src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecRFCut_VdwLJSw_GeomP1P1_avx_256_single.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_256_single kernel generator.
  37  */
  38 #include "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/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_ElecRFCut_VdwLJSw_GeomP1P1_VF_avx_256_single
  54  * Electrostatics interaction: ReactionField
  55  * VdW interaction:            LennardJones
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecRFCut_VdwLJSw_GeomP1P1_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     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
  91     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  92     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  93     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
  94     real             *charge;
  95     int              nvdwtype;
  96     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  97     int              *vdwtype;
  98     real             *vdwparam;
  99     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 100     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 101     __m256           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 102     real             rswitch_scalar,d_scalar;
 103     __m256           dummy_mask,cutoff_mask;
 104     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 105     __m256           one     = _mm256_set1_ps(1.0);
 106     __m256           two     = _mm256_set1_ps(2.0);
 107     x                = xx[0];
 108     f                = ff[0];
 109
 110     nri              = nlist->nri;
 111     iinr             = nlist->iinr;
 112     jindex           = nlist->jindex;
 113     jjnr             = nlist->jjnr;
 114     shiftidx         = nlist->shift;
 115     gid              = nlist->gid;
 116     shiftvec         = fr->shift_vec[0];
 117     fshift           = fr->fshift[0];
 118     facel            = _mm256_set1_ps(fr->epsfac);
 119     charge           = mdatoms->chargeA;
 120     krf              = _mm256_set1_ps(fr->ic->k_rf);
 121     krf2             = _mm256_set1_ps(fr->ic->k_rf*2.0);
 122     crf              = _mm256_set1_ps(fr->ic->c_rf);
 123     nvdwtype         = fr->ntype;
 124     vdwparam         = fr->nbfp;
 125     vdwtype          = mdatoms->typeA;
 126
 127     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 128     rcutoff_scalar   = fr->rcoulomb;
 129     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
 130     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
 131
 132     rswitch_scalar   = fr->rvdw_switch;
 133     rswitch          = _mm256_set1_ps(rswitch_scalar);
 134     /* Setup switch parameters */
 135     d_scalar         = rcutoff_scalar-rswitch_scalar;
 136     d                = _mm256_set1_ps(d_scalar);
 137     swV3             = _mm256_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
 138     swV4             = _mm256_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 139     swV5             = _mm256_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 140     swF2             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
 141     swF3             = _mm256_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 142     swF4             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 143
 144     /* Avoid stupid compiler warnings */
 145     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 146     j_coord_offsetA = 0;
 147     j_coord_offsetB = 0;
 148     j_coord_offsetC = 0;
 149     j_coord_offsetD = 0;
 150     j_coord_offsetE = 0;
 151     j_coord_offsetF = 0;
 152     j_coord_offsetG = 0;
 153     j_coord_offsetH = 0;
 154
 155     outeriter        = 0;
 156     inneriter        = 0;
 157
 158     for(iidx=0;iidx<4*DIM;iidx++)
 159     {
 160         scratch[iidx] = 0.0;
 161     }
 162
 163     /* Start outer loop over neighborlists */
 164     for(iidx=0; iidx<nri; iidx++)
 165     {
 166         /* Load shift vector for this list */
 167         i_shift_offset   = DIM*shiftidx[iidx];
 168
 169         /* Load limits for loop over neighbors */
 170         j_index_start    = jindex[iidx];
 171         j_index_end      = jindex[iidx+1];
 172
 173         /* Get outer coordinate index */
 174         inr              = iinr[iidx];
 175         i_coord_offset   = DIM*inr;
 176
 177         /* Load i particle coords and add shift vector */
 178         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 179
 180         fix0             = _mm256_setzero_ps();
 181         fiy0             = _mm256_setzero_ps();
 182         fiz0             = _mm256_setzero_ps();
 183
 184         /* Load parameters for i particles */
 185         iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 186         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 187
 188         /* Reset potential sums */
 189         velecsum         = _mm256_setzero_ps();
 190         vvdwsum          = _mm256_setzero_ps();
 191
 192         /* Start inner kernel loop */
 193         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 194         {
 195
 196             /* Get j neighbor index, and coordinate index */
 197             jnrA             = jjnr[jidx];
 198             jnrB             = jjnr[jidx+1];
 199             jnrC             = jjnr[jidx+2];
 200             jnrD             = jjnr[jidx+3];
 201             jnrE             = jjnr[jidx+4];
 202             jnrF             = jjnr[jidx+5];
