src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecEw_VdwLJEw_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 #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_ElecEw_VdwLJEw_GeomP1P1_VF_avx_256_single
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            LJEwald
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEw_VdwLJEw_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     real *           vdwgridioffsetptr0;
  90     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  91     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
  92     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  93     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  94     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     int              nvdwtype;
  97     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  98     int              *vdwtype;
  99     real             *vdwparam;
 100     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 101     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 102     __m256           c6grid_00;
 103     real             *vdwgridparam;
 104     __m256           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 105     __m256           one_half  = _mm256_set1_ps(0.5);
 106     __m256           minus_one = _mm256_set1_ps(-1.0);
 107     __m256i          ewitab;
 108     __m128i          ewitab_lo,ewitab_hi;
 109     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 110     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 111     real             *ewtab;
 112     __m256           dummy_mask,cutoff_mask;
 113     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 114     __m256           one     = _mm256_set1_ps(1.0);
 115     __m256           two     = _mm256_set1_ps(2.0);
 116     x                = xx[0];
 117     f                = ff[0];
 118
 119     nri              = nlist->nri;
 120     iinr             = nlist->iinr;
 121     jindex           = nlist->jindex;
 122     jjnr             = nlist->jjnr;
 123     shiftidx         = nlist->shift;
 124     gid              = nlist->gid;
 125     shiftvec         = fr->shift_vec[0];
 126     fshift           = fr->fshift[0];
 127     facel            = _mm256_set1_ps(fr->epsfac);
 128     charge           = mdatoms->chargeA;
 129     nvdwtype         = fr->ntype;
 130     vdwparam         = fr->nbfp;
 131     vdwtype          = mdatoms->typeA;
 132     vdwgridparam     = fr->ljpme_c6grid;
 133     sh_lj_ewald      = _mm256_set1_ps(fr->ic->sh_lj_ewald);
 134     ewclj            = _mm256_set1_ps(fr->ewaldcoeff_lj);
 135     ewclj2           = _mm256_mul_ps(minus_one,_mm256_mul_ps(ewclj,ewclj));
 136
 137     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
 138     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
 139     beta2            = _mm256_mul_ps(beta,beta);
 140     beta3            = _mm256_mul_ps(beta,beta2);
 141
 142     ewtab            = fr->ic->tabq_coul_FDV0;
 143     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
 144     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
 145
 146     /* Avoid stupid compiler warnings */
 147     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 148     j_coord_offsetA = 0;
 149     j_coord_offsetB = 0;
 150     j_coord_offsetC = 0;
 151     j_coord_offsetD = 0;
 152     j_coord_offsetE = 0;
 153     j_coord_offsetF = 0;
 154     j_coord_offsetG = 0;
 155     j_coord_offsetH = 0;
 156
 157     outeriter        = 0;
 158     inneriter        = 0;
 159
 160     for(iidx=0;iidx<4*DIM;iidx++)
 161     {
 162         scratch[iidx] = 0.0;
 163     }
 164
 165     /* Start outer loop over neighborlists */
 166     for(iidx=0; iidx<nri; iidx++)
 167     {
 168         /* Load shift vector for this list */
 169         i_shift_offset   = DIM*shiftidx[iidx];
 170
 171         /* Load limits for loop over neighbors */
 172         j_index_start    = jindex[iidx];
 173         j_index_end      = jindex[iidx+1];
 174
 175         /* Get outer coordinate index */
 176         inr              = iinr[iidx];
 177         i_coord_offset   = DIM*inr;
 178
 179         /* Load i particle coords and add shift vector */
 180         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 181
 182         fix0             = _mm256_setzero_ps();
 183         fiy0             = _mm256_setzero_ps();
 184         fiz0             = _mm256_setzero_ps();
 185
 186         /* Load parameters for i particles */
 187         iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 188         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 189         vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
 190
 191         /* Reset potential sums */
 192         velecsum         = _mm256_setzero_ps();
 193         vvdwsum          = _mm256_setzero_ps();
 194
 195         /* Start inner kernel loop */
 196         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 197         {
 198
 199             /* Get j neighbor index, and coordinate index */
 200             jnrA             = jjnr[jidx];
 201             jnrB             = jjnr[jidx+1];
 202             jnrC             = jjnr[jidx+2];
 203             jnrD             = jjnr[jidx+3];
 204             jnrE             = jjnr[jidx+4];
 205             jnrF             = jjnr[jidx+5];
 206             jnrG             = jjnr[jidx+6];
 207             jnrH             = jjnr[jidx+7];
 208             j_coord_offsetA  = DIM*jnrA;
 209             j_coord_offsetB  = DIM*jnrB;
 210             j_coord_offsetC  = DIM*jnrC;
 211             j_coord_offsetD  = DIM*jnrD;
 212             j_coord_offsetE  = DIM*jnrE;
 213             j_coord_offsetF  = DIM*jnrF;
 214             j_coord_offsetG  = DIM*jnrG;
 215             j_coord_offsetH  = DIM*jnrH;
 216
 217             /* load j atom coordinates */
 218             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 219                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 220                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 221                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 222                                                  &jx0,&jy0,&jz0);
