src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_avx_256_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_256_single kernel generator.
  37  */
  38 #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 "gromacs/math/vec.h"
  47 #include "nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_256_single.h"
  50 #include "kernelutil_x86_avx_256_single.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_VF_avx_256_single
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            LennardJones
  56  * Geometry:                   Water4-Water4
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_VF_avx_256_single
  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,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight 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              jnrE,jnrF,jnrG,jnrH;
  78     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  79     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  80     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  81     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
  82     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  83     real             rcutoff_scalar;
  84     real             *shiftvec,*fshift,*x,*f;
  85     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
  86     real             scratch[4*DIM];
  87     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  88     real *           vdwioffsetptr0;
  89     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  90     real *           vdwioffsetptr1;
  91     __m256           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  92     real *           vdwioffsetptr2;
  93     __m256           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  94     real *           vdwioffsetptr3;
  95     __m256           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  96     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
  97     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  98     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D,vdwjidx1E,vdwjidx1F,vdwjidx1G,vdwjidx1H;
  99     __m256           jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
 100     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D,vdwjidx2E,vdwjidx2F,vdwjidx2G,vdwjidx2H;
 101     __m256           jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
 102     int              vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D,vdwjidx3E,vdwjidx3F,vdwjidx3G,vdwjidx3H;
 103     __m256           jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
 104     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 105     __m256           dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
 106     __m256           dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
 107     __m256           dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
 108     __m256           dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 109     __m256           dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 110     __m256           dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
 111     __m256           dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
 112     __m256           dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
 113     __m256           dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
 114     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
 115     real             *charge;
 116     int              nvdwtype;
 117     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 118     int              *vdwtype;
 119     real             *vdwparam;
 120     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 121     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 122     __m256i          ewitab;
 123     __m128i          ewitab_lo,ewitab_hi;
 124     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 125     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 126     real             *ewtab;
 127     __m256           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 128     real             rswitch_scalar,d_scalar;
 129     __m256           dummy_mask,cutoff_mask;
 130     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 131     __m256           one     = _mm256_set1_ps(1.0);
 132     __m256           two     = _mm256_set1_ps(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_ps(fr->epsfac);
 145     charge           = mdatoms->chargeA;
 146     nvdwtype         = fr->ntype;
 147     vdwparam         = fr->nbfp;
 148     vdwtype          = mdatoms->typeA;
 149
 150     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
 151     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
 152     beta2            = _mm256_mul_ps(beta,beta);
 153     beta3            = _mm256_mul_ps(beta,beta2);
 154
 155     ewtab            = fr->ic->tabq_coul_FDV0;
 156     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
 157     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
 158
 159     /* Setup water-specific parameters */
 160     inr              = nlist->iinr[0];
 161     iq1              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
 162     iq2              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
 163     iq3              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+3]));
 164     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 165
 166     jq1              = _mm256_set1_ps(charge[inr+1]);
 167     jq2              = _mm256_set1_ps(charge[inr+2]);
 168     jq3              = _mm256_set1_ps(charge[inr+3]);
 169     vdwjidx0A        = 2*vdwtype[inr+0];
 170     c6_00            = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A]);
 171     c12_00           = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A+1]);
 172     qq11             = _mm256_mul_ps(iq1,jq1);
 173     qq12             = _mm256_mul_ps(iq1,jq2);
 174     qq13             = _mm256_mul_ps(iq1,jq3);
 175     qq21             = _mm256_mul_ps(iq2,jq1);
 176     qq22             = _mm256_mul_ps(iq2,jq2);
 177     qq23             = _mm256_mul_ps(iq2,jq3);
 178     qq31             = _mm256_mul_ps(iq3,jq1);
 179     qq32             = _mm256_mul_ps(iq3,jq2);
 180     qq33             = _mm256_mul_ps(iq3,jq3);
 181
 182     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 183     rcutoff_scalar   = fr->rcoulomb;
 184     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
 185     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
 186
 187     rswitch_scalar   = fr->rcoulomb_switch;
 188     rswitch          = _mm256_set1_ps(rswitch_scalar);
 189     /* Setup switch parameters */
 190     d_scalar         = rcutoff_scalar-rswitch_scalar;
 191     d                = _mm256_set1_ps(d_scalar);
 192     swV3             = _mm256_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
 193     swV4             = _mm256_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 194     swV5             = _mm256_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 195     swF2             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
 196     swF3             = _mm256_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 197     swF4             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 198
 199     /* Avoid stupid compiler warnings */
 200     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 201     j_coord_offsetA = 0;
 202     j_coord_offsetB = 0;
 203     j_coord_offsetC = 0;
 204     j_coord_offsetD = 0;
 205     j_coord_offsetE = 0;
 206     j_coord_offsetF = 0;
 207     j_coord_offsetG = 0;
 208     j_coord_offsetH = 0;
 209
 210     outeriter        = 0;
 211     inneriter        = 0;
 212
 213     for(iidx=0;iidx<4*DIM;iidx++)
 214     {
 215         scratch[iidx] = 0.0;
 216     }
 217
 218     /* Start outer loop over neighborlists */
 219     for(iidx=0; iidx<nri; iidx++)
 220     {
 221         /* Load shift vector for this list */
 222         i_shift_offset   = DIM*shiftidx[iidx];
 223
 224         /* Load limits for loop over neighbors */
 225         j_index_start    = jindex[iidx];
 226         j_index_end      = jindex[iidx+1];
 227
 228         /* Get outer coordinate index */
 229         inr              = iinr[iidx];
 230         i_coord_offset   = DIM*inr;
 231
 232         /* Load i particle coords and add shift vector */
 233         gmx_mm256_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
 234                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 235
 236         fix0             = _mm256_setzero_ps();
 237         fiy0             = _mm256_setzero_ps();
 238         fiz0             = _mm256_setzero_ps();
 239         fix1             = _mm256_setzero_ps();
 240         fiy1             = _mm256_setzero_ps();
 241         fiz1             = _mm256_setzero_ps();
 242         fix2             = _mm256_setzero_ps();
 243         fiy2             = _mm256_setzero_ps();
 244         fiz2             = _mm256_setzero_ps();
 245         fix3             = _mm256_setzero_ps();
 246         fiy3             = _mm256_setzero_ps();
 247         fiz3             = _mm256_setzero_ps();
 248
 249         /* Reset potential sums */
 250         velecsum         = _mm256_setzero_ps();
 251         vvdwsum          = _mm256_setzero_ps();
 252
 253         /* Start inner kernel loop */
 254         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 255         {
 256
 257             /* Get j neighbor index, and coordinate index */
 258             jnrA             = jjnr[jidx];
 259             jnrB             = jjnr[jidx+1];
 260             jnrC             = jjnr[jidx+2];
 261             jnrD             = jjnr[jidx+3];
 262             jnrE             = jjnr[jidx+4];
 263             jnrF             = jjnr[jidx+5];
 264             jnrG             = jjnr[jidx+6];
 265             jnrH             = jjnr[jidx+7];
 266             j_coord_offsetA  = DIM*jnrA;
 267             j_coord_offsetB  = DIM*jnrB;
 268             j_coord_offsetC  = DIM*jnrC;
 269             j_coord_offsetD  = DIM*jnrD;
 270             j_coord_offsetE  = DIM*jnrE;
 271             j_coord_offsetF  = DIM*jnrF;
 272             j_coord_offsetG  = DIM*jnrG;
 273             j_coord_offsetH  = DIM*jnrH;
 274
 275             /* load j atom coordinates */
 276             gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 277                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 278                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 279                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 280                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 281                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
 282
 283             /* Calculate displacement vector */
 284             dx00             = _mm256_sub_ps(ix0,jx0);
 285             dy00             = _mm256_sub_ps(iy0,jy0);
 286             dz00             = _mm256_sub_ps(iz0,jz0);
 287             dx11             = _mm256_sub_ps(ix1,jx1);
 288             dy11             = _mm256_sub_ps(iy1,jy1);
 289             dz11             = _mm256_sub_ps(iz1,jz1);
 290             dx12             = _mm256_sub_ps(ix1,jx2);
 291             dy12             = _mm256_sub_ps(iy1,jy2);
 292             dz12             = _mm256_sub_ps(iz1,jz2);
 293             dx13             = _mm256_sub_ps(ix1,jx3);
 294             dy13             = _mm256_sub_ps(iy1,jy3);
 295             dz13             = _mm256_sub_ps(iz1,jz3);
 296             dx21             = _mm256_sub_ps(ix2,jx1);
 297             dy21             = _mm256_sub_ps(iy2,jy1);
 298             dz21             = _mm256_sub_ps(iz2,jz1);
 299             dx22             = _mm256_sub_ps(ix2,jx2);
 300             dy22             = _mm256_sub_ps(iy2,jy2);
 301             dz22             = _mm256_sub_ps(iz2,jz2);
 302             dx23             = _mm256_sub_ps(ix2,jx3);
 303             dy23             = _mm256_sub_ps(iy2,jy3);
 304             dz23             = _mm256_sub_ps(iz2,jz3);
 305             dx31             = _mm256_sub_ps(ix3,jx1);
 306             dy31             = _mm256_sub_ps(iy3,jy1);
 307             dz31             = _mm256_sub_ps(iz3,jz1);
 308             dx32             = _mm256_sub_ps(ix3,jx2);
 309             dy32             = _mm256_sub_ps(iy3,jy2);
 310             dz32             = _mm256_sub_ps(iz3,jz2);
 311             dx33             = _mm256_sub_ps(ix3,jx3);
 312             dy33             = _mm256_sub_ps(iy3,jy3);
 313             dz33             = _mm256_sub_ps(iz3,jz3);
 314
 315             /* Calculate squared distance and things based on it */
 316             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 317             rsq11            = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
 318             rsq12            = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
 319             rsq13            = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
 320             rsq21            = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
 321             rsq22            = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
 322             rsq23            = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
 323             rsq31            = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
 324             rsq32            = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
 325             rsq33            = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
 326
 327             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 328             rinv11           = gmx_mm256_invsqrt_ps(rsq11);
 329             rinv12           = gmx_mm256_invsqrt_ps(rsq12);
 330             rinv13           = gmx_mm256_invsqrt_ps(rsq13);
 331             rinv21           = gmx_mm256_invsqrt_ps(rsq21);
 332             rinv22           = gmx_mm256_invsqrt_ps(rsq22);
 333             rinv23           = gmx_mm256_invsqrt_ps(rsq23);
 334             rinv31           = gmx_mm256_invsqrt_ps(rsq31);
 335             rinv32           = gmx_mm256_invsqrt_ps(rsq32);
 336             rinv33           = gmx_mm256_invsqrt_ps(rsq33);
 337
 338             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 339             rinvsq11         = _mm256_mul_ps(rinv11,rinv11);
 340             rinvsq12         = _mm256_mul_ps(rinv12,rinv12);
 341             rinvsq13         = _mm256_mul_ps(rinv13,rinv13);
 342             rinvsq21         = _mm256_mul_ps(rinv21,rinv21);
 343             rinvsq22         = _mm256_mul_ps(rinv22,rinv22);
 344             rinvsq23         = _mm256_mul_ps(rinv23,rinv23);
 345             rinvsq31         = _mm256_mul_ps(rinv31,rinv31);
 346             rinvsq32         = _mm256_mul_ps(rinv32,rinv32);
 347             rinvsq33         = _mm256_mul_ps(rinv33,rinv33);
 348
 349             fjx0             = _mm256_setzero_ps();
 350             fjy0             = _mm256_setzero_ps();
 351             fjz0             = _mm256_setzero_ps();
 352             fjx1             = _mm256_setzero_ps();
 353             fjy1             = _mm256_setzero_ps();
 354             fjz1             = _mm256_setzero_ps();
 355             fjx2             = _mm256_setzero_ps();
 356             fjy2             = _mm256_setzero_ps();
 357             fjz2             = _mm256_setzero_ps();
 358             fjx3             = _mm256_setzero_ps();
 359             fjy3             = _mm256_setzero_ps();
 360             fjz3             = _mm256_setzero_ps();
 361
 362             /**************************
 363              * CALCULATE INTERACTIONS *
 364              **************************/
 365
 366             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 367             {
 368
 369             r00              = _mm256_mul_ps(rsq00,rinv00);
 370
 371             /* LENNARD-JONES DISPERSION/REPULSION */
 372
 373             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 374             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 375             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 376             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 377             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 378
 379             d                = _mm256_sub_ps(r00,rswitch);
 380             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 381             d2               = _mm256_mul_ps(d,d);
 382             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 383
 384             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 385
 386             /* Evaluate switch function */
 387             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 388             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
 389             vvdw             = _mm256_mul_ps(vvdw,sw);
 390             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 391
 392             /* Update potential sum for this i atom from the interaction with this j atom. */
 393             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
 394             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 395
 396             fscal            = fvdw;
 397
 398             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 399
 400             /* Calculate temporary vectorial force */
 401             tx               = _mm256_mul_ps(fscal,dx00);
 402             ty               = _mm256_mul_ps(fscal,dy00);
 403             tz               = _mm256_mul_ps(fscal,dz00);
 404
 405             /* Update vectorial force */
 406             fix0             = _mm256_add_ps(fix0,tx);
 407             fiy0             = _mm256_add_ps(fiy0,ty);
 408             fiz0             = _mm256_add_ps(fiz0,tz);
 409
 410             fjx0             = _mm256_add_ps(fjx0,tx);
 411             fjy0             = _mm256_add_ps(fjy0,ty);
 412             fjz0             = _mm256_add_ps(fjz0,tz);
 413
 414             }
 415
 416             /**************************
 417              * CALCULATE INTERACTIONS *
 418              **************************/
 419
 420             if (gmx_mm256_any_lt(rsq11,rcutoff2))
 421             {
 422
 423             r11              = _mm256_mul_ps(rsq11,rinv11);
 424
 425             /* EWALD ELECTROSTATICS */
 426
 427             /* Analytical PME correction */
 428             zeta2            = _mm256_mul_ps(beta2,rsq11);
 429             rinv3            = _mm256_mul_ps(rinvsq11,rinv11);
 430             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 431             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 432             felec            = _mm256_mul_ps(qq11,felec);
 433             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 434             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 435             velec            = _mm256_sub_ps(rinv11,pmecorrV);
 436             velec            = _mm256_mul_ps(qq11,velec);
 437
 438             d                = _mm256_sub_ps(r11,rswitch);
 439             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 440             d2               = _mm256_mul_ps(d,d);
 441             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 442
 443             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 444
 445             /* Evaluate switch function */
 446             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 447             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv11,_mm256_mul_ps(velec,dsw)) );
 448             velec            = _mm256_mul_ps(velec,sw);
 449             cutoff_mask      = _mm256_cmp_ps(rsq11,rcutoff2,_CMP_LT_OQ);
 450
 451             /* Update potential sum for this i atom from the interaction with this j atom. */
 452             velec            = _mm256_and_ps(velec,cutoff_mask);
 453             velecsum         = _mm256_add_ps(velecsum,velec);
 454
 455             fscal            = felec;
 456
 457             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 458
 459             /* Calculate temporary vectorial force */
 460             tx               = _mm256_mul_ps(fscal,dx11);
 461             ty               = _mm256_mul_ps(fscal,dy11);
 462             tz               = _mm256_mul_ps(fscal,dz11);
 463
 464             /* Update vectorial force */
 465             fix1             = _mm256_add_ps(fix1,tx);
 466             fiy1             = _mm256_add_ps(fiy1,ty);
 467             fiz1             = _mm256_add_ps(fiz1,tz);
 468
 469             fjx1             = _mm256_add_ps(fjx1,tx);
 470             fjy1             = _mm256_add_ps(fjy1,ty);
 471             fjz1             = _mm256_add_ps(fjz1,tz);
 472
 473             }
 474
 475             /**************************
 476              * CALCULATE INTERACTIONS *
 477              **************************/
 478
 479             if (gmx_mm256_any_lt(rsq12,rcutoff2))
 480             {
 481
 482             r12              = _mm256_mul_ps(rsq12,rinv12);
 483
 484             /* EWALD ELECTROSTATICS */
 485
 486             /* Analytical PME correction */
 487             zeta2            = _mm256_mul_ps(beta2,rsq12);
 488             rinv3            = _mm256_mul_ps(rinvsq12,rinv12);
 489             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 490             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 491             felec            = _mm256_mul_ps(qq12,felec);
 492             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 493             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 494             velec            = _mm256_sub_ps(rinv12,pmecorrV);
 495             velec            = _mm256_mul_ps(qq12,velec);
 496
 497             d                = _mm256_sub_ps(r12,rswitch);
 498             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 499             d2               = _mm256_mul_ps(d,d);
 500             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 501
 502             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 503
