setzero_pd();
2154             fjz0             = _mm256_setzero_pd();
2155             fjx1             = _mm256_setzero_pd();
2156             fjy1             = _mm256_setzero_pd();
2157             fjz1             = _mm256_setzero_pd();
2158             fjx2             = _mm256_setzero_pd();
2159             fjy2             = _mm256_setzero_pd();
2160             fjz2             = _mm256_setzero_pd();
2161             fjx3             = _mm256_setzero_pd();
2162             fjy3             = _mm256_setzero_pd();
2163             fjz3             = _mm256_setzero_pd();
2164
2165             /**************************
2166              * CALCULATE INTERACTIONS *
2167              **************************/
2168
2169             r00              = _mm256_mul_pd(rsq00,rinv00);
2170             r00              = _mm256_andnot_pd(dummy_mask,r00);
2171
2172             /* Calculate table index by multiplying r with table scale and truncate to integer */
2173             rt               = _mm256_mul_pd(r00,vftabscale);
2174             vfitab           = _mm256_cvttpd_epi32(rt);
2175             vfeps            = _mm256_sub_pd(rt,_mm256_round_pd(rt, _MM_FROUND_FLOOR));
2176             vfitab           = _mm_slli_epi32(vfitab,3);
2177
2178             /* CUBIC SPLINE TABLE DISPERSION */
2179             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2180             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
2181             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
2182             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
2183             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
2184             Heps             = _mm256_mul_pd(vfeps,H);
2185             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
2186             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
2187             fvdw6            = _mm256_mul_pd(c6_00,FF);
2188
2189             /* CUBIC SPLINE TABLE REPULSION */
2190             vfitab           = _mm_add_epi32(vfitab,ifour);
2191             Y                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
2192             F                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
2193             G                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,2) );
2194             H                = _mm256_load_pd( vftab + _mm_extract_epi32(vfitab,3) );
2195             GMX_MM256_FULLTRANSPOSE4_PD(Y,F,G,H);
2196             Heps             = _mm256_mul_pd(vfeps,H);
2197             Fp               = _mm256_add_pd(F,_mm256_mul_pd(vfeps,_mm256_add_pd(G,Heps)));
2198             FF               = _mm256_add_pd(Fp,_mm256_mul_pd(vfeps,_mm256_add_pd(G,_mm256_add_pd(Heps,Heps))));
2199             fvdw12           = _mm256_mul_pd(c12_00,FF);
2200             fvdw             = _mm256_xor_pd(signbit,_mm256_mul_pd(_mm256_add_pd(fvdw6,fvdw12),_mm256_mul_pd(vftabscale,rinv00)));
2201
2202             fscal            = fvdw;
2203
2204             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2205
2206             /* Calculate temporary vectorial force */
2207             tx               = _mm256_mul_pd(fscal,dx00);
2208             ty               = _mm256_mul_pd(fscal,dy00);
2209             tz               = _mm256_mul_pd(fscal,dz00);
2210
2211             /* Update vectorial force */
2212             fix0             = _mm256_add_pd(fix0,tx);
2213             fiy0             = _mm256_add_pd(fiy0,ty);
2214             fiz0             = _mm256_add_pd(fiz0,tz);
2215
2216             fjx0             = _mm256_add_pd(fjx0,tx);
2217             fjy0             = _mm256_add_pd(fjy0,ty);
2218             fjz0             = _mm256_add_pd(fjz0,tz);
2219
2220             /**************************
2221              * CALCULATE INTERACTIONS *
2222              **************************/
2223
2224             r11              = _mm256_mul_pd(rsq11,rinv11);
2225             r11              = _mm256_andnot_pd(dummy_mask,r11);
2226
2227             /* EWALD ELECTROSTATICS */
2228
2229             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2230             ewrt             = _mm256_mul_pd(r11,ewtabscale);
2231             ewitab           = _mm256_cvttpd_epi32(ewrt);
2232             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2233             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2234                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2235                                             &ewtabF,&ewtabFn);
2236             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2237             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
2238
2239             fscal            = felec;
2240
2241             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2242
2243             /* Calculate temporary vectorial force */
2244             tx               = _mm256_mul_pd(fscal,dx11);
2245             ty               = _mm256_mul_pd(fscal,dy11);
2246             tz               = _mm256_mul_pd(fscal,dz11);
2247
2248             /* Update vectorial force */
2249             fix1             = _mm256_add_pd(fix1,tx);