 203             jnrG             = jjnr[jidx+6];
 204             jnrH             = jjnr[jidx+7];
 205             j_coord_offsetA  = DIM*jnrA;
 206             j_coord_offsetB  = DIM*jnrB;
 207             j_coord_offsetC  = DIM*jnrC;
 208             j_coord_offsetD  = DIM*jnrD;
 209             j_coord_offsetE  = DIM*jnrE;
 210             j_coord_offsetF  = DIM*jnrF;
 211             j_coord_offsetG  = DIM*jnrG;
 212             j_coord_offsetH  = DIM*jnrH;
 213
 214             /* load j atom coordinates */
 215             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 216                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 217                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 218                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 219                                                  &jx0,&jy0,&jz0);
 220
 221             /* Calculate displacement vector */
 222             dx00             = _mm256_sub_ps(ix0,jx0);
 223             dy00             = _mm256_sub_ps(iy0,jy0);
 224             dz00             = _mm256_sub_ps(iz0,jz0);
 225
 226             /* Calculate squared distance and things based on it */
 227             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 228
 229             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 230
 231             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 232
 233             /* Load parameters for j particles */
 234             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 235                                                                  charge+jnrC+0,charge+jnrD+0,
 236                                                                  charge+jnrE+0,charge+jnrF+0,
 237                                                                  charge+jnrG+0,charge+jnrH+0);
 238             vdwjidx0A        = 2*vdwtype[jnrA+0];
 239             vdwjidx0B        = 2*vdwtype[jnrB+0];
 240             vdwjidx0C        = 2*vdwtype[jnrC+0];
 241             vdwjidx0D        = 2*vdwtype[jnrD+0];
 242             vdwjidx0E        = 2*vdwtype[jnrE+0];
 243             vdwjidx0F        = 2*vdwtype[jnrF+0];
 244             vdwjidx0G        = 2*vdwtype[jnrG+0];
 245             vdwjidx0H        = 2*vdwtype[jnrH+0];
 246
 247             /**************************
 248              * CALCULATE INTERACTIONS *
 249              **************************/
 250
 251             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 252             {
 253
 254             r00              = _mm256_mul_ps(rsq00,rinv00);
 255
 256             /* Compute parameters for interactions between i and j atoms */
 257             qq00             = _mm256_mul_ps(iq0,jq0);
 258             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 259                                             vdwioffsetptr0+vdwjidx0B,
 260                                             vdwioffsetptr0+vdwjidx0C,
 261                                             vdwioffsetptr0+vdwjidx0D,
 262                                             vdwioffsetptr0+vdwjidx0E,
 263                                             vdwioffsetptr0+vdwjidx0F,
 264                                             vdwioffsetptr0+vdwjidx0G,
 265                                             vdwioffsetptr0+vdwjidx0H,
 266                                             &c6_00,&c12_00);
 267
 268             /* REACTION-FIELD ELECTROSTATICS */
 269             velec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_add_ps(rinv00,_mm256_mul_ps(krf,rsq00)),crf));
 270             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 271
 272             /* LENNARD-JONES DISPERSION/REPULSION */
 273
 274             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 275             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 276             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 277             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 278             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 279
 280             d                = _mm256_sub_ps(r00,rswitch);
 281             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 282             d2               = _mm256_mul_ps(d,d);
 283             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)))))));
 284
 285             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 286
 287             /* Evaluate switch function */
 288             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 289             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
 290             vvdw             = _mm256_mul_ps(vvdw,sw);
 291             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 292
 293             /* Update potential sum for this i atom from the interaction with this j atom. */
 294             velec            = _mm256_and_ps(velec,cutoff_mask);
 295             velecsum         = _mm256_add_ps(velecsum,velec);
 296             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
 297             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 298
 299             fscal            = _mm256_add_ps(felec,fvdw);
 300
 301             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 302
 303             /* Calculate temporary vectorial force */
 304             tx               = _mm256_mul_ps(fscal,dx00);
 305             ty               = _mm256_mul_ps(fscal,dy00);