 504             /* Evaluate switch function */
 505             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 506             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv12,_mm256_mul_ps(velec,dsw)) );
 507             velec            = _mm256_mul_ps(velec,sw);
 508             cutoff_mask      = _mm256_cmp_ps(rsq12,rcutoff2,_CMP_LT_OQ);
 509
 510             /* Update potential sum for this i atom from the interaction with this j atom. */
 511             velec            = _mm256_and_ps(velec,cutoff_mask);
 512             velecsum         = _mm256_add_ps(velecsum,velec);
 513
 514             fscal            = felec;
 515
 516             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 517
 518             /* Calculate temporary vectorial force */
 519             tx               = _mm256_mul_ps(fscal,dx12);
 520             ty               = _mm256_mul_ps(fscal,dy12);
 521             tz               = _mm256_mul_ps(fscal,dz12);
 522
 523             /* Update vectorial force */
 524             fix1             = _mm256_add_ps(fix1,tx);
 525             fiy1             = _mm256_add_ps(fiy1,ty);
 526             fiz1             = _mm256_add_ps(fiz1,tz);
 527
 528             fjx2             = _mm256_add_ps(fjx2,tx);
 529             fjy2             = _mm256_add_ps(fjy2,ty);
 530             fjz2             = _mm256_add_ps(fjz2,tz);
 531
 532             }
 533
 534             /**************************
 535              * CALCULATE INTERACTIONS *
 536              **************************/
 537
 538             if (gmx_mm256_any_lt(rsq13,rcutoff2))
 539             {
 540
 541             r13              = _mm256_mul_ps(rsq13,rinv13);
 542
 543             /* EWALD ELECTROSTATICS */
 544
 545             /* Analytical PME correction */
 546             zeta2            = _mm256_mul_ps(beta2,rsq13);
 547             rinv3            = _mm256_mul_ps(rinvsq13,rinv13);
 548             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 549             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 550             felec            = _mm256_mul_ps(qq13,felec);
 551             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 552             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 553             velec            = _mm256_sub_ps(rinv13,pmecorrV);
 554             velec            = _mm256_mul_ps(qq13,velec);
 555
 556             d                = _mm256_sub_ps(r13,rswitch);
 557             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 558             d2               = _mm256_mul_ps(d,d);
 559             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 560
 561             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 562
 563             /* Evaluate switch function */
 564             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 565             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv13,_mm256_mul_ps(velec,dsw)) );
 566             velec            = _mm256_mul_ps(velec,sw);
 567             cutoff_mask      = _mm256_cmp_ps(rsq13,rcutoff2,_CMP_LT_OQ);
 568
 569             /* Update potential sum for this i atom from the interaction with this j atom. */
 570             velec            = _mm256_and_ps(velec,cutoff_mask);
 571             velecsum         = _mm256_add_ps(velecsum,velec);
 572
 573             fscal            = felec;
 574
 575             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 576
 577             /* Calculate temporary vectorial force */
 578             tx               = _mm256_mul_ps(fscal,dx13);
 579             ty               = _mm256_mul_ps(fscal,dy13);
 580             tz               = _mm256_mul_ps(fscal,dz13);
 581
 582             /* Update vectorial force */
 583             fix1             = _mm256_add_ps(fix1,tx);
 584             fiy1             = _mm256_add_ps(fiy1,ty);
 585             fiz1             = _mm256_add_ps(fiz1,tz);
 586
 587             fjx3             = _mm256_add_ps(fjx3,tx);
 588             fjy3             = _mm256_add_ps(fjy3,ty);
 589             fjz3             = _mm256_add_ps(fjz3,tz);
 590
 591             }
 592
 593             /**************************
 594              * CALCULATE INTERACTIONS *
 595              **************************/
 596
 597             if (gmx_mm256_any_lt(rsq21,rcutoff2))
 598             {
 599
 600             r21              = _mm256_mul_ps(rsq21,rinv21);
 601
 602             /* EWALD ELECTROSTATICS */
 603
 604             /* Analytical PME correction */
 605             zeta2            = _mm256_mul_ps(beta2,rsq21);
 606             rinv3            = _mm256_mul_ps(rinvsq21,rinv21);
 607             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 608             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 609             felec            = _mm256_mul_ps(qq21,felec);
 610             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 611             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 612             velec            = _mm256_sub_ps(rinv21,pmecorrV);
 613             velec            = _mm256_mul_ps(qq21,velec);
 614
 615             d                = _mm256_sub_ps(r21,rswitch);
 616             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 617             d2               = _mm256_mul_ps(d,d);
 618             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 619
 620             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 621
 622             /* Evaluate switch function */
 623             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 624             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv21,_mm256_mul_ps(velec,dsw)) );
 625             velec            = _mm256_mul_ps(velec,sw);
 626             cutoff_mask      = _mm256_cmp_ps(rsq21,rcutoff2,_CMP_LT_OQ);
 627
 628             /* Update potential sum for this i atom from the interaction with this j atom. */
 629             velec            = _mm256_and_ps(velec,cutoff_mask);
 630             velecsum         = _mm256_add_ps(velecsum,velec);
 631
 632             fscal            = felec;
 633
 634             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 635
 636             /* Calculate temporary vectorial force */
 637             tx               = _mm256_mul_ps(fscal,dx21);
 638             ty               = _mm256_mul_ps(fscal,dy21);
 639             tz               = _mm256_mul_ps(fscal,dz21);
 640
 641             /* Update vectorial force */
 642             fix2             = _mm256_add_ps(fix2,tx);
 643             fiy2             = _mm256_add_ps(fiy2,ty);
 644             fiz2             = _mm256_add_ps(fiz2,tz);
 645
 646             fjx1             = _mm256_add_ps(fjx1,tx);
 647             fjy1             = _mm256_add_ps(fjy1,ty);
 648             fjz1             = _mm256_add_ps(fjz1,tz);
 649
 650             }
 651
 652             /**************************
 653              * CALCULATE INTERACTIONS *
 654              **************************/
 655
 656             if (gmx_mm256_any_lt(rsq22,rcutoff2))
 657             {
 658
 659             r22              = _mm256_mul_ps(rsq22,rinv22);
 660
 661             /* EWALD ELECTROSTATICS */
 662
 663             /* Analytical PME correction */
 664             zeta2            = _mm256_mul_ps(beta2,rsq22);
 665             rinv3            = _mm256_mul_ps(rinvsq22,rinv22);
 666             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 667             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 668             felec            = _mm256_mul_ps(qq22,felec);
 669             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 670             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 671             velec            = _mm256_sub_ps(rinv22,pmecorrV);
 672             velec            = _mm256_mul_ps(qq22,velec);
 673
 674             d                = _mm256_sub_ps(r22,rswitch);
 675             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 676             d2               = _mm256_mul_ps(d,d);
 677             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 678
 679             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 680
 681             /* Evaluate switch function */
 682             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 683             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv22,_mm256_mul_ps(velec,dsw)) );
 684             velec            = _mm256_mul_ps(velec,sw);
 685             cutoff_mask      = _mm256_cmp_ps(rsq22,rcutoff2,_CMP_LT_OQ);
 686
 687             /* Update potential sum for this i atom from the interaction with this j atom. */
 688             velec            = _mm256_and_ps(velec,cutoff_mask);
 689             velecsum         = _mm256_add_ps(velecsum,velec);
 690
 691             fscal            = felec;
 692
 693             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 694
 695             /* Calculate temporary vectorial force */
 696             tx               = _mm256_mul_ps(fscal,dx22);
 697             ty               = _mm256_mul_ps(fscal,dy22);
 698             tz               = _mm256_mul_ps(fscal,dz22);
 699
 700             /* Update vectorial force */
 701             fix2             = _mm256_add_ps(fix2,tx);
 702             fiy2             = _mm256_add_ps(fiy2,ty);
 703             fiz2             = _mm256_add_ps(fiz2,tz);
 704
 705             fjx2             = _mm256_add_ps(fjx2,tx);
 706             fjy2             = _mm256_add_ps(fjy2,ty);
 707             fjz2             = _mm256_add_ps(fjz2,tz);
 708
 709             }
 710
 711             /**************************
 712              * CALCULATE INTERACTIONS *
 713              **************************/
 714
 715             if (gmx_mm256_any_lt(rsq23,rcutoff2))
 716             {
 717
 718             r23              = _mm256_mul_ps(rsq23,rinv23);
 719
 720             /* EWALD ELECTROSTATICS */
 721
 722             /* Analytical PME correction */
 723             zeta2            = _mm256_mul_ps(beta2,rsq23);
 724             rinv3            = _mm256_mul_ps(rinvsq23,rinv23);
 725             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 726             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 727             felec            = _mm256_mul_ps(qq23,felec);
 728             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 729             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 730             velec            = _mm256_sub_ps(rinv23,pmecorrV);
 731             velec            = _mm256_mul_ps(qq23,velec);
 732
 733             d                = _mm256_sub_ps(r23,rswitch);
 734             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 735             d2               = _mm256_mul_ps(d,d);
 736             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 737
 738             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 739
 740             /* Evaluate switch function */
 741             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 742             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv23,_mm256_mul_ps(velec,dsw)) );
 743             velec            = _mm256_mul_ps(velec,sw);
 744             cutoff_mask      = _mm256_cmp_ps(rsq23,rcutoff2,_CMP_LT_OQ);
 745
 746             /* Update potential sum for this i atom from the interaction with this j atom. */
 747             velec            = _mm256_and_ps(velec,cutoff_mask);
 748             velecsum         = _mm256_add_ps(velecsum,velec);
 749
 750             fscal            = felec;
 751
 752             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 753
 754             /* Calculate temporary vectorial force */
 755             tx               = _mm256_mul_ps(fscal,dx23);
 756             ty               = _mm256_mul_ps(fscal,dy23);
 757             tz               = _mm256_mul_ps(fscal,dz23);
 758
 759             /* Update vectorial force */
 760             fix2             = _mm256_add_ps(fix2,tx);
 761             fiy2             = _mm256_add_ps(fiy2,ty);
 762             fiz2             = _mm256_add_ps(fiz2,tz);
 763
 764             fjx3             = _mm256_add_ps(fjx3,tx);
 765             fjy3             = _mm256_add_ps(fjy3,ty);
 766             fjz3             = _mm256_add_ps(fjz3,tz);
 767
 768             }
 769
 770             /**************************
 771              * CALCULATE INTERACTIONS *
 772              **************************/
 773
 774             if (gmx_mm256_any_lt(rsq31,rcutoff2))
 775             {
 776
 777             r31              = _mm256_mul_ps(rsq31,rinv31);
 778
 779             /* EWALD ELECTROSTATICS */
 780
 781             /* Analytical PME correction */
 782             zeta2            = _mm256_mul_ps(beta2,rsq31);
 783             rinv3            = _mm256_mul_ps(rinvsq31,rinv31);
 784             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 785             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 786             felec            = _mm256_mul_ps(qq31,felec);
 787             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 788             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 789             velec            = _mm256_sub_ps(rinv31,pmecorrV);
 790             velec            = _mm256_mul_ps(qq31,velec);
 791
 792             d                = _mm256_sub_ps(r31,rswitch);
 793             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 794             d2               = _mm256_mul_ps(d,d);
 795             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 796
 797             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 798
 799             /* Evaluate switch function */
 800             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 801             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv31,_mm256_mul_ps(velec,dsw)) );
 802             velec            = _mm256_mul_ps(velec,sw);
 803             cutoff_mask      = _mm256_cmp_ps(rsq31,rcutoff2,_CMP_LT_OQ);
 804
 805             /* Update potential sum for this i atom from the interaction with this j atom. */
 806             velec            = _mm256_and_ps(velec,cutoff_mask);
 807             velecsum         = _mm256_add_ps(velecsum,velec);
 808
 809             fscal            = felec;
 810
 811             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 812
 813             /* Calculate temporary vectorial force */
 814             tx               = _mm256_mul_ps(fscal,dx31);
 815             ty               = _mm256_mul_ps(fscal,dy31);
 816             tz               = _mm256_mul_ps(fscal,dz31);
 817
 818             /* Update vectorial force */
 819             fix3             = _mm256_add_ps(fix3,tx);
 820             fiy3             = _mm256_add_ps(fiy3,ty);
 821             fiz3             = _mm256_add_ps(fiz3,tz);
 822
 823             fjx1             = _mm256_add_ps(fjx1,tx);
 824             fjy1             = _mm256_add_ps(fjy1,ty);
 825             fjz1             = _mm256_add_ps(fjz1,tz);
 826
 827             }
 828
 829             /**************************
 830              * CALCULATE INTERACTIONS *
 831              **************************/
 832
 833             if (gmx_mm256_any_lt(rsq32,rcutoff2))
 834             {
 835
 836             r32              = _mm256_mul_ps(rsq32,rinv32);
 837
 838             /* EWALD ELECTROSTATICS */
 839
 840             /* Analytical PME correction */
 841             zeta2            = _mm256_mul_ps(beta2,rsq32);
 842             rinv3            = _mm256_mul_ps(rinvsq32,rinv32);
 843             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 844             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 845             felec            = _mm256_mul_ps(qq32,felec);
 846             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 847             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 848             velec            = _mm256_sub_ps(rinv32,pmecorrV);
 849             velec            = _mm256_mul_ps(qq32,velec);
 850
 851             d                = _mm256_sub_ps(r32,rswitch);
 852             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 853             d2               = _mm256_mul_ps(d,d);
 854             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 855
 856             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 857
 858             /* Evaluate switch function */
 859             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 860             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv32,_mm256_mul_ps(velec,dsw)) );
 861             velec            = _mm256_mul_ps(velec,sw);
 862             cutoff_mask      = _mm256_cmp_ps(rsq32,rcutoff2,_CMP_LT_OQ);
 863
 864             /* Update potential sum for this i atom from the interaction with this j atom. */
 865             velec            = _mm256_and_ps(velec,cutoff_mask);
 866             velecsum         = _mm256_add_ps(velecsum,velec);
 867
 868             fscal            = felec;
 869
 870             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 871
 872             /* Calculate temporary vectorial force */
 873             tx               = _mm256_mul_ps(fscal,dx32);
 874             ty               = _mm256_mul_ps(fscal,dy32);
 875             tz               = _mm256_mul_ps(fscal,dz32);
 876
 877             /* Update vectorial force */
 878             fix3             = _mm256_add_ps(fix3,tx);
 879             fiy3             = _mm256_add_ps(fiy3,ty);
 880             fiz3             = _mm256_add_ps(fiz3,tz);
 881
 882             fjx2             = _mm256_add_ps(fjx2,tx);
 883             fjy2             = _mm256_add_ps(fjy2,ty);
 884             fjz2             = _mm256_add_ps(fjz2,tz);
 885
 886             }
 887
 888             /**************************
 889              * CALCULATE INTERACTIONS *
 890              **************************/
 891
 892             if (gmx_mm256_any_lt(rsq33,rcutoff2))
 893             {
 894
 895             r33              = _mm256_mul_ps(rsq33,rinv33);
 896
 897             /* EWALD ELECTROSTATICS */
 898
 899             /* Analytical PME correction */
 900             zeta2            = _mm256_mul_ps(beta2,rsq33);
 901             rinv3            = _mm256_mul_ps(rinvsq33,rinv33);
 902             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 903             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 904             felec            = _mm256_mul_ps(qq33,felec);
 905             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 906             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 907             velec            = _mm256_sub_ps(rinv33,pmecorrV);
 908             velec            = _mm256_mul_ps(qq33,velec);
 909
 910             d                = _mm256_sub_ps(r33,rswitch);
 911             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 912             d2               = _mm256_mul_ps(d,d);
 913             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
 914
 915             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 916
 917             /* Evaluate switch function */
 918             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 919             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv33,_mm256_mul_ps(velec,dsw)) );
 920             velec            = _mm256_mul_ps(velec,sw);
 921             cutoff_mask      = _mm256_cmp_ps(rsq33,rcutoff2,_CMP_LT_OQ);
 922
 923             /* Update potential sum for this i atom from the interaction with this j atom. */
 924             velec            = _mm256_and_ps(velec,cutoff_mask);
 925             velecsum         = _mm256_add_ps(velecsum,velec);
 926
 927             fscal            = felec;
 928
 929             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 930
 931             /* Calculate temporary vectorial force */
 932             tx               = _mm256_mul_ps(fscal,dx33);
 933             ty               = _mm256_mul_ps(fscal,dy33);
 934             tz               = _mm256_mul_ps(fscal,dz33);
 935
 936             /* Update vectorial force */
 937             fix3             = _mm256_add_ps(fix3,tx);
 938             fiy3             = _mm256_add_ps(fiy3,ty);
 939             fiz3             = _mm256_add_ps(fiz3,tz);
 940
 941             fjx3             = _mm256_add_ps(fjx3,tx);
 942             fjy3             = _mm256_add_ps(fjy3,ty);
 943             fjz3             = _mm256_add_ps(fjz3,tz);
 944
 945             }
 946
 947             fjptrA             = f+j_coord_offsetA;
 948             fjptrB             = f+j_coord_offsetB;
 949             fjptrC             = f+j_coord_offsetC;
 950             fjptrD             = f+j_coord_offsetD;
 951             fjptrE             = f+j_coord_offsetE;
 952             fjptrF             = f+j_coord_offsetF;
 953             fjptrG             = f+j_coord_offsetG;
 954             fjptrH             = f+j_coord_offsetH;
 955
 956             gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
 957                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
 958                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 959
 960             /* Inner loop uses 1034 flops */
 961         }
 962
 963         if(jidx<j_index_end)
 964         {
 965
 966             /* Get j neighbor index, and coordinate index */
 967             jnrlistA         = jjnr[jidx];
 968             jnrlistB         = jjnr[jidx+1];
 969             jnrlistC         = jjnr[jidx+2];
 970             jnrlistD         = jjnr[jidx+3];
 971             jnrlistE         = jjnr[jidx+4];
 972             jnrlistF         = jjnr[jidx+5];
 973             jnrlistG         = jjnr[jidx+6];
 974             jnrlistH         = jjnr[jidx+7];
 975             /* Sign of each element will be negative for non-real atoms.