2250             fiy1             = _mm256_add_pd(fiy1,ty);
2251             fiz1             = _mm256_add_pd(fiz1,tz);
2252
2253             fjx1             = _mm256_add_pd(fjx1,tx);
2254             fjy1             = _mm256_add_pd(fjy1,ty);
2255             fjz1             = _mm256_add_pd(fjz1,tz);
2256
2257             /**************************
2258              * CALCULATE INTERACTIONS *
2259              **************************/
2260
2261             r12              = _mm256_mul_pd(rsq12,rinv12);
2262             r12              = _mm256_andnot_pd(dummy_mask,r12);
2263
2264             /* EWALD ELECTROSTATICS */
2265
2266             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2267             ewrt             = _mm256_mul_pd(r12,ewtabscale);
2268             ewitab           = _mm256_cvttpd_epi32(ewrt);
2269             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2270             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2271                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2272                                             &ewtabF,&ewtabFn);
2273             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2274             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
2275
2276             fscal            = felec;
2277
2278             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2279
2280             /* Calculate temporary vectorial force */
2281             tx               = _mm256_mul_pd(fscal,dx12);
2282             ty               = _mm256_mul_pd(fscal,dy12);
2283             tz               = _mm256_mul_pd(fscal,dz12);
2284
2285             /* Update vectorial force */
2286             fix1             = _mm256_add_pd(fix1,tx);
2287             fiy1             = _mm256_add_pd(fiy1,ty);
2288             fiz1             = _mm256_add_pd(fiz1,tz);
2289
2290             fjx2             = _mm256_add_pd(fjx2,tx);
2291             fjy2             = _mm256_add_pd(fjy2,ty);
2292             fjz2             = _mm256_add_pd(fjz2,tz);
2293
2294             /**************************
2295              * CALCULATE INTERACTIONS *
2296              **************************/
2297
2298             r13              = _mm256_mul_pd(rsq13,rinv13);
2299             r13              = _mm256_andnot_pd(dummy_mask,r13);
2300
2301             /* EWALD ELECTROSTATICS */
2302
2303             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2304             ewrt             = _mm256_mul_pd(r13,ewtabscale);
2305             ewitab           = _mm256_cvttpd_epi32(ewrt);
2306             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2307             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2308                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2309                                             &ewtabF,&ewtabFn);
2310             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2311             felec            = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
2312
2313             fscal            = felec;
2314
2315             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2316
2317             /* Calculate temporary vectorial force */
2318             tx               = _mm256_mul_pd(fscal,dx13);
2319             ty               = _mm256_mul_pd(fscal,dy13);
2320             tz               = _mm256_mul_pd(fscal,dz13);
2321
2322             /* Update vectorial force */
2323             fix1             = _mm256_add_pd(fix1,tx);
2324             fiy1             = _mm256_add_pd(fiy1,ty);
2325             fiz1             = _mm256_add_pd(fiz1,tz);
2326
2327             fjx3             = _mm256_add_pd(fjx3,tx);
2328             fjy3             = _mm256_add_pd(fjy3,ty);
2329             fjz3             = _mm256_add_pd(fjz3,tz);
2330
2331             /**************************
2332              * CALCULATE INTERACTIONS *
2333              **************************/
2334
2335             r21              = _mm256_mul_pd(rsq21,rinv21);
2336             r21              = _mm256_andnot_pd(dummy_mask,r21);
2337
2338             /* EWALD ELECTROSTATICS */
2339
2340             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2341             ewrt             = _mm256_mul_pd(r21,ewtabscale);
2342             ewitab           = _mm256_cvttpd_epi32(ewrt);
2343             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2344             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2345                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2346                                             &ewtabF,&ewtabFn);
2347             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2348             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
2349
2350             fscal            = felec;
2351
2352             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2353
2354             /* Calculate temporary vectorial force */
2355             tx               = _mm256_mul_pd(fscal,dx21);
2356             ty               = _mm256_mul_pd(fscal,dy21);
2357             tz               = _mm256_mul_pd(fscal,dz21);
2358
2359             /* Update vectorial force */