 306             tz               = _mm256_mul_ps(fscal,dz00);
 307
 308             /* Update vectorial force */
 309             fix0             = _mm256_add_ps(fix0,tx);
 310             fiy0             = _mm256_add_ps(fiy0,ty);
 311             fiz0             = _mm256_add_ps(fiz0,tz);
 312
 313             fjptrA             = f+j_coord_offsetA;
 314             fjptrB             = f+j_coord_offsetB;
 315             fjptrC             = f+j_coord_offsetC;
 316             fjptrD             = f+j_coord_offsetD;
 317             fjptrE             = f+j_coord_offsetE;
 318             fjptrF             = f+j_coord_offsetF;
 319             fjptrG             = f+j_coord_offsetG;
 320             fjptrH             = f+j_coord_offsetH;
 321             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 322
 323             }
 324
 325             /* Inner loop uses 70 flops */
 326         }
 327
 328         if(jidx<j_index_end)
 329         {
 330
 331             /* Get j neighbor index, and coordinate index */
 332             jnrlistA         = jjnr[jidx];
 333             jnrlistB         = jjnr[jidx+1];
 334             jnrlistC         = jjnr[jidx+2];
 335             jnrlistD         = jjnr[jidx+3];
 336             jnrlistE         = jjnr[jidx+4];
 337             jnrlistF         = jjnr[jidx+5];
 338             jnrlistG         = jjnr[jidx+6];
 339             jnrlistH         = jjnr[jidx+7];
 340             /* Sign of each element will be negative for non-real atoms.
 341              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 342              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 343              */
 344             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 345                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 346
 347             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 348             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 349             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 350             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 351             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 352             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 353             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 354             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 355             j_coord_offsetA  = DIM*jnrA;
 356             j_coord_offsetB  = DIM*jnrB;
 357             j_coord_offsetC  = DIM*jnrC;
 358             j_coord_offsetD  = DIM*jnrD;
 359             j_coord_offsetE  = DIM*jnrE;
 360             j_coord_offsetF  = DIM*jnrF;
 361             j_coord_offsetG  = DIM*jnrG;
 362             j_coord_offsetH  = DIM*jnrH;
 363
 364             /* load j atom coordinates */
 365             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 366                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 367                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 368                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 369                                                  &jx0,&jy0,&jz0);
 370
 371             /* Calculate displacement vector */
 372             dx00             = _mm256_sub_ps(ix0,jx0);
 373             dy00             = _mm256_sub_ps(iy0,jy0);
 374             dz00             = _mm256_sub_ps(iz0,jz0);
 375
 376             /* Calculate squared distance and things based on it */
 377             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 378
 379             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 380
 381             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 382
 383             /* Load parameters for j particles */
 384             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 385                                                                  charge+jnrC+0,charge+jnrD+0,
 386                                                                  charge+jnrE+0,charge+jnrF+0,
 387                                                                  charge+jnrG+0,charge+jnrH+0);
 388             vdwjidx0A        = 2*vdwtype[jnrA+0];
 389             vdwjidx0B        = 2*vdwtype[jnrB+0];
 390             vdwjidx0C        = 2*vdwtype[jnrC+0];
 391             vdwjidx0D        = 2*vdwtype[jnrD+0];
 392             vdwjidx0E        = 2*vdwtype[jnrE+0];
 393             vdwjidx0F        = 2*vdwtype[jnrF+0];
 394             vdwjidx0G        = 2*vdwtype[jnrG+0];
 395             vdwjidx0H        = 2*vdwtype[jnrH+0];
 396
 397             /**************************
 398              * CALCULATE INTERACTIONS *
 399              **************************/
 400
 401             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 402             {
 403
 404             r00              = _mm256_mul_ps(rsq00,rinv00);
 405             r00              = _mm256_andnot_ps(dummy_mask,r00);
 406
 407             /* Compute parameters for interactions between i and j atoms */
 408             qq00             = _mm256_mul_ps(iq0,jq0);
 409             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 410                                             vdwioffsetptr0+vdwjidx0B,
 411                                             vdwioffsetptr0+vdwjidx0C,
 412                                             vdwioffsetptr0+vdwjidx0D,
 413                                             vdwioffsetptr0+vdwjidx0E,