 976              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 977              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 978              */
 979             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 980                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 981
 982             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 983             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 984             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 985             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 986             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 987             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 988             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 989             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 990             j_coord_offsetA  = DIM*jnrA;
 991             j_coord_offsetB  = DIM*jnrB;
 992             j_coord_offsetC  = DIM*jnrC;
 993             j_coord_offsetD  = DIM*jnrD;
 994             j_coord_offsetE  = DIM*jnrE;
 995             j_coord_offsetF  = DIM*jnrF;
 996             j_coord_offsetG  = DIM*jnrG;
 997             j_coord_offsetH  = DIM*jnrH;
 998
 999             /* load j atom coordinates */
1000             gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1001                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1002                                                  x+j_coord_offsetE,x+j_coord_offsetF,
1003                                                  x+j_coord_offsetG,x+j_coord_offsetH,
1004                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1005                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
1006
1007             /* Calculate displacement vector */
1008             dx00             = _mm256_sub_ps(ix0,jx0);
1009             dy00             = _mm256_sub_ps(iy0,jy0);
1010             dz00             = _mm256_sub_ps(iz0,jz0);
1011             dx11             = _mm256_sub_ps(ix1,jx1);
1012             dy11             = _mm256_sub_ps(iy1,jy1);
1013             dz11             = _mm256_sub_ps(iz1,jz1);
1014             dx12             = _mm256_sub_ps(ix1,jx2);
1015             dy12             = _mm256_sub_ps(iy1,jy2);
1016             dz12             = _mm256_sub_ps(iz1,jz2);
1017             dx13             = _mm256_sub_ps(ix1,jx3);
1018             dy13             = _mm256_sub_ps(iy1,jy3);
1019             dz13             = _mm256_sub_ps(iz1,jz3);
1020             dx21             = _mm256_sub_ps(ix2,jx1);
1021             dy21             = _mm256_sub_ps(iy2,jy1);
1022             dz21             = _mm256_sub_ps(iz2,jz1);
1023             dx22             = _mm256_sub_ps(ix2,jx2);
1024             dy22             = _mm256_sub_ps(iy2,jy2);
1025             dz22             = _mm256_sub_ps(iz2,jz2);
1026             dx23             = _mm256_sub_ps(ix2,jx3);
1027             dy23             = _mm256_sub_ps(iy2,jy3);
1028             dz23             = _mm256_sub_ps(iz2,jz3);
1029             dx31             = _mm256_sub_ps(ix3,jx1);
1030             dy31             = _mm256_sub_ps(iy3,jy1);
1031             dz31             = _mm256_sub_ps(iz3,jz1);
1032             dx32             = _mm256_sub_ps(ix3,jx2);
1033             dy32             = _mm256_sub_ps(iy3,jy2);
1034             dz32             = _mm256_sub_ps(iz3,jz2);
1035             dx33             = _mm256_sub_ps(ix3,jx3);
1036             dy33             = _mm256_sub_ps(iy3,jy3);
1037             dz33             = _mm256_sub_ps(iz3,jz3);
1038
1039             /* Calculate squared distance and things based on it */
1040             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
1041             rsq11            = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
1042             rsq12            = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
1043             rsq13            = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
1044             rsq21            = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
1045             rsq22            = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
1046             rsq23            = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
1047             rsq31            = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
1048             rsq32            = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
1049             rsq33            = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
1050
1051             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
1052             rinv11           = gmx_mm256_invsqrt_ps(rsq11);
1053             rinv12           = gmx_mm256_invsqrt_ps(rsq12);
1054             rinv13           = gmx_mm256_invsqrt_ps(rsq13);
1055             rinv21           = gmx_mm256_invsqrt_ps(rsq21);
1056             rinv22           = gmx_mm256_invsqrt_ps(rsq22);
1057             rinv23           = gmx_mm256_invsqrt_ps(rsq23);
1058             rinv31           = gmx_mm256_invsqrt_ps(rsq31);
1059             rinv32           = gmx_mm256_invsqrt_ps(rsq32);
1060             rinv33           = gmx_mm256_invsqrt_ps(rsq33);
1061
1062             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
1063             rinvsq11         = _mm256_mul_ps(rinv11,rinv11);
1064             rinvsq12         = _mm256_mul_ps(rinv12,rinv12);
1065             rinvsq13         = _mm256_mul_ps(rinv13,rinv13);
1066             rinvsq21         = _mm256_mul_ps(rinv21,rinv21);
1067             rinvsq22         = _mm256_mul_ps(rinv22,rinv22);
1068             rinvsq23         = _mm256_mul_ps(rinv23,rinv23);
1069             rinvsq31         = _mm256_mul_ps(rinv31,rinv31);
1070             rinvsq32         = _mm256_mul_ps(rinv32,rinv32);
1071             rinvsq33         = _mm256_mul_ps(rinv33,rinv33);
1072
1073             fjx0             = _mm256_setzero_ps();
1074             fjy0             = _mm256_setzero_ps();
1075             fjz0             = _mm256_setzero_ps();
1076             fjx1             = _mm256_setzero_ps();
1077             fjy1             = _mm256_setzero_ps();
1078             fjz1             = _mm256_setzero_ps();
1079             fjx2             = _mm256_setzero_ps();
1080             fjy2             = _mm256_setzero_ps();
1081             fjz2             = _mm256_setzero_ps();
1082             fjx3             = _mm256_setzero_ps();
1083             fjy3             = _mm256_setzero_ps();
1084             fjz3             = _mm256_setzero_ps();
1085
1086             /**************************
1087              * CALCULATE INTERACTIONS *
1088              **************************/
1089
1090             if (gmx_mm256_any_lt(rsq00,rcutoff2))
1091             {
1092
1093             r00              = _mm256_mul_ps(rsq00,rinv00);
1094             r00              = _mm256_andnot_ps(dummy_mask,r00);
1095
1096             /* LENNARD-JONES DISPERSION/REPULSION */
1097
1098             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1099             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
1100             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
1101             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
1102             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
1103
1104             d                = _mm256_sub_ps(r00,rswitch);
1105             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1106             d2               = _mm256_mul_ps(d,d);
1107             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1108
1109             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1110
1111             /* Evaluate switch function */
1112             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1113             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
1114             vvdw             = _mm256_mul_ps(vvdw,sw);
1115             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
1116
1117             /* Update potential sum for this i atom from the interaction with this j atom. */
1118             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
1119             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
1120             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
1121
1122             fscal            = fvdw;
1123
1124             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1125
1126             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1127
1128             /* Calculate temporary vectorial force */
1129             tx               = _mm256_mul_ps(fscal,dx00);
1130             ty               = _mm256_mul_ps(fscal,dy00);
1131             tz               = _mm256_mul_ps(fscal,dz00);
1132
1133             /* Update vectorial force */
1134             fix0             = _mm256_add_ps(fix0,tx);
1135             fiy0             = _mm256_add_ps(fiy0,ty);
1136             fiz0             = _mm256_add_ps(fiz0,tz);
1137
1138             fjx0             = _mm256_add_ps(fjx0,tx);
1139             fjy0             = _mm256_add_ps(fjy0,ty);
1140             fjz0             = _mm256_add_ps(fjz0,tz);
1141
1142             }
1143
1144             /**************************
1145              * CALCULATE INTERACTIONS *
1146              **************************/
1147
1148             if (gmx_mm256_any_lt(rsq11,rcutoff2))
1149             {
1150
1151             r11              = _mm256_mul_ps(rsq11,rinv11);
1152             r11              = _mm256_andnot_ps(dummy_mask,r11);
1153
1154             /* EWALD ELECTROSTATICS */
1155
1156             /* Analytical PME correction */
1157             zeta2            = _mm256_mul_ps(beta2,rsq11);
1158             rinv3            = _mm256_mul_ps(rinvsq11,rinv11);
1159             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1160             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1161             felec            = _mm256_mul_ps(qq11,felec);
1162             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1163             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1164             velec            = _mm256_sub_ps(rinv11,pmecorrV);
1165             velec            = _mm256_mul_ps(qq11,velec);
1166
1167             d                = _mm256_sub_ps(r11,rswitch);
1168             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1169             d2               = _mm256_mul_ps(d,d);
1170             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1171
1172             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1173
1174             /* Evaluate switch function */
1175             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1176             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv11,_mm256_mul_ps(velec,dsw)) );
1177             velec            = _mm256_mul_ps(velec,sw);
1178             cutoff_mask      = _mm256_cmp_ps(rsq11,rcutoff2,_CMP_LT_OQ);
1179
1180             /* Update potential sum for this i atom from the interaction with this j atom. */
1181             velec            = _mm256_and_ps(velec,cutoff_mask);
1182             velec            = _mm256_andnot_ps(dummy_mask,velec);
1183             velecsum         = _mm256_add_ps(velecsum,velec);
1184
1185             fscal            = felec;
1186
1187             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1188
1189             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1190
1191             /* Calculate temporary vectorial force */
1192             tx               = _mm256_mul_ps(fscal,dx11);
1193             ty               = _mm256_mul_ps(fscal,dy11);
1194             tz               = _mm256_mul_ps(fscal,dz11);
1195
1196             /* Update vectorial force */
1197             fix1             = _mm256_add_ps(fix1,tx);
1198             fiy1             = _mm256_add_ps(fiy1,ty);
1199             fiz1             = _mm256_add_ps(fiz1,tz);
1200
1201             fjx1             = _mm256_add_ps(fjx1,tx);
1202             fjy1             = _mm256_add_ps(fjy1,ty);
1203             fjz1             = _mm256_add_ps(fjz1,tz);
1204
1205             }
1206
1207             /**************************
1208              * CALCULATE INTERACTIONS *
1209              **************************/
1210
1211             if (gmx_mm256_any_lt(rsq12,rcutoff2))
1212             {
1213
1214             r12              = _mm256_mul_ps(rsq12,rinv12);
1215             r12              = _mm256_andnot_ps(dummy_mask,r12);
1216
1217             /* EWALD ELECTROSTATICS */
1218
1219             /* Analytical PME correction */
1220             zeta2            = _mm256_mul_ps(beta2,rsq12);
1221             rinv3            = _mm256_mul_ps(rinvsq12,rinv12);
1222             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1223             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1224             felec            = _mm256_mul_ps(qq12,felec);
1225             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1226             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1227             velec            = _mm256_sub_ps(rinv12,pmecorrV);
1228             velec            = _mm256_mul_ps(qq12,velec);
1229
1230             d                = _mm256_sub_ps(r12,rswitch);
1231             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1232             d2               = _mm256_mul_ps(d,d);
1233             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1234
1235             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1236
1237             /* Evaluate switch function */
1238             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1239             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv12,_mm256_mul_ps(velec,dsw)) );
1240             velec            = _mm256_mul_ps(velec,sw);
1241             cutoff_mask      = _mm256_cmp_ps(rsq12,rcutoff2,_CMP_LT_OQ);
1242
1243             /* Update potential sum for this i atom from the interaction with this j atom. */
1244             velec            = _mm256_and_ps(velec,cutoff_mask);
1245             velec            = _mm256_andnot_ps(dummy_mask,velec);
1246             velecsum         = _mm256_add_ps(velecsum,velec);
1247
1248             fscal            = felec;
1249
1250             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1251
1252             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1253
1254             /* Calculate temporary vectorial force */
1255             tx               = _mm256_mul_ps(fscal,dx12);
1256             ty               = _mm256_mul_ps(fscal,dy12);
1257             tz               = _mm256_mul_ps(fscal,dz12);
1258
1259             /* Update vectorial force */
1260             fix1             = _mm256_add_ps(fix1,tx);
1261             fiy1             = _mm256_add_ps(fiy1,ty);
1262             fiz1             = _mm256_add_ps(fiz1,tz);
1263
1264             fjx2             = _mm256_add_ps(fjx2,tx);
1265             fjy2             = _mm256_add_ps(fjy2,ty);
1266             fjz2             = _mm256_add_ps(fjz2,tz);
1267
1268             }
1269
1270             /**************************
1271              * CALCULATE INTERACTIONS *
1272              **************************/
1273
1274             if (gmx_mm256_any_lt(rsq13,rcutoff2))
1275             {
1276
1277             r13              = _mm256_mul_ps(rsq13,rinv13);
1278             r13              = _mm256_andnot_ps(dummy_mask,r13);
1279
1280             /* EWALD ELECTROSTATICS */
1281
1282             /* Analytical PME correction */
1283             zeta2            = _mm256_mul_ps(beta2,rsq13);
1284             rinv3            = _mm256_mul_ps(rinvsq13,rinv13);
1285             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1286             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1287             felec            = _mm256_mul_ps(qq13,felec);
1288             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1289             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1290             velec            = _mm256_sub_ps(rinv13,pmecorrV);
1291             velec            = _mm256_mul_ps(qq13,velec);
1292
1293             d                = _mm256_sub_ps(r13,rswitch);
1294             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1295             d2               = _mm256_mul_ps(d,d);
1296             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1297
1298             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1299
1300             /* Evaluate switch function */
1301             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1302             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv13,_mm256_mul_ps(velec,dsw)) );
1303             velec            = _mm256_mul_ps(velec,sw);
1304             cutoff_mask      = _mm256_cmp_ps(rsq13,rcutoff2,_CMP_LT_OQ);
1305
1306             /* Update potential sum for this i atom from the interaction with this j atom. */
1307             velec            = _mm256_and_ps(velec,cutoff_mask);
1308             velec            = _mm256_andnot_ps(dummy_mask,velec);
1309             velecsum         = _mm256_add_ps(velecsum,velec);
1310
1311             fscal            = felec;
1312
1313             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1314
1315             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1316
1317             /* Calculate temporary vectorial force */
1318             tx               = _mm256_mul_ps(fscal,dx13);
1319             ty               = _mm256_mul_ps(fscal,dy13);
1320             tz               = _mm256_mul_ps(fscal,dz13);
1321
1322             /* Update vectorial force */
1323             fix1             = _mm256_add_ps(fix1,tx);
1324             fiy1             = _mm256_add_ps(fiy1,ty);
1325             fiz1             = _mm256_add_ps(fiz1,tz);
1326
1327             fjx3             = _mm256_add_ps(fjx3,tx);
1328             fjy3             = _mm256_add_ps(fjy3,ty);
1329             fjz3             = _mm256_add_ps(fjz3,tz);
1330
1331             }
1332
1333             /**************************
1334              * CALCULATE INTERACTIONS *
1335              **************************/
1336
1337             if (gmx_mm256_any_lt(rsq21,rcutoff2))
1338             {
1339
1340             r21              = _mm256_mul_ps(rsq21,rinv21);
1341             r21              = _mm256_andnot_ps(dummy_mask,r21);
1342
1343             /* EWALD ELECTROSTATICS */
1344
1345             /* Analytical PME correction */
1346             zeta2            = _mm256_mul_ps(beta2,rsq21);
1347             rinv3            = _mm256_mul_ps(rinvsq21,rinv21);
1348             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1349             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1350             felec            = _mm256_mul_ps(qq21,felec);
1351             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1352             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1353             velec            = _mm256_sub_ps(rinv21,pmecorrV);
1354             velec            = _mm256_mul_ps(qq21,velec);
1355
1356             d                = _mm256_sub_ps(r21,rswitch);
1357             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1358             d2               = _mm256_mul_ps(d,d);