2360             fix2             = _mm256_add_pd(fix2,tx);
2361             fiy2             = _mm256_add_pd(fiy2,ty);
2362             fiz2             = _mm256_add_pd(fiz2,tz);
2363
2364             fjx1             = _mm256_add_pd(fjx1,tx);
2365             fjy1             = _mm256_add_pd(fjy1,ty);
2366             fjz1             = _mm256_add_pd(fjz1,tz);
2367
2368             /**************************
2369              * CALCULATE INTERACTIONS *
2370              **************************/
2371
2372             r22              = _mm256_mul_pd(rsq22,rinv22);
2373             r22              = _mm256_andnot_pd(dummy_mask,r22);
2374
2375             /* EWALD ELECTROSTATICS */
2376
2377             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2378             ewrt             = _mm256_mul_pd(r22,ewtabscale);
2379             ewitab           = _mm256_cvttpd_epi32(ewrt);
2380             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2381             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2382                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2383                                             &ewtabF,&ewtabFn);
2384             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2385             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
2386
2387             fscal            = felec;
2388
2389             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2390
2391             /* Calculate temporary vectorial force */
2392             tx               = _mm256_mul_pd(fscal,dx22);
2393             ty               = _mm256_mul_pd(fscal,dy22);
2394             tz               = _mm256_mul_pd(fscal,dz22);
2395
2396             /* Update vectorial force */
2397             fix2             = _mm256_add_pd(fix2,tx);
2398             fiy2             = _mm256_add_pd(fiy2,ty);
2399             fiz2             = _mm256_add_pd(fiz2,tz);
2400
2401             fjx2             = _mm256_add_pd(fjx2,tx);
2402             fjy2             = _mm256_add_pd(fjy2,ty);
2403             fjz2             = _mm256_add_pd(fjz2,tz);
2404
2405             /**************************
2406              * CALCULATE INTERACTIONS *
2407              **************************/
2408
2409             r23              = _mm256_mul_pd(rsq23,rinv23);
2410             r23              = _mm256_andnot_pd(dummy_mask,r23);
2411
2412             /* EWALD ELECTROSTATICS */
2413
2414             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2415             ewrt             = _mm256_mul_pd(r23,ewtabscale);
2416             ewitab           = _mm256_cvttpd_epi32(ewrt);
2417             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2418             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2419                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2420                                             &ewtabF,&ewtabFn);
2421             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2422             felec            = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
2423
2424             fscal            = felec;
2425
2426             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2427
2428             /* Calculate temporary vectorial force */
2429             tx               = _mm256_mul_pd(fscal,dx23);
2430             ty               = _mm256_mul_pd(fscal,dy23);
2431             tz               = _mm256_mul_pd(fscal,dz23);
2432
2433             /* Update vectorial force */
2434             fix2             = _mm256_add_pd(fix2,tx);
2435             fiy2             = _mm256_add_pd(fiy2,ty);
2436             fiz2             = _mm256_add_pd(fiz2,tz);
2437
2438             fjx3             = _mm256_add_pd(fjx3,tx);
2439             fjy3             = _mm256_add_pd(fjy3,ty);
2440             fjz3             = _mm256_add_pd(fjz3,tz);
2441
2442             /**************************
2443              * CALCULATE INTERACTIONS *
2444              **************************/
2445
2446             r31              = _mm256_mul_pd(rsq31,rinv31);
2447             r31              = _mm256_andnot_pd(dummy_mask,r31);
2448
2449             /* EWALD ELECTROSTATICS */
2450
2451             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2452             ewrt             = _mm256_mul_pd(r31,ewtabscale);
2453             ewitab           = _mm256_cvttpd_epi32(ewrt);
2454             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2455             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2456                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2457                                             &ewtabF,&ewtabFn);
2458             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2459             felec            = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
2460
2461             fscal            = felec;
2462
2463             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2464
2465             /* Calculate temporary vectorial force */
2466             tx               = _mm256_mul_pd(fscal,dx31);
2467             ty               = _mm256_mul_pd(fscal,dy31);