 414                                             vdwioffsetptr0+vdwjidx0F,
 415                                             vdwioffsetptr0+vdwjidx0G,
 416                                             vdwioffsetptr0+vdwjidx0H,
 417                                             &c6_00,&c12_00);
 418
 419             /* REACTION-FIELD ELECTROSTATICS */
 420             velec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_add_ps(rinv00,_mm256_mul_ps(krf,rsq00)),crf));
 421             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 422
 423             /* LENNARD-JONES DISPERSION/REPULSION */
 424
 425             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 426             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 427             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 428             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 429             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 430
 431             d                = _mm256_sub_ps(r00,rswitch);
 432             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 433             d2               = _mm256_mul_ps(d,d);
 434             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)))))));
 435
 436             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 437
 438             /* Evaluate switch function */
 439             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 440             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
 441             vvdw             = _mm256_mul_ps(vvdw,sw);
 442             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 443
 444             /* Update potential sum for this i atom from the interaction with this j atom. */
 445             velec            = _mm256_and_ps(velec,cutoff_mask);
 446             velec            = _mm256_andnot_ps(dummy_mask,velec);
 447             velecsum         = _mm256_add_ps(velecsum,velec);
 448             vvdw             = _mm256_and_ps(vvdw,cutoff_mask);
 449             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
 450             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 451
 452             fscal            = _mm256_add_ps(felec,fvdw);
 453
 454             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 455
 456             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 457
 458             /* Calculate temporary vectorial force */
 459             tx               = _mm256_mul_ps(fscal,dx00);
 460             ty               = _mm256_mul_ps(fscal,dy00);
 461             tz               = _mm256_mul_ps(fscal,dz00);
 462
 463             /* Update vectorial force */
 464             fix0             = _mm256_add_ps(fix0,tx);
 465             fiy0             = _mm256_add_ps(fiy0,ty);
 466             fiz0             = _mm256_add_ps(fiz0,tz);
 467
 468             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 469             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 470             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 471             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 472             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 473             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 474             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 475             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 476             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 477
 478             }
 479
 480             /* Inner loop uses 71 flops */
 481         }
 482
 483         /* End of innermost loop */
 484
 485         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 486                                                  f+i_coord_offset,fshift+i_shift_offset);
 487
 488         ggid                        = gid[iidx];
 489         /* Update potential energies */
 490         gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 491         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 492
 493         /* Increment number of inner iterations */
 494         inneriter                  += j_index_end - j_index_start;
 495
 496         /* Outer loop uses 9 flops */
 497     }
 498
 499     /* Increment number of outer iterations */
 500     outeriter        += nri;
 501
 502     /* Update outer/inner flops */
 503
 504     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*71);
 505 }
 506 /*
 507  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwLJSw_GeomP1P1_F_avx_256_single
 508  * Electrostatics interaction: ReactionField
 509  * VdW interaction:            LennardJones
 510  * Geometry:                   Particle-Particle
 511  * Calculate force/pot:        Force
 512  */
 513 void
 514 nb_kernel_ElecRFCut_VdwLJSw_GeomP1P1_F_avx_256_single
 515                     (t_nblist                    * gmx_restrict       nlist,
 516                      rvec                        * gmx_restrict          xx,
 517                      rvec                        * gmx_restrict          ff,
 518                      t_forcerec                  * gmx_restrict          fr,
 519                      t_mdatoms                   * gmx_restrict     mdatoms,
 520                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 521                      t_nrnb                      * gmx_restrict        nrnb)
 522 {
 523     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 524      * just 0 for non-waters.