 223
 224             /* Calculate displacement vector */
 225             dx00             = _mm256_sub_ps(ix0,jx0);
 226             dy00             = _mm256_sub_ps(iy0,jy0);
 227             dz00             = _mm256_sub_ps(iz0,jz0);
 228
 229             /* Calculate squared distance and things based on it */
 230             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 231
 232             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 233
 234             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 235
 236             /* Load parameters for j particles */
 237             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 238                                                                  charge+jnrC+0,charge+jnrD+0,
 239                                                                  charge+jnrE+0,charge+jnrF+0,
 240                                                                  charge+jnrG+0,charge+jnrH+0);
 241             vdwjidx0A        = 2*vdwtype[jnrA+0];
 242             vdwjidx0B        = 2*vdwtype[jnrB+0];
 243             vdwjidx0C        = 2*vdwtype[jnrC+0];
 244             vdwjidx0D        = 2*vdwtype[jnrD+0];
 245             vdwjidx0E        = 2*vdwtype[jnrE+0];
 246             vdwjidx0F        = 2*vdwtype[jnrF+0];
 247             vdwjidx0G        = 2*vdwtype[jnrG+0];
 248             vdwjidx0H        = 2*vdwtype[jnrH+0];
 249
 250             /**************************
 251              * CALCULATE INTERACTIONS *
 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             c6grid_00       = gmx_mm256_load_8real_swizzle_ps(vdwgridioffsetptr0+vdwjidx0A,
 269                                                                   vdwgridioffsetptr0+vdwjidx0B,
 270                                                                   vdwgridioffsetptr0+vdwjidx0C,
 271                                                                   vdwgridioffsetptr0+vdwjidx0D,
 272                                                                   vdwgridioffsetptr0+vdwjidx0E,
 273                                                                   vdwgridioffsetptr0+vdwjidx0F,
 274                                                                   vdwgridioffsetptr0+vdwjidx0G,
 275                                                                   vdwgridioffsetptr0+vdwjidx0H);
 276
 277             /* EWALD ELECTROSTATICS */
 278
 279             /* Analytical PME correction */
 280             zeta2            = _mm256_mul_ps(beta2,rsq00);
 281             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 282             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 283             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 284             felec            = _mm256_mul_ps(qq00,felec);
 285             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 286             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 287             velec            = _mm256_sub_ps(rinv00,pmecorrV);
 288             velec            = _mm256_mul_ps(qq00,velec);
 289
 290             /* Analytical LJ-PME */
 291             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 292             ewcljrsq         = _mm256_mul_ps(ewclj2,rsq00);
 293             ewclj6           = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
 294             exponent         = gmx_simd_exp_r(ewcljrsq);
 295             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 296             poly             = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
 297             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 298             vvdw6            = _mm256_mul_ps(_mm256_sub_ps(c6_00,_mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly))),rinvsix);
 299             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 300             vvdw             = _mm256_sub_ps(_mm256_mul_ps(vvdw12,one_twelfth),_mm256_mul_ps(vvdw6,one_sixth));
 301             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 302             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,_mm256_sub_ps(vvdw6,_mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6)))),rinvsq00);
 303
 304             /* Update potential sum for this i atom from the interaction with this j atom. */
 305             velecsum         = _mm256_add_ps(velecsum,velec);
 306             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 307
 308             fscal            = _mm256_add_ps(felec,fvdw);
 309
 310             /* Calculate temporary vectorial force */
 311             tx               = _mm256_mul_ps(fscal,dx00);
 312             ty               = _mm256_mul_ps(fscal,dy00);
 313             tz               = _mm256_mul_ps(fscal,dz00);
 314
 315             /* Update vectorial force */
 316             fix0             = _mm256_add_ps(fix0,tx);
 317             fiy0             = _mm256_add_ps(fiy0,ty);
 318             fiz0             = _mm256_add_ps(fiz0,tz);
 319
 320             fjptrA             = f+j_coord_offsetA;
 321             fjptrB             = f+j_coord_offsetB;
 322             fjptrC             = f+j_coord_offsetC;
 323             fjptrD             = f+j_coord_offsetD;
 324             fjptrE             = f+j_coord_offsetE;
 325             fjptrF             = f+j_coord_offsetF;
 326             fjptrG             = f+j_coord_offsetG;
 327             fjptrH             = f+j_coord_offsetH;
 328             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 329
 330             /* Inner loop uses 112 flops */
 331         }
 332
 333         if(jidx<j_index_end)
 334         {
 335
 336             /* Get j neighbor index, and coordinate index */
 337             jnrlistA         = jjnr[jidx];
 338             jnrlistB         = jjnr[jidx+1];
 339             jnrlistC         = jjnr[jidx+2];
 340             jnrlistD         = jjnr[jidx+3];
 341             jnrlistE         = jjnr[jidx+4];
 342             jnrlistF         = jjnr[jidx+5];
 343             jnrlistG         = jjnr[jidx+6];
 344             jnrlistH         = jjnr[jidx+7];
 345             /* Sign of each element will be negative for non-real atoms.