1359             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1360
1361             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1362
1363             /* Evaluate switch function */
1364             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1365             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv21,_mm256_mul_ps(velec,dsw)) );
1366             velec            = _mm256_mul_ps(velec,sw);
1367             cutoff_mask      = _mm256_cmp_ps(rsq21,rcutoff2,_CMP_LT_OQ);
1368
1369             /* Update potential sum for this i atom from the interaction with this j atom. */
1370             velec            = _mm256_and_ps(velec,cutoff_mask);
1371             velec            = _mm256_andnot_ps(dummy_mask,velec);
1372             velecsum         = _mm256_add_ps(velecsum,velec);
1373
1374             fscal            = felec;
1375
1376             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1377
1378             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1379
1380             /* Calculate temporary vectorial force */
1381             tx               = _mm256_mul_ps(fscal,dx21);
1382             ty               = _mm256_mul_ps(fscal,dy21);
1383             tz               = _mm256_mul_ps(fscal,dz21);
1384
1385             /* Update vectorial force */
1386             fix2             = _mm256_add_ps(fix2,tx);
1387             fiy2             = _mm256_add_ps(fiy2,ty);
1388             fiz2             = _mm256_add_ps(fiz2,tz);
1389
1390             fjx1             = _mm256_add_ps(fjx1,tx);
1391             fjy1             = _mm256_add_ps(fjy1,ty);
1392             fjz1             = _mm256_add_ps(fjz1,tz);
1393
1394             }
1395
1396             /**************************
1397              * CALCULATE INTERACTIONS *
1398              **************************/
1399
1400             if (gmx_mm256_any_lt(rsq22,rcutoff2))
1401             {
1402
1403             r22              = _mm256_mul_ps(rsq22,rinv22);
1404             r22              = _mm256_andnot_ps(dummy_mask,r22);
1405
1406             /* EWALD ELECTROSTATICS */
1407
1408             /* Analytical PME correction */
1409             zeta2            = _mm256_mul_ps(beta2,rsq22);
1410             rinv3            = _mm256_mul_ps(rinvsq22,rinv22);
1411             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1412             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1413             felec            = _mm256_mul_ps(qq22,felec);
1414             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1415             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1416             velec            = _mm256_sub_ps(rinv22,pmecorrV);
1417             velec            = _mm256_mul_ps(qq22,velec);
1418
1419             d                = _mm256_sub_ps(r22,rswitch);
1420             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1421             d2               = _mm256_mul_ps(d,d);
1422             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1423
1424             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1425
1426             /* Evaluate switch function */
1427             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1428             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv22,_mm256_mul_ps(velec,dsw)) );
1429             velec            = _mm256_mul_ps(velec,sw);
1430             cutoff_mask      = _mm256_cmp_ps(rsq22,rcutoff2,_CMP_LT_OQ);
1431
1432             /* Update potential sum for this i atom from the interaction with this j atom. */
1433             velec            = _mm256_and_ps(velec,cutoff_mask);
1434             velec            = _mm256_andnot_ps(dummy_mask,velec);
1435             velecsum         = _mm256_add_ps(velecsum,velec);
1436
1437             fscal            = felec;
1438
1439             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1440
1441             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1442
1443             /* Calculate temporary vectorial force */
1444             tx               = _mm256_mul_ps(fscal,dx22);
1445             ty               = _mm256_mul_ps(fscal,dy22);
1446             tz               = _mm256_mul_ps(fscal,dz22);
1447
1448             /* Update vectorial force */
1449             fix2             = _mm256_add_ps(fix2,tx);
1450             fiy2             = _mm256_add_ps(fiy2,ty);
1451             fiz2             = _mm256_add_ps(fiz2,tz);
1452
1453             fjx2             = _mm256_add_ps(fjx2,tx);
1454             fjy2             = _mm256_add_ps(fjy2,ty);
1455             fjz2             = _mm256_add_ps(fjz2,tz);
1456
1457             }
1458
1459             /**************************
1460              * CALCULATE INTERACTIONS *
1461              **************************/
1462
1463             if (gmx_mm256_any_lt(rsq23,rcutoff2))
1464             {
1465
1466             r23              = _mm256_mul_ps(rsq23,rinv23);
1467             r23              = _mm256_andnot_ps(dummy_mask,r23);
1468
1469             /* EWALD ELECTROSTATICS */
1470
1471             /* Analytical PME correction */
1472             zeta2            = _mm256_mul_ps(beta2,rsq23);
1473             rinv3            = _mm256_mul_ps(rinvsq23,rinv23);
1474             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1475             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1476             felec            = _mm256_mul_ps(qq23,felec);
1477             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1478             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1479             velec            = _mm256_sub_ps(rinv23,pmecorrV);
1480             velec            = _mm256_mul_ps(qq23,velec);
1481
1482             d                = _mm256_sub_ps(r23,rswitch);
1483             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1484             d2               = _mm256_mul_ps(d,d);
1485             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1486
1487             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1488
1489             /* Evaluate switch function */
1490             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1491             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv23,_mm256_mul_ps(velec,dsw)) );
1492             velec            = _mm256_mul_ps(velec,sw);
1493             cutoff_mask      = _mm256_cmp_ps(rsq23,rcutoff2,_CMP_LT_OQ);
1494
1495             /* Update potential sum for this i atom from the interaction with this j atom. */
1496             velec            = _mm256_and_ps(velec,cutoff_mask);
1497             velec            = _mm256_andnot_ps(dummy_mask,velec);
1498             velecsum         = _mm256_add_ps(velecsum,velec);
1499
1500             fscal            = felec;
1501
1502             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1503
1504             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1505
1506             /* Calculate temporary vectorial force */
1507             tx               = _mm256_mul_ps(fscal,dx23);
1508             ty               = _mm256_mul_ps(fscal,dy23);
1509             tz               = _mm256_mul_ps(fscal,dz23);
1510
1511             /* Update vectorial force */
1512             fix2             = _mm256_add_ps(fix2,tx);
1513             fiy2             = _mm256_add_ps(fiy2,ty);
1514             fiz2             = _mm256_add_ps(fiz2,tz);
1515
1516             fjx3             = _mm256_add_ps(fjx3,tx);
1517             fjy3             = _mm256_add_ps(fjy3,ty);
1518             fjz3             = _mm256_add_ps(fjz3,tz);
1519
1520             }
1521
1522             /**************************
1523              * CALCULATE INTERACTIONS *
1524              **************************/
1525
1526             if (gmx_mm256_any_lt(rsq31,rcutoff2))
1527             {
1528
1529             r31              = _mm256_mul_ps(rsq31,rinv31);
1530             r31              = _mm256_andnot_ps(dummy_mask,r31);
1531
1532             /* EWALD ELECTROSTATICS */
1533
1534             /* Analytical PME correction */
1535             zeta2            = _mm256_mul_ps(beta2,rsq31);
1536             rinv3            = _mm256_mul_ps(rinvsq31,rinv31);
1537             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1538             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1539             felec            = _mm256_mul_ps(qq31,felec);
1540             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1541             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1542             velec            = _mm256_sub_ps(rinv31,pmecorrV);
1543             velec            = _mm256_mul_ps(qq31,velec);
1544
1545             d                = _mm256_sub_ps(r31,rswitch);
1546             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1547             d2               = _mm256_mul_ps(d,d);
1548             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1549
1550             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1551
1552             /* Evaluate switch function */
1553             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1554             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv31,_mm256_mul_ps(velec,dsw)) );
1555             velec            = _mm256_mul_ps(velec,sw);
1556             cutoff_mask      = _mm256_cmp_ps(rsq31,rcutoff2,_CMP_LT_OQ);
1557
1558             /* Update potential sum for this i atom from the interaction with this j atom. */
1559             velec            = _mm256_and_ps(velec,cutoff_mask);
1560             velec            = _mm256_andnot_ps(dummy_mask,velec);
1561             velecsum         = _mm256_add_ps(velecsum,velec);
1562
1563             fscal            = felec;
1564
1565             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1566
1567             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1568
1569             /* Calculate temporary vectorial force */
1570             tx               = _mm256_mul_ps(fscal,dx31);
1571             ty               = _mm256_mul_ps(fscal,dy31);
1572             tz               = _mm256_mul_ps(fscal,dz31);
1573
1574             /* Update vectorial force */
1575             fix3             = _mm256_add_ps(fix3,tx);
1576             fiy3             = _mm256_add_ps(fiy3,ty);
1577             fiz3             = _mm256_add_ps(fiz3,tz);
1578
1579             fjx1             = _mm256_add_ps(fjx1,tx);
1580             fjy1             = _mm256_add_ps(fjy1,ty);
1581             fjz1             = _mm256_add_ps(fjz1,tz);
1582
1583             }
1584
1585             /**************************
1586              * CALCULATE INTERACTIONS *
1587              **************************/
1588
1589             if (gmx_mm256_any_lt(rsq32,rcutoff2))
1590             {
1591
1592             r32              = _mm256_mul_ps(rsq32,rinv32);
1593             r32              = _mm256_andnot_ps(dummy_mask,r32);
1594
1595             /* EWALD ELECTROSTATICS */
1596
1597             /* Analytical PME correction */
1598             zeta2            = _mm256_mul_ps(beta2,rsq32);
1599             rinv3            = _mm256_mul_ps(rinvsq32,rinv32);
1600             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1601             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1602             felec            = _mm256_mul_ps(qq32,felec);
1603             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1604             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1605             velec            = _mm256_sub_ps(rinv32,pmecorrV);
1606             velec            = _mm256_mul_ps(qq32,velec);
1607
1608             d                = _mm256_sub_ps(r32,rswitch);
1609             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1610             d2               = _mm256_mul_ps(d,d);
1611             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1612
1613             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1614
1615             /* Evaluate switch function */
1616             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1617             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv32,_mm256_mul_ps(velec,dsw)) );
1618             velec            = _mm256_mul_ps(velec,sw);
1619             cutoff_mask      = _mm256_cmp_ps(rsq32,rcutoff2,_CMP_LT_OQ);
1620
1621             /* Update potential sum for this i atom from the interaction with this j atom. */
1622             velec            = _mm256_and_ps(velec,cutoff_mask);
1623             velec            = _mm256_andnot_ps(dummy_mask,velec);
1624             velecsum         = _mm256_add_ps(velecsum,velec);
1625
1626             fscal            = felec;
1627
1628             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1629
1630             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1631
1632             /* Calculate temporary vectorial force */
1633             tx               = _mm256_mul_ps(fscal,dx32);
1634             ty               = _mm256_mul_ps(fscal,dy32);
1635             tz               = _mm256_mul_ps(fscal,dz32);
1636
1637             /* Update vectorial force */
1638             fix3             = _mm256_add_ps(fix3,tx);
1639             fiy3             = _mm256_add_ps(fiy3,ty);
1640             fiz3             = _mm256_add_ps(fiz3,tz);
1641
1642             fjx2             = _mm256_add_ps(fjx2,tx);
1643             fjy2             = _mm256_add_ps(fjy2,ty);
1644             fjz2             = _mm256_add_ps(fjz2,tz);
1645
1646             }
1647
1648             /**************************
1649              * CALCULATE INTERACTIONS *
1650              **************************/
1651
1652             if (gmx_mm256_any_lt(rsq33,rcutoff2))
1653             {
1654
1655             r33              = _mm256_mul_ps(rsq33,rinv33);
1656             r33              = _mm256_andnot_ps(dummy_mask,r33);
1657
1658             /* EWALD ELECTROSTATICS */
1659
1660             /* Analytical PME correction */
1661             zeta2            = _mm256_mul_ps(beta2,rsq33);
1662             rinv3            = _mm256_mul_ps(rinvsq33,rinv33);
1663             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
1664             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1665             felec            = _mm256_mul_ps(qq33,felec);
1666             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
1667             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
1668             velec            = _mm256_sub_ps(rinv33,pmecorrV);
1669             velec            = _mm256_mul_ps(qq33,velec);
1670
1671             d                = _mm256_sub_ps(r33,rswitch);
1672             d                = _mm256_max_ps(d,_mm256_setzero_ps());
1673             d2               = _mm256_mul_ps(d,d);
1674             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
1675
1676             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
1677
1678             /* Evaluate switch function */
1679             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1680             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv33,_mm256_mul_ps(velec,dsw)) );
1681             velec            = _mm256_mul_ps(velec,sw);
1682             cutoff_mask      = _mm256_cmp_ps(rsq33,rcutoff2,_CMP_LT_OQ);
1683
1684             /* Update potential sum for this i atom from the interaction with this j atom. */
1685             velec            = _mm256_and_ps(velec,cutoff_mask);
1686             velec            = _mm256_andnot_ps(dummy_mask,velec);
1687             velecsum         = _mm256_add_ps(velecsum,velec);
1688
1689             fscal            = felec;
1690
1691             fscal            = _mm256_and_ps(fscal,cutoff_mask);
1692
1693             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
1694
1695             /* Calculate temporary vectorial force */
1696             tx               = _mm256_mul_ps(fscal,dx33);
1697             ty               = _mm256_mul_ps(fscal,dy33);
1698             tz               = _mm256_mul_ps(fscal,dz33);
1699
1700             /* Update vectorial force */
1701             fix3             = _mm256_add_ps(fix3,tx);
1702             fiy3             = _mm256_add_ps(fiy3,ty);
1703             fiz3             = _mm256_add_ps(fiz3,tz);
1704
1705             fjx3             = _mm256_add_ps(fjx3,tx);
1706             fjy3             = _mm256_add_ps(fjy3,ty);
1707             fjz3             = _mm256_add_ps(fjz3,tz);
1708
1709             }
1710
1711             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1712             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1713             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1714             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1715             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
1716             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
1717             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
1718             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
1719
1720             gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
1721                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1722                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1723
1724             /* Inner loop uses 1044 flops */
1725         }
1726
1727         /* End of innermost loop */
1728
1729         gmx_mm256_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1730                                                  f+i_coord_offset,fshift+i_shift_offset);
1731
1732         ggid                        = gid[iidx];
1733         /* Update potential energies */
1734         gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
1735         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
1736
1737         /* Increment number of inner iterations */
1738         inneriter                  += j_index_end - j_index_start;
1739
1740         /* Outer loop uses 26 flops */
1741     }
1742
1743     /* Increment number of outer iterations */
1744     outeriter        += nri;
1745
1746     /* Update outer/inner flops */
1747
1748     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*1044);
1749 }
1750 /*
1751  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_avx_256_single
1752  * Electrostatics interaction: Ewald
1753  * VdW interaction:            LennardJones
1754  * Geometry:                   Water4-Water4
1755  * Calculate force/pot:        Force
1756  */
1757 void
1758 nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_avx_256_single
1759                     (t_nblist                    * gmx_restrict       nlist,
1760                      rvec                        * gmx_restrict          xx,
1761                      rvec                        * gmx_restrict          ff,
1762                      t_forcerec                  * gmx_restrict          fr,
1763                      t_mdatoms                   * gmx_restrict     mdatoms,
1764                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1765                      t_nrnb                      * gmx_restrict        nrnb)
1766 {
1767     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1768      * just 0 for non-waters.