 525      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
 526      * jnr indices corresponding to data put in the four positions in the SIMD register.
 527      */
 528     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 529     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 530     int              jnrA,jnrB,jnrC,jnrD;
 531     int              jnrE,jnrF,jnrG,jnrH;
 532     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 533     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 534     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 535     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
 536     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 537     real             rcutoff_scalar;
 538     real             *shiftvec,*fshift,*x,*f;
 539     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
 540     real             scratch[4*DIM];
 541     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 542     real *           vdwioffsetptr0;
 543     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 544     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
 545     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 546     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 547     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
 548     real             *charge;
 549     int              nvdwtype;
 550     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 551     int              *vdwtype;
 552     real             *vdwparam;
 553     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 554     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 555     __m256           rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 556     real             rswitch_scalar,d_scalar;
 557     __m256           dummy_mask,cutoff_mask;
 558     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 559     __m256           one     = _mm256_set1_ps(1.0);
 560     __m256           two     = _mm256_set1_ps(2.0);
 561     x                = xx[0];
 562     f                = ff[0];
 563
 564     nri              = nlist->nri;
 565     iinr             = nlist->iinr;
 566     jindex           = nlist->jindex;
 567     jjnr             = nlist->jjnr;
 568     shiftidx         = nlist->shift;
 569     gid              = nlist->gid;
 570     shiftvec         = fr->shift_vec[0];
 571     fshift           = fr->fshift[0];
 572     facel            = _mm256_set1_ps(fr->epsfac);
 573     charge           = mdatoms->chargeA;
 574     krf              = _mm256_set1_ps(fr->ic->k_rf);
 575     krf2             = _mm256_set1_ps(fr->ic->k_rf*2.0);
 576     crf              = _mm256_set1_ps(fr->ic->c_rf);
 577     nvdwtype         = fr->ntype;
 578     vdwparam         = fr->nbfp;
 579     vdwtype          = mdatoms->typeA;
 580
 581     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 582     rcutoff_scalar   = fr->rcoulomb;
 583     rcutoff          = _mm256_set1_ps(rcutoff_scalar);
 584     rcutoff2         = _mm256_mul_ps(rcutoff,rcutoff);
 585
 586     rswitch_scalar   = fr->rvdw_switch;
 587     rswitch          = _mm256_set1_ps(rswitch_scalar);
 588     /* Setup switch parameters */
 589     d_scalar         = rcutoff_scalar-rswitch_scalar;
 590     d                = _mm256_set1_ps(d_scalar);
 591     swV3             = _mm256_set1_ps(-10.0/(d_scalar*d_scalar*d_scalar));
 592     swV4             = _mm256_set1_ps( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 593     swV5             = _mm256_set1_ps( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 594     swF2             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar));
 595     swF3             = _mm256_set1_ps( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
 596     swF4             = _mm256_set1_ps(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
 597
 598     /* Avoid stupid compiler warnings */
 599     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 600     j_coord_offsetA = 0;
 601     j_coord_offsetB = 0;
 602     j_coord_offsetC = 0;
 603     j_coord_offsetD = 0;
 604     j_coord_offsetE = 0;
 605     j_coord_offsetF = 0;
 606     j_coord_offsetG = 0;
 607     j_coord_offsetH = 0;
 608
 609     outeriter        = 0;
 610     inneriter        = 0;
 611
 612     for(iidx=0;iidx<4*DIM;iidx++)
 613     {
 614         scratch[iidx] = 0.0;
 615     }
 616
 617     /* Start outer loop over neighborlists */
 618     for(iidx=0; iidx<nri; iidx++)
 619     {
 620         /* Load shift vector for this list */
 621         i_shift_offset   = DIM*shiftidx[iidx];
 622
 623         /* Load limits for loop over neighbors */
 624         j_index_start    = jindex[iidx];
 625         j_index_end      = jindex[iidx+1];
 626
 627         /* Get outer coordinate index */
 628         inr              = iinr[iidx];
 629         i_coord_offset   = DIM*inr;
 630
 631         /* Load i particle coords and add shift vector */
 632         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 633