2468             tz               = _mm256_mul_pd(fscal,dz31);
2469
2470             /* Update vectorial force */
2471             fix3             = _mm256_add_pd(fix3,tx);
2472             fiy3             = _mm256_add_pd(fiy3,ty);
2473             fiz3             = _mm256_add_pd(fiz3,tz);
2474
2475             fjx1             = _mm256_add_pd(fjx1,tx);
2476             fjy1             = _mm256_add_pd(fjy1,ty);
2477             fjz1             = _mm256_add_pd(fjz1,tz);
2478
2479             /**************************
2480              * CALCULATE INTERACTIONS *
2481              **************************/
2482
2483             r32              = _mm256_mul_pd(rsq32,rinv32);
2484             r32              = _mm256_andnot_pd(dummy_mask,r32);
2485
2486             /* EWALD ELECTROSTATICS */
2487
2488             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2489             ewrt             = _mm256_mul_pd(r32,ewtabscale);
2490             ewitab           = _mm256_cvttpd_epi32(ewrt);
2491             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2492             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2493                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2494                                             &ewtabF,&ewtabFn);
2495             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2496             felec            = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
2497
2498             fscal            = felec;
2499
2500             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2501
2502             /* Calculate temporary vectorial force */
2503             tx               = _mm256_mul_pd(fscal,dx32);
2504             ty               = _mm256_mul_pd(fscal,dy32);
2505             tz               = _mm256_mul_pd(fscal,dz32);
2506
2507             /* Update vectorial force */
2508             fix3             = _mm256_add_pd(fix3,tx);
2509             fiy3             = _mm256_add_pd(fiy3,ty);
2510             fiz3             = _mm256_add_pd(fiz3,tz);
2511
2512             fjx2             = _mm256_add_pd(fjx2,tx);
2513             fjy2             = _mm256_add_pd(fjy2,ty);
2514             fjz2             = _mm256_add_pd(fjz2,tz);
2515
2516             /**************************
2517              * CALCULATE INTERACTIONS *
2518              **************************/
2519
2520             r33              = _mm256_mul_pd(rsq33,rinv33);
2521             r33              = _mm256_andnot_pd(dummy_mask,r33);
2522
2523             /* EWALD ELECTROSTATICS */
2524
2525             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2526             ewrt             = _mm256_mul_pd(r33,ewtabscale);
2527             ewitab           = _mm256_cvttpd_epi32(ewrt);
2528             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2529             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2530                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2531                                             &ewtabF,&ewtabFn);
2532             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2533             felec            = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
2534
2535             fscal            = felec;
2536
2537             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2538
2539             /* Calculate temporary vectorial force */
2540             tx               = _mm256_mul_pd(fscal,dx33);
2541             ty               = _mm256_mul_pd(fscal,dy33);
2542             tz               = _mm256_mul_pd(fscal,dz33);
2543
2544             /* Update vectorial force */
2545             fix3             = _mm256_add_pd(fix3,tx);
2546             fiy3             = _mm256_add_pd(fiy3,ty);
2547             fiz3             = _mm256_add_pd(fiz3,tz);
2548
2549             fjx3             = _mm256_add_pd(fjx3,tx);
2550             fjy3             = _mm256_add_pd(fjy3,ty);
2551             fjz3             = _mm256_add_pd(fjz3,tz);
2552
2553             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
2554             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
2555             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
2556             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
2557
2558             gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
2559                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
2560                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2561
2562             /* Inner loop uses 385 flops */
2563         }
2564
2565         /* End of innermost loop */
2566
2567         gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
2568                                                  f+i_coord_offset,fshift+i_shift_offset);
2569
2570         /* Increment number of inner iterations */
2571         inneriter                  += j_index_end - j_index_start;
2572
2573         /* Outer loop uses 24 flops */
2574     }
2575
2576     /* Increment number of outer iterations */
2577     outeriter        += nri;
2578
2579     /* Update outer/inner flops */
2580
2581     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*385);
2582 }