1769      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
1770      * jnr indices corresponding to data put in the four positions in the SIMD register.
1771      */
1772     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1773     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1774     int              jnrA,jnrB,jnrC,jnrD;
1775     int              jnrE,jnrF,jnrG,jnrH;
1776     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1777     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1778     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1779     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
1780     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1781     real             rcutoff_scalar;
1782     real             *shiftvec,*fshift,*x,*f;
1783     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
1784     real             scratch[4*DIM];
1785     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1786     real *           vdwioffsetptr0;
1787     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1788     real *           vdwioffsetptr1;
1789     __m256           ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1790     real *           vdwioffsetptr2;
1791     __m256           ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1792     real *           vdwioffsetptr3;
1793     __m256           ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1794     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
1795     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1796     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D,vdwjidx1E,vdwjidx1F,vdwjidx1G,vdwjidx1H;
1797     __m256           jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1798     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D,vdwjidx2E,vdwjidx2F,vdwjidx2G,vdwjidx2H;
1799     __m256           jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1800     int              vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D,vdwjidx3E,vdwjidx3F,vdwjidx3G,vdwjidx3H;
1801     __m256           jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1802     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1803     __m256           dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1804     __m256           dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1805     __m256           dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1806     __m256           dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1807     __m256           dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1808     __m256           dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1809     __m256           dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1810     __m256           dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1811     __m256           dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1812     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
1813     real             *charge;
1814     int              nvdwtype;
1815     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1816     int              *vdwtype;
1817     real             *vdwparam;
1818     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
1819     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
1820     __m256i          ewitab;
1821     __m128i          ewitab_lo,ewitab_hi;
1822     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1823     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
1824     real             *ewtab;
1825     __m256           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
1826     real             rswitch_scalar,d_scalar;
1827     __m256           dummy_mask,cutoff_mask;
1828     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
1829     __m256           one     = _mm256_set1_ps(1.0);
1830     __m256           two     = _mm256_set1_ps(2.0);
1831     x                = xx[0];
1832     f                = ff[0];
1833
1834     nri              = nlist->nri;
1835     iinr             = nlist->iinr;
1836     jindex           = nlist->jindex;
1837     jjnr             = nlist->jjnr;
1838     shiftidx         = nlist->shift;
1839     gid              = nlist->gid;
1840     shiftvec         = fr->shift_vec[0];
1841     fshift           = fr->fshift[0];
1842     facel            = _mm256_set1_ps(fr->epsfac);
1843     charge           = mdatoms->chargeA;
1844     nvdwtype         = fr->ntype;
1845     vdwparam         = fr->nbfp;
1846     vdwtype          = mdatoms->typeA;
1847
1848     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
1849     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
1850     beta2            = _mm256_mul_ps(beta,beta);
1851     beta3            = _mm256_mul_ps(beta,beta2);
1852
1853     ewtab            = fr->ic->tabq_coul_FDV0;
1854     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
1855     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
1856
1857     /* Setup water-specific parameters */
1858     inr              = nlist->iinr[0];
1859     iq1              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
1860     iq2              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
1861     iq3              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+3]));
1862     vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
1863
1864     jq1              = _mm256_set1_ps(charge[inr+1]);
1865     jq2              = _mm256_set1_ps(charge[inr+2]);
1866     jq3              = _mm256_set1_ps(charge[inr+3]);
1867     vdwjidx0A        = 2*vdwtype[inr+0];
1868     c6_00            = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A]);
1869     c12_00           = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A+1]);
1870     qq11             = _mm256_mul_ps(iq1,jq1);
1871     qq12             = _mm256_mul_ps(iq1,jq2);
1872     qq13             = _mm256_mul_ps(iq1,jq3);
1873     qq21             = _mm256_mul_ps(iq2,jq1);
1874     qq22             = _mm256_mul_ps(iq2,jq2);
1875     qq23             = _mm256_mul_ps(iq2,jq3);
1876     qq31             = _mm256_mul_ps(iq3,jq1);
1877     qq32             = _mm256_mul_ps(iq3,jq2);
1878     qq33             = _mm256_mul_ps(iq3,jq3);
1879
1880     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
1881     rcutoff_scalar   = fr->rcoulomb;
1882     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
1883     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
1884
1885     rswitch_scalar   = fr->rcoulomb_switch;
1886     rswitch          = _mm256_set1_ps(rswitch_scalar);
1887     /* Setup switch parameters */
1888     d_scalar         = rcutoff_scalar-rswitch_scalar;
1889     d                = _mm256_set1_ps(d_scalar);
1890     swV3             = _mm256_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
1891     swV4             = _mm256_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1892     swV5             = _mm256_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1893     swF2             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
1894     swF3             = _mm256_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1895     swF4             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1896
1897     /* Avoid stupid compiler warnings */
1898     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
1899     j_coord_offsetA = 0;
1900     j_coord_offsetB = 0;
1901     j_coord_offsetC = 0;
1902     j_coord_offsetD = 0;
1903     j_coord_offsetE = 0;
1904     j_coord_offsetF = 0;
1905     j_coord_offsetG = 0;
1906     j_coord_offsetH = 0;
1907
1908     outeriter        = 0;
1909     inneriter        = 0;
1910
1911     for(iidx=0;iidx<4*DIM;iidx++)
1912     {
1913         scratch[iidx] = 0.0;
1914     }
1915
1916     /* Start outer loop over neighborlists */
1917     for(iidx=0; iidx<nri; iidx++)
1918     {
1919         /* Load shift vector for this list */
1920         i_shift_offset   = DIM*shiftidx[iidx];
1921
1922         /* Load limits for loop over neighbors */
1923         j_index_start    = jindex[iidx];
1924         j_index_end      = jindex[iidx+1];
1925
1926         /* Get outer coordinate index */
1927         inr              = iinr[iidx];
1928         i_coord_offset   = DIM*inr;
1929
1930         /* Load i particle coords and add shift vector */
1931         gmx_mm256_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
1932                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1933
1934         fix0             = _mm256_setzero_ps();
1935         fiy0             = _mm256_setzero_ps();
1936         fiz0             = _mm256_setzero_ps();
1937         fix1             = _mm256_setzero_ps();
1938         fiy1             = _mm256_setzero_ps();
1939         fiz1             = _mm256_setzero_ps();
1940         fix2             = _mm256_setzero_ps();
1941         fiy2             = _mm256_setzero_ps();
1942         fiz2             = _mm256_setzero_ps();
1943         fix3             = _mm256_setzero_ps();
1944         fiy3             = _mm256_setzero_ps();
1945         fiz3             = _mm256_setzero_ps();
1946
1947         /* Start inner kernel loop */
1948         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
1949         {
1950
1951             /* Get j neighbor index, and coordinate index */
1952             jnrA             = jjnr[jidx];
1953             jnrB             = jjnr[jidx+1];
1954             jnrC             = jjnr[jidx+2];
1955             jnrD             = jjnr[jidx+3];
1956             jnrE             = jjnr[jidx+4];
1957             jnrF             = jjnr[jidx+5];
1958             jnrG             = jjnr[jidx+6];
1959             jnrH             = jjnr[jidx+7];
1960             j_coord_offsetA  = DIM*jnrA;
1961             j_coord_offsetB  = DIM*jnrB;
1962             j_coord_offsetC  = DIM*jnrC;
1963             j_coord_offsetD  = DIM*jnrD;
1964             j_coord_offsetE  = DIM*jnrE;
1965             j_coord_offsetF  = DIM*jnrF;
1966             j_coord_offsetG  = DIM*jnrG;
1967             j_coord_offsetH  = DIM*jnrH;
1968
1969             /* load j atom coordinates */
1970             gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1971                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1972                                                  x+j_coord_offsetE,x+j_coord_offsetF,
1973                                                  x+j_coord_offsetG,x+j_coord_offsetH,
1974                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1975                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
1976
1977             /* Calculate displacement vector */
1978             dx00             = _mm256_sub_ps(ix0,jx0);
1979             dy00             = _mm256_sub_ps(iy0,jy0);
1980             dz00             = _mm256_sub_ps(iz0,jz0);
1981             dx11             = _mm256_sub_ps(ix1,jx1);
1982             dy11             = _mm256_sub_ps(iy1,jy1);
1983             dz11             = _mm256_sub_ps(iz1,jz1);
1984             dx12             = _mm256_sub_ps(ix1,jx2);
1985             dy12             = _mm256_sub_ps(iy1,jy2);
1986             dz12             = _mm256_sub_ps(iz1,jz2);
1987             dx13             = _mm256_sub_ps(ix1,jx3);
1988             dy13             = _mm256_sub_ps(iy1,jy3);
1989             dz13             = _mm256_sub_ps(iz1,jz3);
1990             dx21             = _mm256_sub_ps(ix2,jx1);
1991             dy21             = _mm256_sub_ps(iy2,jy1);
1992             dz21             = _mm256_sub_ps(iz2,jz1);
1993             dx22             = _mm256_sub_ps(ix2,jx2);
1994             dy22             = _mm256_sub_ps(iy2,jy2);
1995             dz22             = _mm256_sub_ps(iz2,jz2);
1996             dx23             = _mm256_sub_ps(ix2,jx3);
1997             dy23             = _mm256_sub_ps(iy2,jy3);
1998             dz23             = _mm256_sub_ps(iz2,jz3);
1999             dx31             = _mm256_sub_ps(ix3,jx1);
2000             dy31             = _mm256_sub_ps(iy3,jy1);
2001             dz31             = _mm256_sub_ps(iz3,jz1);
2002             dx32             = _mm256_sub_ps(ix3,jx2);
2003             dy32             = _mm256_sub_ps(iy3,jy2);
2004             dz32             = _mm256_sub_ps(iz3,jz2);
2005             dx33             = _mm256_sub_ps(ix3,jx3);
2006             dy33             = _mm256_sub_ps(iy3,jy3);
2007             dz33             = _mm256_sub_ps(iz3,jz3);
2008
2009             /* Calculate squared distance and things based on it */
2010             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
2011             rsq11            = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
2012             rsq12            = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
2013             rsq13            = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
2014             rsq21            = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
2015             rsq22            = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
2016             rsq23            = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
2017             rsq31            = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
2018             rsq32            = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
2019             rsq33            = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
2020
2021             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
2022             rinv11           = gmx_mm256_invsqrt_ps(rsq11);
2023             rinv12           = gmx_mm256_invsqrt_ps(rsq12);
2024             rinv13           = gmx_mm256_invsqrt_ps(rsq13);
2025             rinv21           = gmx_mm256_invsqrt_ps(rsq21);
2026             rinv22           = gmx_mm256_invsqrt_ps(rsq22);
2027             rinv23           = gmx_mm256_invsqrt_ps(rsq23);
2028             rinv31           = gmx_mm256_invsqrt_ps(rsq31);
2029             rinv32           = gmx_mm256_invsqrt_ps(rsq32);
2030             rinv33           = gmx_mm256_invsqrt_ps(rsq33);
2031
2032             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
2033             rinvsq11         = _mm256_mul_ps(rinv11,rinv11);
2034             rinvsq12         = _mm256_mul_ps(rinv12,rinv12);
2035             rinvsq13         = _mm256_mul_ps(rinv13,rinv13);
2036             rinvsq21         = _mm256_mul_ps(rinv21,rinv21);
2037             rinvsq22         = _mm256_mul_ps(rinv22,rinv22);
2038             rinvsq23         = _mm256_mul_ps(rinv23,rinv23);
2039             rinvsq31         = _mm256_mul_ps(rinv31,rinv31);
2040             rinvsq32         = _mm256_mul_ps(rinv32,rinv32);
2041             rinvsq33         = _mm256_mul_ps(rinv33,rinv33);
2042
2043             fjx0             = _mm256_setzero_ps();
2044             fjy0             = _mm256_setzero_ps();
2045             fjz0             = _mm256_setzero_ps();
2046             fjx1             = _mm256_setzero_ps();
2047             fjy1             = _mm256_setzero_ps();
2048             fjz1             = _mm256_setzero_ps();
2049             fjx2             = _mm256_setzero_ps();
2050             fjy2             = _mm256_setzero_ps();
2051             fjz2             = _mm256_setzero_ps();
2052             fjx3             = _mm256_setzero_ps();
2053             fjy3             = _mm256_setzero_ps();
2054             fjz3             = _mm256_setzero_ps();
2055
2056             /**************************
2057              * CALCULATE INTERACTIONS *
2058              **************************/
2059
2060             if (gmx_mm256_any_lt(rsq00,rcutoff2))
2061             {
2062
2063             r00              = _mm256_mul_ps(rsq00,rinv00);
2064
2065             /* LENNARD-JONES DISPERSION/REPULSION */
2066
2067             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
2068             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
2069             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
2070             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
2071             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
2072
2073             d                = _mm256_sub_ps(r00,rswitch);
2074             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2075             d2               = _mm256_mul_ps(d,d);
2076             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2077
2078             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2079
2080             /* Evaluate switch function */
2081             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2082             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
2083             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
2084
2085             fscal            = fvdw;
2086
2087             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2088
2089             /* Calculate temporary vectorial force */
2090             tx               = _mm256_mul_ps(fscal,dx00);
2091             ty               = _mm256_mul_ps(fscal,dy00);
2092             tz               = _mm256_mul_ps(fscal,dz00);
2093
2094             /* Update vectorial force */
2095             fix0             = _mm256_add_ps(fix0,tx);
2096             fiy0             = _mm256_add_ps(fiy0,ty);
2097             fiz0             = _mm256_add_ps(fiz0,tz);
2098
2099             fjx0             = _mm256_add_ps(fjx0,tx);
2100             fjy0             = _mm256_add_ps(fjy0,ty);
2101             fjz0             = _mm256_add_ps(fjz0,tz);
2102
2103             }
2104
2105             /**************************
2106              * CALCULATE INTERACTIONS *
2107              **************************/
2108
2109             if (gmx_mm256_any_lt(rsq11,rcutoff2))
2110             {
2111
2112             r11              = _mm256_mul_ps(rsq11,rinv11);
2113
2114             /* EWALD ELECTROSTATICS */
2115
2116             /* Analytical PME correction */
2117             zeta2            = _mm256_mul_ps(beta2,rsq11);
2118             rinv3            = _mm256_mul_ps(rinvsq11,rinv11);
2119             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2120             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2121             felec            = _mm256_mul_ps(qq11,felec);
2122             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2123             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2124             velec            = _mm256_sub_ps(rinv11,pmecorrV);
2125             velec            = _mm256_mul_ps(qq11,velec);
2126
2127             d                = _mm256_sub_ps(r11,rswitch);
2128             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2129             d2               = _mm256_mul_ps(d,d);
2130             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2131
2132             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2133
2134             /* Evaluate switch function */
2135             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2136             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv11,_mm256_mul_ps(velec,dsw)) );
2137             cutoff_mask      = _mm256_cmp_ps(rsq11,rcutoff2,_CMP_LT_OQ);
2138
2139             fscal            = felec;
2140
2141             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2142
2143             /* Calculate temporary vectorial force */
2144             tx               = _mm256_mul_ps(fscal,dx11);
2145             ty               = _mm256_mul_ps(fscal,dy11);
2146             tz               = _mm256_mul_ps(fscal,dz11);
2147
2148             /* Update vectorial force */
2149             fix1             = _mm256_add_ps(fix1,tx);
2150             fiy1             = _mm256_add_ps(fiy1,ty);
2151             fiz1             = _mm256_add_ps(fiz1,tz);
2152
2153             fjx1             = _mm256_add_ps(fjx1,tx);
2154             fjy1             = _mm256_add_ps(fjy1,ty);
2155             fjz1             = _mm256_add_ps(fjz1,tz);
2156
2157             }
2158
2159             /**************************
2160              * CALCULATE INTERACTIONS *
2161              **************************/
2162
2163             if (gmx_mm256_any_lt(rsq12,rcutoff2))
2164             {
2165
2166             r12              = _mm256_mul_ps(rsq12,rinv12);
2167
2168             /* EWALD ELECTROSTATICS */
2169
2170             /* Analytical PME correction */
2171             zeta2            = _mm256_mul_ps(beta2,rsq12);
2172             rinv3            = _mm256_mul_ps(rinvsq12,rinv12);