 634         fix0             = _mm256_setzero_ps();
 635         fiy0             = _mm256_setzero_ps();
 636         fiz0             = _mm256_setzero_ps();
 637
 638         /* Load parameters for i particles */
 639         iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 640         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 641
 642         /* Start inner kernel loop */
 643         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 644         {
 645
 646             /* Get j neighbor index, and coordinate index */
 647             jnrA             = jjnr[jidx];
 648             jnrB             = jjnr[jidx+1];
 649             jnrC             = jjnr[jidx+2];
 650             jnrD             = jjnr[jidx+3];
 651             jnrE             = jjnr[jidx+4];
 652             jnrF             = jjnr[jidx+5];
 653             jnrG             = jjnr[jidx+6];
 654             jnrH             = jjnr[jidx+7];
 655             j_coord_offsetA  = DIM*jnrA;
 656             j_coord_offsetB  = DIM*jnrB;
 657             j_coord_offsetC  = DIM*jnrC;
 658             j_coord_offsetD  = DIM*jnrD;
 659             j_coord_offsetE  = DIM*jnrE;
 660             j_coord_offsetF  = DIM*jnrF;
 661             j_coord_offsetG  = DIM*jnrG;
 662             j_coord_offsetH  = DIM*jnrH;
 663
 664             /* load j atom coordinates */
 665             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 666                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 667                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 668                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 669                                                  &jx0,&jy0,&jz0);
 670
 671             /* Calculate displacement vector */
 672             dx00             = _mm256_sub_ps(ix0,jx0);
 673             dy00             = _mm256_sub_ps(iy0,jy0);
 674             dz00             = _mm256_sub_ps(iz0,jz0);
 675
 676             /* Calculate squared distance and things based on it */
 677             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 678
 679             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 680
 681             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 682
 683             /* Load parameters for j particles */
 684             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 685                                                                  charge+jnrC+0,charge+jnrD+0,
 686                                                                  charge+jnrE+0,charge+jnrF+0,
 687                                                                  charge+jnrG+0,charge+jnrH+0);
 688             vdwjidx0A        = 2*vdwtype[jnrA+0];
 689             vdwjidx0B        = 2*vdwtype[jnrB+0];
 690             vdwjidx0C        = 2*vdwtype[jnrC+0];
 691             vdwjidx0D        = 2*vdwtype[jnrD+0];
 692             vdwjidx0E        = 2*vdwtype[jnrE+0];
 693             vdwjidx0F        = 2*vdwtype[jnrF+0];
 694             vdwjidx0G        = 2*vdwtype[jnrG+0];
 695             vdwjidx0H        = 2*vdwtype[jnrH+0];
 696
 697             /**************************
 698              * CALCULATE INTERACTIONS *
 699              **************************/
 700
 701             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 702             {
 703
 704             r00              = _mm256_mul_ps(rsq00,rinv00);
 705
 706             /* Compute parameters for interactions between i and j atoms */
 707             qq00             = _mm256_mul_ps(iq0,jq0);
 708             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 709                                             vdwioffsetptr0+vdwjidx0B,
 710                                             vdwioffsetptr0+vdwjidx0C,
 711                                             vdwioffsetptr0+vdwjidx0D,
 712                                             vdwioffsetptr0+vdwjidx0E,
 713                                             vdwioffsetptr0+vdwjidx0F,
 714                                             vdwioffsetptr0+vdwjidx0G,
 715                                             vdwioffsetptr0+vdwjidx0H,
 716                                             &c6_00,&c12_00);
 717
 718             /* REACTION-FIELD ELECTROSTATICS */
 719             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 720
 721             /* LENNARD-JONES DISPERSION/REPULSION */
 722
 723             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 724             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 725             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 726             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 727             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 728
 729             d                = _mm256_sub_ps(r00,rswitch);
 730             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 731             d2               = _mm256_mul_ps(d,d);
 732             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)))))));
 733
 734             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 735
 736             /* Evaluate switch function */
 737             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 738             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
 739             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 740
 741             fscal            = _mm256_add_ps(felec,fvdw);
 742
 743             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 744
 745             /* Calculate temporary vectorial force */
 746             tx               = _mm256_mul_ps(fscal,dx00);
 747             ty               = _mm256_mul_ps(fscal,dy00);
 748             tz               = _mm256_mul_ps(fscal,dz00);
 749
 750             /* Update vectorial force */
 751             fix0             = _mm256_add_ps(fix0,tx);
 752             fiy0             = _mm256_add_ps(fiy0,ty);
 753             fiz0             = _mm256_add_ps(fiz0,tz);
 754
 755             fjptrA             = f+j_coord_offsetA;
 756             fjptrB             = f+j_coord_offsetB;
 757             fjptrC             = f+j_coord_offsetC;
 758             fjptrD             = f+j_coord_offsetD;
 759             fjptrE             = f+j_coord_offsetE;
 760             fjptrF             = f+j_coord_offsetF;
 761             fjptrG             = f+j_coord_offsetG;
 762             fjptrH             = f+j_coord_offsetH;
 763             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 764
 765             }
 766
 767             /* Inner loop uses 61 flops */
 768         }
 769
 770         if(jidx<j_index_end)
 771         {
 772
 773             /* Get j neighbor index, and coordinate index */
 774             jnrlistA         = jjnr[jidx];
 775             jnrlistB         = jjnr[jidx+1];
 776             jnrlistC         = jjnr[jidx+2];
 777             jnrlistD         = jjnr[jidx+3];
 778             jnrlistE         = jjnr[jidx+4];
 779             jnrlistF         = jjnr[jidx+5];
 780             jnrlistG         = jjnr[jidx+6];
 781             jnrlistH         = jjnr[jidx+7];
 782             /* Sign of each element will be negative for non-real atoms.
 783              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 784              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 785              */
 786             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 787                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 788
 789             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 790             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 791             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 792             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 793             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 794             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 795             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 796             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 797             j_coord_offsetA  = DIM*jnrA;
 798             j_coord_offsetB  = DIM*jnrB;
 799             j_coord_offsetC  = DIM*jnrC;
 800             j_coord_offsetD  = DIM*jnrD;
 801             j_coord_offsetE  = DIM*jnrE;
 802             j_coord_offsetF  = DIM*jnrF;
 803             j_coord_offsetG  = DIM*jnrG;
 804             j_coord_offsetH  = DIM*jnrH;
 805
 806             /* load j atom coordinates */
 807             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 808                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 809                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 810                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 811                                                  &jx0,&jy0,&jz0);
 812
 813             /* Calculate displacement vector */
 814             dx00             = _mm256_sub_ps(ix0,jx0);
 815             dy00             = _mm256_sub_ps(iy0,jy0);
 816             dz00             = _mm256_sub_ps(iz0,jz0);
 817
 818             /* Calculate squared distance and things based on it */
 819             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 820
 821             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 822
 823             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 824
 825             /* Load parameters for j particles */
 826             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 827                                                                  charge+jnrC+0,charge+jnrD+0,
 828                                                                  charge+jnrE+0,charge+jnrF+0,
 829                                                                  charge+jnrG+0,charge+jnrH+0);
 830             vdwjidx0A        = 2*vdwtype[jnrA+0];