2173             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2174             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2175             felec            = _mm256_mul_ps(qq12,felec);
2176             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2177             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2178             velec            = _mm256_sub_ps(rinv12,pmecorrV);
2179             velec            = _mm256_mul_ps(qq12,velec);
2180
2181             d                = _mm256_sub_ps(r12,rswitch);
2182             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2183             d2               = _mm256_mul_ps(d,d);
2184             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2185
2186             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2187
2188             /* Evaluate switch function */
2189             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2190             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv12,_mm256_mul_ps(velec,dsw)) );
2191             cutoff_mask      = _mm256_cmp_ps(rsq12,rcutoff2,_CMP_LT_OQ);
2192
2193             fscal            = felec;
2194
2195             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2196
2197             /* Calculate temporary vectorial force */
2198             tx               = _mm256_mul_ps(fscal,dx12);
2199             ty               = _mm256_mul_ps(fscal,dy12);
2200             tz               = _mm256_mul_ps(fscal,dz12);
2201
2202             /* Update vectorial force */
2203             fix1             = _mm256_add_ps(fix1,tx);
2204             fiy1             = _mm256_add_ps(fiy1,ty);
2205             fiz1             = _mm256_add_ps(fiz1,tz);
2206
2207             fjx2             = _mm256_add_ps(fjx2,tx);
2208             fjy2             = _mm256_add_ps(fjy2,ty);
2209             fjz2             = _mm256_add_ps(fjz2,tz);
2210
2211             }
2212
2213             /**************************
2214              * CALCULATE INTERACTIONS *
2215              **************************/
2216
2217             if (gmx_mm256_any_lt(rsq13,rcutoff2))
2218             {
2219
2220             r13              = _mm256_mul_ps(rsq13,rinv13);
2221
2222             /* EWALD ELECTROSTATICS */
2223
2224             /* Analytical PME correction */
2225             zeta2            = _mm256_mul_ps(beta2,rsq13);
2226             rinv3            = _mm256_mul_ps(rinvsq13,rinv13);
2227             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2228             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2229             felec            = _mm256_mul_ps(qq13,felec);
2230             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2231             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2232             velec            = _mm256_sub_ps(rinv13,pmecorrV);
2233             velec            = _mm256_mul_ps(qq13,velec);
2234
2235             d                = _mm256_sub_ps(r13,rswitch);
2236             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2237             d2               = _mm256_mul_ps(d,d);
2238             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2239
2240             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2241
2242             /* Evaluate switch function */
2243             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2244             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv13,_mm256_mul_ps(velec,dsw)) );
2245             cutoff_mask      = _mm256_cmp_ps(rsq13,rcutoff2,_CMP_LT_OQ);
2246
2247             fscal            = felec;
2248
2249             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2250
2251             /* Calculate temporary vectorial force */
2252             tx               = _mm256_mul_ps(fscal,dx13);
2253             ty               = _mm256_mul_ps(fscal,dy13);
2254             tz               = _mm256_mul_ps(fscal,dz13);
2255
2256             /* Update vectorial force */
2257             fix1             = _mm256_add_ps(fix1,tx);
2258             fiy1             = _mm256_add_ps(fiy1,ty);
2259             fiz1             = _mm256_add_ps(fiz1,tz);
2260
2261             fjx3             = _mm256_add_ps(fjx3,tx);
2262             fjy3             = _mm256_add_ps(fjy3,ty);
2263             fjz3             = _mm256_add_ps(fjz3,tz);
2264
2265             }
2266
2267             /**************************
2268              * CALCULATE INTERACTIONS *
2269              **************************/
2270
2271             if (gmx_mm256_any_lt(rsq21,rcutoff2))
2272             {
2273
2274             r21              = _mm256_mul_ps(rsq21,rinv21);
2275
2276             /* EWALD ELECTROSTATICS */
2277
2278             /* Analytical PME correction */
2279             zeta2            = _mm256_mul_ps(beta2,rsq21);
2280             rinv3            = _mm256_mul_ps(rinvsq21,rinv21);
2281             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2282             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2283             felec            = _mm256_mul_ps(qq21,felec);
2284             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2285             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2286             velec            = _mm256_sub_ps(rinv21,pmecorrV);
2287             velec            = _mm256_mul_ps(qq21,velec);
2288
2289             d                = _mm256_sub_ps(r21,rswitch);
2290             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2291             d2               = _mm256_mul_ps(d,d);
2292             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2293
2294             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2295
2296             /* Evaluate switch function */
2297             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2298             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv21,_mm256_mul_ps(velec,dsw)) );
2299             cutoff_mask      = _mm256_cmp_ps(rsq21,rcutoff2,_CMP_LT_OQ);
2300
2301             fscal            = felec;
2302
2303             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2304
2305             /* Calculate temporary vectorial force */
2306             tx               = _mm256_mul_ps(fscal,dx21);
2307             ty               = _mm256_mul_ps(fscal,dy21);
2308             tz               = _mm256_mul_ps(fscal,dz21);
2309
2310             /* Update vectorial force */
2311             fix2             = _mm256_add_ps(fix2,tx);
2312             fiy2             = _mm256_add_ps(fiy2,ty);
2313             fiz2             = _mm256_add_ps(fiz2,tz);
2314
2315             fjx1             = _mm256_add_ps(fjx1,tx);
2316             fjy1             = _mm256_add_ps(fjy1,ty);
2317             fjz1             = _mm256_add_ps(fjz1,tz);
2318
2319             }
2320
2321             /**************************
2322              * CALCULATE INTERACTIONS *
2323              **************************/
2324
2325             if (gmx_mm256_any_lt(rsq22,rcutoff2))
2326             {
2327
2328             r22              = _mm256_mul_ps(rsq22,rinv22);
2329
2330             /* EWALD ELECTROSTATICS */
2331
2332             /* Analytical PME correction */
2333             zeta2            = _mm256_mul_ps(beta2,rsq22);
2334             rinv3            = _mm256_mul_ps(rinvsq22,rinv22);
2335             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2336             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2337             felec            = _mm256_mul_ps(qq22,felec);
2338             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2339             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2340             velec            = _mm256_sub_ps(rinv22,pmecorrV);
2341             velec            = _mm256_mul_ps(qq22,velec);
2342
2343             d                = _mm256_sub_ps(r22,rswitch);
2344             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2345             d2               = _mm256_mul_ps(d,d);
2346             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2347
2348             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2349
2350             /* Evaluate switch function */
2351             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2352             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv22,_mm256_mul_ps(velec,dsw)) );
2353             cutoff_mask      = _mm256_cmp_ps(rsq22,rcutoff2,_CMP_LT_OQ);
2354
2355             fscal            = felec;
2356
2357             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2358
2359             /* Calculate temporary vectorial force */
2360             tx               = _mm256_mul_ps(fscal,dx22);
2361             ty               = _mm256_mul_ps(fscal,dy22);
2362             tz               = _mm256_mul_ps(fscal,dz22);
2363
2364             /* Update vectorial force */
2365             fix2             = _mm256_add_ps(fix2,tx);
2366             fiy2             = _mm256_add_ps(fiy2,ty);
2367             fiz2             = _mm256_add_ps(fiz2,tz);
2368
2369             fjx2             = _mm256_add_ps(fjx2,tx);
2370             fjy2             = _mm256_add_ps(fjy2,ty);
2371             fjz2             = _mm256_add_ps(fjz2,tz);
2372
2373             }
2374
2375             /**************************
2376              * CALCULATE INTERACTIONS *
2377              **************************/
2378
2379             if (gmx_mm256_any_lt(rsq23,rcutoff2))
2380             {
2381
2382             r23              = _mm256_mul_ps(rsq23,rinv23);
2383
2384             /* EWALD ELECTROSTATICS */
2385
2386             /* Analytical PME correction */
2387             zeta2            = _mm256_mul_ps(beta2,rsq23);
2388             rinv3            = _mm256_mul_ps(rinvsq23,rinv23);
2389             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2390             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2391             felec            = _mm256_mul_ps(qq23,felec);
2392             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2393             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2394             velec            = _mm256_sub_ps(rinv23,pmecorrV);
2395             velec            = _mm256_mul_ps(qq23,velec);
2396
2397             d                = _mm256_sub_ps(r23,rswitch);
2398             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2399             d2               = _mm256_mul_ps(d,d);
2400             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2401
2402             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2403
2404             /* Evaluate switch function */
2405             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2406             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv23,_mm256_mul_ps(velec,dsw)) );
2407             cutoff_mask      = _mm256_cmp_ps(rsq23,rcutoff2,_CMP_LT_OQ);
2408
2409             fscal            = felec;
2410
2411             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2412
2413             /* Calculate temporary vectorial force */
2414             tx               = _mm256_mul_ps(fscal,dx23);
2415             ty               = _mm256_mul_ps(fscal,dy23);
2416             tz               = _mm256_mul_ps(fscal,dz23);
2417
2418             /* Update vectorial force */
2419             fix2             = _mm256_add_ps(fix2,tx);
2420             fiy2             = _mm256_add_ps(fiy2,ty);
2421             fiz2             = _mm256_add_ps(fiz2,tz);
2422
2423             fjx3             = _mm256_add_ps(fjx3,tx);
2424             fjy3             = _mm256_add_ps(fjy3,ty);
2425             fjz3             = _mm256_add_ps(fjz3,tz);
2426
2427             }
2428
2429             /**************************
2430              * CALCULATE INTERACTIONS *
2431              **************************/
2432
2433             if (gmx_mm256_any_lt(rsq31,rcutoff2))
2434             {
2435
2436             r31              = _mm256_mul_ps(rsq31,rinv31);
2437
2438             /* EWALD ELECTROSTATICS */
2439
2440             /* Analytical PME correction */
2441             zeta2            = _mm256_mul_ps(beta2,rsq31);
2442             rinv3            = _mm256_mul_ps(rinvsq31,rinv31);
2443             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2444             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2445             felec            = _mm256_mul_ps(qq31,felec);
2446             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2447             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2448             velec            = _mm256_sub_ps(rinv31,pmecorrV);
2449             velec            = _mm256_mul_ps(qq31,velec);
2450
2451             d                = _mm256_sub_ps(r31,rswitch);
2452             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2453             d2               = _mm256_mul_ps(d,d);
2454             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2455
2456             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2457
2458             /* Evaluate switch function */
2459             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2460             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv31,_mm256_mul_ps(velec,dsw)) );
2461             cutoff_mask      = _mm256_cmp_ps(rsq31,rcutoff2,_CMP_LT_OQ);
2462
2463             fscal            = felec;
2464
2465             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2466
2467             /* Calculate temporary vectorial force */
2468             tx               = _mm256_mul_ps(fscal,dx31);
2469             ty               = _mm256_mul_ps(fscal,dy31);
2470             tz               = _mm256_mul_ps(fscal,dz31);
2471
2472             /* Update vectorial force */
2473             fix3             = _mm256_add_ps(fix3,tx);
2474             fiy3             = _mm256_add_ps(fiy3,ty);
2475             fiz3             = _mm256_add_ps(fiz3,tz);
2476
2477             fjx1             = _mm256_add_ps(fjx1,tx);
2478             fjy1             = _mm256_add_ps(fjy1,ty);
2479             fjz1             = _mm256_add_ps(fjz1,tz);
2480
2481             }
2482
2483             /**************************
2484              * CALCULATE INTERACTIONS *
2485              **************************/
2486
2487             if (gmx_mm256_any_lt(rsq32,rcutoff2))
2488             {
2489
2490             r32              = _mm256_mul_ps(rsq32,rinv32);
2491
2492             /* EWALD ELECTROSTATICS */
2493
2494             /* Analytical PME correction */
2495             zeta2            = _mm256_mul_ps(beta2,rsq32);
2496             rinv3            = _mm256_mul_ps(rinvsq32,rinv32);
2497             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2498             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2499             felec            = _mm256_mul_ps(qq32,felec);
2500             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2501             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2502             velec            = _mm256_sub_ps(rinv32,pmecorrV);
2503             velec            = _mm256_mul_ps(qq32,velec);
2504
2505             d                = _mm256_sub_ps(r32,rswitch);
2506             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2507             d2               = _mm256_mul_ps(d,d);
2508             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2509
2510             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2511
2512             /* Evaluate switch function */
2513             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2514             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv32,_mm256_mul_ps(velec,dsw)) );
2515             cutoff_mask      = _mm256_cmp_ps(rsq32,rcutoff2,_CMP_LT_OQ);
2516
2517             fscal            = felec;
2518
2519             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2520
2521             /* Calculate temporary vectorial force */
2522             tx               = _mm256_mul_ps(fscal,dx32);
2523             ty               = _mm256_mul_ps(fscal,dy32);
2524             tz               = _mm256_mul_ps(fscal,dz32);
2525
2526             /* Update vectorial force */
2527             fix3             = _mm256_add_ps(fix3,tx);
2528             fiy3             = _mm256_add_ps(fiy3,ty);
2529             fiz3             = _mm256_add_ps(fiz3,tz);
2530
2531             fjx2             = _mm256_add_ps(fjx2,tx);
2532             fjy2             = _mm256_add_ps(fjy2,ty);
2533             fjz2             = _mm256_add_ps(fjz2,tz);
2534
2535             }
2536
2537             /**************************
2538              * CALCULATE INTERACTIONS *
2539              **************************/
2540
2541             if (gmx_mm256_any_lt(rsq33,rcutoff2))
2542             {
2543
2544             r33              = _mm256_mul_ps(rsq33,rinv33);
2545
2546             /* EWALD ELECTROSTATICS */
2547
2548             /* Analytical PME correction */
2549             zeta2            = _mm256_mul_ps(beta2,rsq33);
2550             rinv3            = _mm256_mul_ps(rinvsq33,rinv33);
2551             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2552             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2553             felec            = _mm256_mul_ps(qq33,felec);
2554             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2555             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2556             velec            = _mm256_sub_ps(rinv33,pmecorrV);
2557             velec            = _mm256_mul_ps(qq33,velec);
2558
2559             d                = _mm256_sub_ps(r33,rswitch);
2560             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2561             d2               = _mm256_mul_ps(d,d);
2562             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2563
2564             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2565
2566             /* Evaluate switch function */
2567             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2568             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv33,_mm256_mul_ps(velec,dsw)) );
2569             cutoff_mask      = _mm256_cmp_ps(rsq33,rcutoff2,_CMP_LT_OQ);
2570
2571             fscal            = felec;
2572
2573             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2574
2575             /* Calculate temporary vectorial force */
2576             tx               = _mm256_mul_ps(fscal,dx33);
2577             ty               = _mm256_mul_ps(fscal,dy33);
2578             tz               = _mm256_mul_ps(fscal,dz33);
2579
2580             /* Update vectorial force */
2581             fix3             = _mm256_add_ps(fix3,tx);
2582             fiy3             = _mm256_add_ps(fiy3,ty);
2583             fiz3             = _mm256_add_ps(fiz3,tz);
2584
2585             fjx3             = _mm256_add_ps(fjx3,tx);
2586             fjy3             = _mm256_add_ps(fjy3,ty);
2587             fjz3             = _mm256_add_ps(fjz3,tz);
2588
2589             }
2590
2591             fjptrA             = f+j_coord_offsetA;
2592             fjptrB             = f+j_coord_offsetB;
2593             fjptrC             = f+j_coord_offsetC;
2594             fjptrD             = f+j_coord_offsetD;
2595             fjptrE             = f+j_coord_offsetE;
2596             fjptrF             = f+j_coord_offsetF;
2597             fjptrG             = f+j_coord_offsetG;
2598             fjptrH             = f+j_coord_offsetH;
2599
2600             gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
2601                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
2602                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2603
2604             /* Inner loop uses 1004 flops */
2605         }
2606
2607         if(jidx<j_index_end)
2608         {
2609
2610             /* Get j neighbor index, and coordinate index */
2611             jnrlistA         = jjnr[jidx];
2612             jnrlistB         = jjnr[jidx+1];
2613             jnrlistC         = jjnr[jidx+2];
2614             jnrlistD         = jjnr[jidx+3];
2615             jnrlistE         = jjnr[jidx+4];
2616             jnrlistF         = jjnr[jidx+5];
2617             jnrlistG         = jjnr[jidx+6];
2618             jnrlistH         = jjnr[jidx+7];
2619             /* Sign of each element will be negative for non-real atoms.