 831             vdwjidx0B        = 2*vdwtype[jnrB+0];
 832             vdwjidx0C        = 2*vdwtype[jnrC+0];
 833             vdwjidx0D        = 2*vdwtype[jnrD+0];
 834             vdwjidx0E        = 2*vdwtype[jnrE+0];
 835             vdwjidx0F        = 2*vdwtype[jnrF+0];
 836             vdwjidx0G        = 2*vdwtype[jnrG+0];
 837             vdwjidx0H        = 2*vdwtype[jnrH+0];
 838
 839             /**************************
 840              * CALCULATE INTERACTIONS *
 841              **************************/
 842
 843             if (gmx_mm256_any_lt(rsq00,rcutoff2))
 844             {
 845
 846             r00              = _mm256_mul_ps(rsq00,rinv00);
 847             r00              = _mm256_andnot_ps(dummy_mask,r00);
 848
 849             /* Compute parameters for interactions between i and j atoms */
 850             qq00             = _mm256_mul_ps(iq0,jq0);
 851             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 852                                             vdwioffsetptr0+vdwjidx0B,
 853                                             vdwioffsetptr0+vdwjidx0C,
 854                                             vdwioffsetptr0+vdwjidx0D,
 855                                             vdwioffsetptr0+vdwjidx0E,
 856                                             vdwioffsetptr0+vdwjidx0F,
 857                                             vdwioffsetptr0+vdwjidx0G,
 858                                             vdwioffsetptr0+vdwjidx0H,
 859                                             &c6_00,&c12_00);
 860
 861             /* REACTION-FIELD ELECTROSTATICS */
 862             felec            = _mm256_mul_ps(qq00,_mm256_sub_ps(_mm256_mul_ps(rinv00,rinvsq00),krf2));
 863
 864             /* LENNARD-JONES DISPERSION/REPULSION */
 865
 866             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 867             vvdw6            = _mm256_mul_ps(c6_00,rinvsix);
 868             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 869             vvdw             = _mm256_sub_ps( _mm256_mul_ps(vvdw12,one_twelfth) , _mm256_mul_ps(vvdw6,one_sixth) );
 870             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,vvdw6),rinvsq00);
 871
 872             d                = _mm256_sub_ps(r00,rswitch);
 873             d                = _mm256_max_ps(d,_mm256_setzero_ps());
 874             d2               = _mm256_mul_ps(d,d);
 875             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)))))));
 876
 877             dsw              = _mm256_mul_ps(d2,_mm256_add_ps(swF2,_mm256_mul_ps(d,_mm256_add_ps(swF3,_mm256_mul_ps(d,swF4)))));
 878
 879             /* Evaluate switch function */
 880             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 881             fvdw             = _mm256_sub_ps( _mm256_mul_ps(fvdw,sw) , _mm256_mul_ps(rinv00,_mm256_mul_ps(vvdw,dsw)) );
 882             cutoff_mask      = _mm256_cmp_ps(rsq00,rcutoff2,_CMP_LT_OQ);
 883
 884             fscal            = _mm256_add_ps(felec,fvdw);
 885
 886             fscal            = _mm256_and_ps(fscal,cutoff_mask);
 887
 888             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 889
 890             /* Calculate temporary vectorial force */
 891             tx               = _mm256_mul_ps(fscal,dx00);
 892             ty               = _mm256_mul_ps(fscal,dy00);
 893             tz               = _mm256_mul_ps(fscal,dz00);
 894
 895             /* Update vectorial force */
 896             fix0             = _mm256_add_ps(fix0,tx);
 897             fiy0             = _mm256_add_ps(fiy0,ty);
 898             fiz0             = _mm256_add_ps(fiz0,tz);
 899
 900             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 901             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 902             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 903             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 904             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 905             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 906             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 907             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 908             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 909
 910             }
 911
 912             /* Inner loop uses 62 flops */
 913         }
 914
 915         /* End of innermost loop */
 916
 917         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 918                                                  f+i_coord_offset,fshift+i_shift_offset);
 919
 920         /* Increment number of inner iterations */
 921         inneriter                  += j_index_end - j_index_start;
 922
 923         /* Outer loop uses 7 flops */
 924     }
 925
 926     /* Increment number of outer iterations */
 927     outeriter        += nri;
 928
 929     /* Update outer/inner flops */
 930
 931     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*62);
 932 }