2620              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
2621              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
2622              */
2623             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
2624                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
2625
2626             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
2627             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
2628             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
2629             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
2630             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
2631             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
2632             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
2633             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
2634             j_coord_offsetA  = DIM*jnrA;
2635             j_coord_offsetB  = DIM*jnrB;
2636             j_coord_offsetC  = DIM*jnrC;
2637             j_coord_offsetD  = DIM*jnrD;
2638             j_coord_offsetE  = DIM*jnrE;
2639             j_coord_offsetF  = DIM*jnrF;
2640             j_coord_offsetG  = DIM*jnrG;
2641             j_coord_offsetH  = DIM*jnrH;
2642
2643             /* load j atom coordinates */
2644             gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
2645                                                  x+j_coord_offsetC,x+j_coord_offsetD,
2646                                                  x+j_coord_offsetE,x+j_coord_offsetF,
2647                                                  x+j_coord_offsetG,x+j_coord_offsetH,
2648                                                  &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
2649                                                  &jy2,&jz2,&jx3,&jy3,&jz3);
2650
2651             /* Calculate displacement vector */
2652             dx00             = _mm256_sub_ps(ix0,jx0);
2653             dy00             = _mm256_sub_ps(iy0,jy0);
2654             dz00             = _mm256_sub_ps(iz0,jz0);
2655             dx11             = _mm256_sub_ps(ix1,jx1);
2656             dy11             = _mm256_sub_ps(iy1,jy1);
2657             dz11             = _mm256_sub_ps(iz1,jz1);
2658             dx12             = _mm256_sub_ps(ix1,jx2);
2659             dy12             = _mm256_sub_ps(iy1,jy2);
2660             dz12             = _mm256_sub_ps(iz1,jz2);
2661             dx13             = _mm256_sub_ps(ix1,jx3);
2662             dy13             = _mm256_sub_ps(iy1,jy3);
2663             dz13             = _mm256_sub_ps(iz1,jz3);
2664             dx21             = _mm256_sub_ps(ix2,jx1);
2665             dy21             = _mm256_sub_ps(iy2,jy1);
2666             dz21             = _mm256_sub_ps(iz2,jz1);
2667             dx22             = _mm256_sub_ps(ix2,jx2);
2668             dy22             = _mm256_sub_ps(iy2,jy2);
2669             dz22             = _mm256_sub_ps(iz2,jz2);
2670             dx23             = _mm256_sub_ps(ix2,jx3);
2671             dy23             = _mm256_sub_ps(iy2,jy3);
2672             dz23             = _mm256_sub_ps(iz2,jz3);
2673             dx31             = _mm256_sub_ps(ix3,jx1);
2674             dy31             = _mm256_sub_ps(iy3,jy1);
2675             dz31             = _mm256_sub_ps(iz3,jz1);
2676             dx32             = _mm256_sub_ps(ix3,jx2);
2677             dy32             = _mm256_sub_ps(iy3,jy2);
2678             dz32             = _mm256_sub_ps(iz3,jz2);
2679             dx33             = _mm256_sub_ps(ix3,jx3);
2680             dy33             = _mm256_sub_ps(iy3,jy3);
2681             dz33             = _mm256_sub_ps(iz3,jz3);
2682
2683             /* Calculate squared distance and things based on it */
2684             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
2685             rsq11            = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
2686             rsq12            = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
2687             rsq13            = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
2688             rsq21            = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
2689             rsq22            = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
2690             rsq23            = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
2691             rsq31            = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
2692             rsq32            = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
2693             rsq33            = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
2694
2695             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
2696             rinv11           = gmx_mm256_invsqrt_ps(rsq11);
2697             rinv12           = gmx_mm256_invsqrt_ps(rsq12);
2698             rinv13           = gmx_mm256_invsqrt_ps(rsq13);
2699             rinv21           = gmx_mm256_invsqrt_ps(rsq21);
2700             rinv22           = gmx_mm256_invsqrt_ps(rsq22);
2701             rinv23           = gmx_mm256_invsqrt_ps(rsq23);
2702             rinv31           = gmx_mm256_invsqrt_ps(rsq31);
2703             rinv32           = gmx_mm256_invsqrt_ps(rsq32);
2704             rinv33           = gmx_mm256_invsqrt_ps(rsq33);
2705
2706             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
2707             rinvsq11         = _mm256_mul_ps(rinv11,rinv11);
2708             rinvsq12         = _mm256_mul_ps(rinv12,rinv12);
2709             rinvsq13         = _mm256_mul_ps(rinv13,rinv13);
2710             rinvsq21         = _mm256_mul_ps(rinv21,rinv21);
2711             rinvsq22         = _mm256_mul_ps(rinv22,rinv22);
2712             rinvsq23         = _mm256_mul_ps(rinv23,rinv23);
2713             rinvsq31         = _mm256_mul_ps(rinv31,rinv31);
2714             rinvsq32         = _mm256_mul_ps(rinv32,rinv32);
2715             rinvsq33         = _mm256_mul_ps(rinv33,rinv33);
2716
2717             fjx0             = _mm256_setzero_ps();
2718             fjy0             = _mm256_setzero_ps();
2719             fjz0             = _mm256_setzero_ps();
2720             fjx1             = _mm256_setzero_ps();
2721             fjy1             = _mm256_setzero_ps();
2722             fjz1             = _mm256_setzero_ps();
2723             fjx2             = _mm256_setzero_ps();
2724             fjy2             = _mm256_setzero_ps();
2725             fjz2             = _mm256_setzero_ps();
2726             fjx3             = _mm256_setzero_ps();
2727             fjy3             = _mm256_setzero_ps();
2728             fjz3             = _mm256_setzero_ps();
2729
2730             /**************************
2731              * CALCULATE INTERACTIONS *
2732              **************************/
2733
2734             if (gmx_mm256_any_lt(rsq00,rcutoff2))
2735             {
2736
2737             r00              = _mm256_mul_ps(rsq00,rinv00);
2738             r00              = _mm256_andnot_ps(dummy_mask,r00);
2739
2740             /* LENNARD-JONES DISPERSION/REPULSION */
2741
2742             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
2743             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
2744             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
2745             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
2746             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
2747
2748             d                = _mm256_sub_ps(r00,rswitch);
2749             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2750             d2               = _mm256_mul_ps(d,d);
2751             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2752
2753             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2754
2755             /* Evaluate switch function */
2756             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2757             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
2758             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
2759
2760             fscal            = fvdw;
2761
2762             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2763
2764             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
2765
2766             /* Calculate temporary vectorial force */
2767             tx               = _mm256_mul_ps(fscal,dx00);
2768             ty               = _mm256_mul_ps(fscal,dy00);
2769             tz               = _mm256_mul_ps(fscal,dz00);
2770
2771             /* Update vectorial force */
2772             fix0             = _mm256_add_ps(fix0,tx);
2773             fiy0             = _mm256_add_ps(fiy0,ty);
2774             fiz0             = _mm256_add_ps(fiz0,tz);
2775
2776             fjx0             = _mm256_add_ps(fjx0,tx);
2777             fjy0             = _mm256_add_ps(fjy0,ty);
2778             fjz0             = _mm256_add_ps(fjz0,tz);
2779
2780             }
2781
2782             /**************************
2783              * CALCULATE INTERACTIONS *
2784              **************************/
2785
2786             if (gmx_mm256_any_lt(rsq11,rcutoff2))
2787             {
2788
2789             r11              = _mm256_mul_ps(rsq11,rinv11);
2790             r11              = _mm256_andnot_ps(dummy_mask,r11);
2791
2792             /* EWALD ELECTROSTATICS */
2793
2794             /* Analytical PME correction */
2795             zeta2            = _mm256_mul_ps(beta2,rsq11);
2796             rinv3            = _mm256_mul_ps(rinvsq11,rinv11);
2797             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2798             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2799             felec            = _mm256_mul_ps(qq11,felec);
2800             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2801             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2802             velec            = _mm256_sub_ps(rinv11,pmecorrV);
2803             velec            = _mm256_mul_ps(qq11,velec);
2804
2805             d                = _mm256_sub_ps(r11,rswitch);
2806             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2807             d2               = _mm256_mul_ps(d,d);
2808             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2809
2810             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2811
2812             /* Evaluate switch function */
2813             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2814             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv11,_mm256_mul_ps(velec,dsw)) );
2815             cutoff_mask      = _mm256_cmp_ps(rsq11,rcutoff2,_CMP_LT_OQ);
2816
2817             fscal            = felec;
2818
2819             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2820
2821             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
2822
2823             /* Calculate temporary vectorial force */
2824             tx               = _mm256_mul_ps(fscal,dx11);
2825             ty               = _mm256_mul_ps(fscal,dy11);
2826             tz               = _mm256_mul_ps(fscal,dz11);
2827
2828             /* Update vectorial force */
2829             fix1             = _mm256_add_ps(fix1,tx);
2830             fiy1             = _mm256_add_ps(fiy1,ty);
2831             fiz1             = _mm256_add_ps(fiz1,tz);
2832
2833             fjx1             = _mm256_add_ps(fjx1,tx);
2834             fjy1             = _mm256_add_ps(fjy1,ty);
2835             fjz1             = _mm256_add_ps(fjz1,tz);
2836
2837             }
2838
2839             /**************************
2840              * CALCULATE INTERACTIONS *
2841              **************************/
2842
2843             if (gmx_mm256_any_lt(rsq12,rcutoff2))
2844             {
2845
2846             r12              = _mm256_mul_ps(rsq12,rinv12);
2847             r12              = _mm256_andnot_ps(dummy_mask,r12);
2848
2849             /* EWALD ELECTROSTATICS */
2850
2851             /* Analytical PME correction */
2852             zeta2            = _mm256_mul_ps(beta2,rsq12);
2853             rinv3            = _mm256_mul_ps(rinvsq12,rinv12);
2854             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2855             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2856             felec            = _mm256_mul_ps(qq12,felec);
2857             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2858             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2859             velec            = _mm256_sub_ps(rinv12,pmecorrV);
2860             velec            = _mm256_mul_ps(qq12,velec);
2861
2862             d                = _mm256_sub_ps(r12,rswitch);
2863             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2864             d2               = _mm256_mul_ps(d,d);
2865             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2866
2867             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2868
2869             /* Evaluate switch function */
2870             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2871             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv12,_mm256_mul_ps(velec,dsw)) );
2872             cutoff_mask      = _mm256_cmp_ps(rsq12,rcutoff2,_CMP_LT_OQ);
2873
2874             fscal            = felec;
2875
2876             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2877
2878             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
2879
2880             /* Calculate temporary vectorial force */
2881             tx               = _mm256_mul_ps(fscal,dx12);
2882             ty               = _mm256_mul_ps(fscal,dy12);
2883             tz               = _mm256_mul_ps(fscal,dz12);
2884
2885             /* Update vectorial force */
2886             fix1             = _mm256_add_ps(fix1,tx);
2887             fiy1             = _mm256_add_ps(fiy1,ty);
2888             fiz1             = _mm256_add_ps(fiz1,tz);
2889
2890             fjx2             = _mm256_add_ps(fjx2,tx);
2891             fjy2             = _mm256_add_ps(fjy2,ty);
2892             fjz2             = _mm256_add_ps(fjz2,tz);
2893
2894             }
2895
2896             /**************************
2897              * CALCULATE INTERACTIONS *
2898              **************************/
2899
2900             if (gmx_mm256_any_lt(rsq13,rcutoff2))
2901             {
2902
2903             r13              = _mm256_mul_ps(rsq13,rinv13);
2904             r13              = _mm256_andnot_ps(dummy_mask,r13);
2905
2906             /* EWALD ELECTROSTATICS */
2907
2908             /* Analytical PME correction */
2909             zeta2            = _mm256_mul_ps(beta2,rsq13);
2910             rinv3            = _mm256_mul_ps(rinvsq13,rinv13);
2911             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2912             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2913             felec            = _mm256_mul_ps(qq13,felec);
2914             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2915             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2916             velec            = _mm256_sub_ps(rinv13,pmecorrV);
2917             velec            = _mm256_mul_ps(qq13,velec);
2918
2919             d                = _mm256_sub_ps(r13,rswitch);
2920             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2921             d2               = _mm256_mul_ps(d,d);
2922             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2923
2924             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2925
2926             /* Evaluate switch function */
2927             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2928             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv13,_mm256_mul_ps(velec,dsw)) );
2929             cutoff_mask      = _mm256_cmp_ps(rsq13,rcutoff2,_CMP_LT_OQ);
2930
2931             fscal            = felec;
2932
2933             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2934
2935             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
2936
2937             /* Calculate temporary vectorial force */
2938             tx               = _mm256_mul_ps(fscal,dx13);
2939             ty               = _mm256_mul_ps(fscal,dy13);
2940             tz               = _mm256_mul_ps(fscal,dz13);
2941
2942             /* Update vectorial force */
2943             fix1             = _mm256_add_ps(fix1,tx);
2944             fiy1             = _mm256_add_ps(fiy1,ty);
2945             fiz1             = _mm256_add_ps(fiz1,tz);
2946
2947             fjx3             = _mm256_add_ps(fjx3,tx);
2948             fjy3             = _mm256_add_ps(fjy3,ty);
2949             fjz3             = _mm256_add_ps(fjz3,tz);
2950
2951             }
2952
2953             /**************************
2954              * CALCULATE INTERACTIONS *
2955              **************************/
2956
2957             if (gmx_mm256_any_lt(rsq21,rcutoff2))
2958             {
2959
2960             r21              = _mm256_mul_ps(rsq21,rinv21);
2961             r21              = _mm256_andnot_ps(dummy_mask,r21);
2962
2963             /* EWALD ELECTROSTATICS */
2964
2965             /* Analytical PME correction */
2966             zeta2            = _mm256_mul_ps(beta2,rsq21);
2967             rinv3            = _mm256_mul_ps(rinvsq21,rinv21);
2968             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
2969             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2970             felec            = _mm256_mul_ps(qq21,felec);
2971             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
2972             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
2973             velec            = _mm256_sub_ps(rinv21,pmecorrV);
2974             velec            = _mm256_mul_ps(qq21,velec);
2975
2976             d                = _mm256_sub_ps(r21,rswitch);
2977             d                = _mm256_max_ps(d,_mm256_setzero_ps());
2978             d2               = _mm256_mul_ps(d,d);
2979             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
2980
2981             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
2982
2983             /* Evaluate switch function */
2984             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2985             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv21,_mm256_mul_ps(velec,dsw)) );
2986             cutoff_mask      = _mm256_cmp_ps(rsq21,rcutoff2,_CMP_LT_OQ);
2987
2988             fscal            = felec;
2989
2990             fscal            = _mm256_and_ps(fscal,cutoff_mask);
2991
2992             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
2993
2994             /* Calculate temporary vectorial force */
2995             tx               = _mm256_mul_ps(fscal,dx21);
2996             ty               = _mm256_mul_ps(fscal,dy21);
2997             tz               = _mm256_mul_ps(fscal,dz21);
2998
2999             /* Update vectorial force */
3000             fix2             = _mm256_add_ps(fix2,tx);
3001             fiy2             = _mm256_add_ps(fiy2,ty);
3002             fiz2             = _mm256_add_ps(fiz2,tz);
3003
3004             fjx1             = _mm256_add_ps(fjx1,tx);
3005             fjy1             = _mm256_add_ps(fjy1,ty);
3006             fjz1             = _mm256_add_ps(fjz1,tz);
3007
3008             }
3009
3010             /**************************
3011              * CALCULATE INTERACTIONS *
3012              **************************/
3013
3014             if (gmx_mm256_any_lt(rsq22,rcutoff2))
3015             {
3016
3017             r22              = _mm256_mul_ps(rsq22,rinv22);
3018             r22              = _mm256_andnot_ps(dummy_mask,r22);
3019
3020             /* EWALD ELECTROSTATICS */
3021
3022             /* Analytical PME correction */
3023             zeta2            = _mm256_mul_ps(beta2,rsq22);
3024             rinv3            = _mm256_mul_ps(rinvsq22,rinv22);
3025             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
3026             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
3027             felec            = _mm256_mul_ps(qq22,felec);
3028             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
3029             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
3030             velec            = _mm256_sub_ps(rinv22,pmecorrV);
3031             velec            = _mm256_mul_ps(qq22,velec);
3032
3033             d                = _mm256_sub_ps(r22,rswitch);
3034             d                = _mm256_max_ps(d,_mm256_setzero_ps());
3035             d2               = _mm256_mul_ps(d,d);
3036             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
3037
3038             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
3039
3040             /* Evaluate switch function */
3041             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3042             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv22,_mm256_mul_ps(velec,dsw)) );
3043             cutoff_mask      = _mm256_cmp_ps(rsq22,rcutoff2,_CMP_LT_OQ);
3044
3045             fscal            = felec;
3046
3047             fscal            = _mm256_and_ps(fscal,cutoff_mask);
3048
3049             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
3050
3051             /* Calculate temporary vectorial force */
3052             tx               = _mm256_mul_ps(fscal,dx22);
3053             ty               = _mm256_mul_ps(fscal,dy22);
3054             tz               = _mm256_mul_ps(fscal,dz22);
3055
3056             /* Update vectorial force */
3057             fix2             = _mm256_add_ps(fix2,tx);
3058             fiy2             = _mm256_add_ps(fiy2,ty);
3059             fiz2             = _mm256_add_ps(fiz2,tz);
3060
3061             fjx2             = _mm256_add_ps(fjx2,tx);
3062             fjy2             = _mm256_add_ps(fjy2,ty);
3063             fjz2             = _mm256_add_ps(fjz2,tz);
3064
3065             }
3066
3067             /**************************
3068              * CALCULATE INTERACTIONS *
3069              **************************/
3070
3071             if (gmx_mm256_any_lt(rsq23,rcutoff2))
3072             {
3073
3074             r23              = _mm256_mul_ps(rsq23,rinv23);
3075             r23              = _mm256_andnot_ps(dummy_mask,r23);
3076
3077             /* EWALD ELECTROSTATICS */
3078
3079             /* Analytical PME correction */
3080             zeta2            = _mm256_mul_ps(beta2,rsq23);
3081             rinv3            = _mm256_mul_ps(rinvsq23,rinv23);
3082             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
3083             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
3084             felec            = _mm256_mul_ps(qq23,felec);
3085             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
3086             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
3087             velec            = _mm256_sub_ps(rinv23,pmecorrV);
3088             velec            = _mm256_mul_ps(qq23,velec);
3089
3090             d                = _mm256_sub_ps(r23,rswitch);
3091             d                = _mm256_max_ps(d,_mm256_setzero_ps());
3092             d2               = _mm256_mul_ps(d,d);
3093             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
3094
3095             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
3096
3097             /* Evaluate switch function */
3098             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3099             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv23,_mm256_mul_ps(velec,dsw)) );
3100             cutoff_mask      = _mm256_cmp_ps(rsq23,rcutoff2,_CMP_LT_OQ);
3101
3102             fscal            = felec;
3103
3104             fscal            = _mm256_and_ps(fscal,cutoff_mask);
3105
3106             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
3107
3108             /* Calculate temporary vectorial force */
3109             tx               = _mm256_mul_ps(fscal,dx23);
3110             ty               = _mm256_mul_ps(fscal,dy23);
3111             tz               = _mm256_mul_ps(fscal,dz23);
3112
3113             /* Update vectorial force */
3114             fix2             = _mm256_add_ps(fix2,tx);
3115             fiy2             = _mm256_add_ps(fiy2,ty);
3116             fiz2             = _mm256_add_ps(fiz2,tz);
3117
3118             fjx3             = _mm256_add_ps(fjx3,tx);
3119             fjy3             = _mm256_add_ps(fjy3,ty);
3120             fjz3             = _mm256_add_ps(fjz3,tz);
3121
3122             }
3123
3124             /**************************
3125              * CALCULATE INTERACTIONS *
3126              **************************/
3127
3128             if (gmx_mm256_any_lt(rsq31,rcutoff2))
3129             {
3130
3131             r31              = _mm256_mul_ps(rsq31,rinv31);
3132             r31              = _mm256_andnot_ps(dummy_mask,r31);
3133
3134             /* EWALD ELECTROSTATICS */
3135
3136             /* Analytical PME correction */
3137             zeta2            = _mm256_mul_ps(beta2,rsq31);
3138             rinv3            = _mm256_mul_ps(rinvsq31,rinv31);
3139             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
3140             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
3141             felec            = _mm256_mul_ps(qq31,felec);
3142             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
3143             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
3144             velec            = _mm256_sub_ps(rinv31,pmecorrV);
3145             velec            = _mm256_mul_ps(qq31,velec);
3146
3147             d                = _mm256_sub_ps(r31,rswitch);
3148             d                = _mm256_max_ps(d,_mm256_setzero_ps());
3149             d2               = _mm256_mul_ps(d,d);
3150             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
3151
3152             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
3153
3154             /* Evaluate switch function */
3155             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3156             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv31,_mm256_mul_ps(velec,dsw)) );
3157             cutoff_mask      = _mm256_cmp_ps(rsq31,rcutoff2,_CMP_LT_OQ);
3158
3159             fscal            = felec;
3160
3161             fscal            = _mm256_and_ps(fscal,cutoff_mask);
3162
3163             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
3164
3165             /* Calculate temporary vectorial force */
3166             tx               = _mm256_mul_ps(fscal,dx31);
3167             ty               = _mm256_mul_ps(fscal,dy31);
3168             tz               = _mm256_mul_ps(fscal,dz31);
3169
3170             /* Update vectorial force */
3171             fix3             = _mm256_add_ps(fix3,tx);
3172             fiy3             = _mm256_add_ps(fiy3,ty);
3173             fiz3             = _mm256_add_ps(fiz3,tz);
3174
3175             fjx1             = _mm256_add_ps(fjx1,tx);
3176             fjy1             = _mm256_add_ps(fjy1,ty);
3177             fjz1             = _mm256_add_ps(fjz1,tz);
3178
3179             }
3180
3181             /**************************
3182              * CALCULATE INTERACTIONS *
3183              **************************/
3184
3185             if (gmx_mm256_any_lt(rsq32,rcutoff2))
3186             {
3187
3188             r32              = _mm256_mul_ps(rsq32,rinv32);
3189             r32              = _mm256_andnot_ps(dummy_mask,r32);
3190
3191             /* EWALD ELECTROSTATICS */
3192
3193             /* Analytical PME correction */
3194             zeta2            = _mm256_mul_ps(beta2,rsq32);
3195             rinv3            = _mm256_mul_ps(rinvsq32,rinv32);
3196             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
3197             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
3198             felec            = _mm256_mul_ps(qq32,felec);
3199             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
3200             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
3201             velec            = _mm256_sub_ps(rinv32,pmecorrV);
3202             velec            = _mm256_mul_ps(qq32,velec);
3203
3204             d                = _mm256_sub_ps(r32,rswitch);
3205             d                = _mm256_max_ps(d,_mm256_setzero_ps());
3206             d2               = _mm256_mul_ps(d,d);
3207             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
3208
3209             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
3210
3211             /* Evaluate switch function */
3212             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3213             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv32,_mm256_mul_ps(velec,dsw)) );
3214             cutoff_mask      = _mm256_cmp_ps(rsq32,rcutoff2,_CMP_LT_OQ);
3215
3216             fscal            = felec;
3217
3218             fscal            = _mm256_and_ps(fscal,cutoff_mask);
3219
3220             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
3221
3222             /* Calculate temporary vectorial force */
3223             tx               = _mm256_mul_ps(fscal,dx32);
3224             ty               = _mm256_mul_ps(fscal,dy32);
3225             tz               = _mm256_mul_ps(fscal,dz32);
3226
3227             /* Update vectorial force */
3228             fix3             = _mm256_add_ps(fix3,tx);
3229             fiy3             = _mm256_add_ps(fiy3,ty);
3230             fiz3             = _mm256_add_ps(fiz3,tz);
3231
3232             fjx2             = _mm256_add_ps(fjx2,tx);
3233             fjy2             = _mm256_add_ps(fjy2,ty);
3234             fjz2             = _mm256_add_ps(fjz2,tz);
3235
3236             }
3237
3238             /**************************
3239              * CALCULATE INTERACTIONS *
3240              **************************/
3241
3242             if (gmx_mm256_any_lt(rsq33,rcutoff2))
3243             {
3244
3245             r33              = _mm256_mul_ps(rsq33,rinv33);
3246             r33              = _mm256_andnot_ps(dummy_mask,r33);
3247
3248             /* EWALD ELECTROSTATICS */
3249
3250             /* Analytical PME correction */
3251             zeta2            = _mm256_mul_ps(beta2,rsq33);
3252             rinv3            = _mm256_mul_ps(rinvsq33,rinv33);
3253             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
3254             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
3255             felec            = _mm256_mul_ps(qq33,felec);
3256             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
3257             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
3258             velec            = _mm256_sub_ps(rinv33,pmecorrV);
3259             velec            = _mm256_mul_ps(qq33,velec);
3260
3261             d                = _mm256_sub_ps(r33,rswitch);
3262             d                = _mm256_max_ps(d,_mm256_setzero_ps());
3263             d2               = _mm256_mul_ps(d,d);
3264             sw               = _mm256_add_ps(one,_mm256_mul_ps(d2,_mm256_mul_ps(d,_mm256_add_ps(swV3,_mm256_mul_ps(d,_mm256_add_ps(swV4,_mm256_mul_ps(d,swV5)))))));
3265
3266             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
3267
3268             /* Evaluate switch function */
3269             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3270             felec            = _mm256_sub_ps( _mm256_mul_ps(felec,sw) , _mm256_mul_ps(rinv33,_mm256_mul_ps(velec,dsw)) );
3271             cutoff_mask      = _mm256_cmp_ps(rsq33,rcutoff2,_CMP_LT_OQ);
3272
3273             fscal            = felec;
3274
3275             fscal            = _mm256_and_ps(fscal,cutoff_mask);
3276
3277             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
3278
3279             /* Calculate temporary vectorial force */
3280             tx               = _mm256_mul_ps(fscal,dx33);
3281             ty               = _mm256_mul_ps(fscal,dy33);
3282             tz               = _mm256_mul_ps(fscal,dz33);
3283
3284             /* Update vectorial force */
3285             fix3             = _mm256_add_ps(fix3,tx);
3286             fiy3             = _mm256_add_ps(fiy3,ty);
3287             fiz3             = _mm256_add_ps(fiz3,tz);
3288
3289             fjx3             = _mm256_add_ps(fjx3,tx);
3290             fjy3             = _mm256_add_ps(fjy3,ty);
3291             fjz3             = _mm256_add_ps(fjz3,tz);
3292
3293             }
3294
3295             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
3296             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
3297             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
3298             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
3299             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
3300             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
3301             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
3302             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
3303
3304             gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
3305                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
3306                                                       fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
3307
3308             /* Inner loop uses 1014 flops */
3309         }
3310
3311         /* End of innermost loop */
3312
3313         gmx_mm256_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
3314                                                  f+i_coord_offset,fshift+i_shift_offset);
3315
3316         /* Increment number of inner iterations */
3317         inneriter                  += j_index_end - j_index_start;
3318
3319         /* Outer loop uses 24 flops */
3320     }
3321
3322     /* Increment number of outer iterations */
3323     outeriter        += nri;
3324
3325     /* Update outer/inner flops */
3326
3327     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*1014);
3328 }