 346              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 347              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 348              */
 349             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 350                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 351
 352             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 353             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 354             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 355             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 356             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 357             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 358             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 359             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 360             j_coord_offsetA  = DIM*jnrA;
 361             j_coord_offsetB  = DIM*jnrB;
 362             j_coord_offsetC  = DIM*jnrC;
 363             j_coord_offsetD  = DIM*jnrD;
 364             j_coord_offsetE  = DIM*jnrE;
 365             j_coord_offsetF  = DIM*jnrF;
 366             j_coord_offsetG  = DIM*jnrG;
 367             j_coord_offsetH  = DIM*jnrH;
 368
 369             /* load j atom coordinates */
 370             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 371                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 372                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 373                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 374                                                  &jx0,&jy0,&jz0);
 375
 376             /* Calculate displacement vector */
 377             dx00             = _mm256_sub_ps(ix0,jx0);
 378             dy00             = _mm256_sub_ps(iy0,jy0);
 379             dz00             = _mm256_sub_ps(iz0,jz0);
 380
 381             /* Calculate squared distance and things based on it */
 382             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 383
 384             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 385
 386             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 387
 388             /* Load parameters for j particles */
 389             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 390                                                                  charge+jnrC+0,charge+jnrD+0,
 391                                                                  charge+jnrE+0,charge+jnrF+0,
 392                                                                  charge+jnrG+0,charge+jnrH+0);
 393             vdwjidx0A        = 2*vdwtype[jnrA+0];
 394             vdwjidx0B        = 2*vdwtype[jnrB+0];
 395             vdwjidx0C        = 2*vdwtype[jnrC+0];
 396             vdwjidx0D        = 2*vdwtype[jnrD+0];
 397             vdwjidx0E        = 2*vdwtype[jnrE+0];
 398             vdwjidx0F        = 2*vdwtype[jnrF+0];
 399             vdwjidx0G        = 2*vdwtype[jnrG+0];
 400             vdwjidx0H        = 2*vdwtype[jnrH+0];
 401
 402             /**************************
 403              * CALCULATE INTERACTIONS *
 404              **************************/
 405
 406             r00              = _mm256_mul_ps(rsq00,rinv00);
 407             r00              = _mm256_andnot_ps(dummy_mask,r00);
 408
 409             /* Compute parameters for interactions between i and j atoms */
 410             qq00             = _mm256_mul_ps(iq0,jq0);
 411             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 412                                             vdwioffsetptr0+vdwjidx0B,
 413                                             vdwioffsetptr0+vdwjidx0C,
 414                                             vdwioffsetptr0+vdwjidx0D,
 415                                             vdwioffsetptr0+vdwjidx0E,
 416                                             vdwioffsetptr0+vdwjidx0F,
 417                                             vdwioffsetptr0+vdwjidx0G,
 418                                             vdwioffsetptr0+vdwjidx0H,
 419                                             &c6_00,&c12_00);
 420
 421             c6grid_00       = gmx_mm256_load_8real_swizzle_ps(vdwgridioffsetptr0+vdwjidx0A,
 422                                                                   vdwgridioffsetptr0+vdwjidx0B,
 423                                                                   vdwgridioffsetptr0+vdwjidx0C,
 424                                                                   vdwgridioffsetptr0+vdwjidx0D,
 425                                                                   vdwgridioffsetptr0+vdwjidx0E,
 426                                                                   vdwgridioffsetptr0+vdwjidx0F,
 427                                                                   vdwgridioffsetptr0+vdwjidx0G,
 428                                                                   vdwgridioffsetptr0+vdwjidx0H);
 429
 430             /* EWALD ELECTROSTATICS */
 431
 432             /* Analytical PME correction */
 433             zeta2            = _mm256_mul_ps(beta2,rsq00);
 434             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 435             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 436             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 437             felec            = _mm256_mul_ps(qq00,felec);
 438             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 439             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 440             velec            = _mm256_sub_ps(rinv00,pmecorrV);
 441             velec            = _mm256_mul_ps(qq00,velec);
 442
 443             /* Analytical LJ-PME */
 444             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 445             ewcljrsq         = _mm256_mul_ps(ewclj2,rsq00);
 446             ewclj6           = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
 447             exponent         = gmx_simd_exp_r(ewcljrsq);
 448             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 449             poly             = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
 450             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 451             vvdw6            = _mm256_mul_ps(_mm256_sub_ps(c6_00,_mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly))),rinvsix);
 452             vvdw12           = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
 453             vvdw             = _mm256_sub_ps(_mm256_mul_ps(vvdw12,one_twelfth),_mm256_mul_ps(vvdw6,one_sixth));
 454             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 455             fvdw             = _mm256_mul_ps(_mm256_sub_ps(vvdw12,_mm256_sub_ps(vvdw6,_mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6)))),rinvsq00);
 456
 457             /* Update potential sum for this i atom from the interaction with this j atom. */
 458             velec            = _mm256_andnot_ps(dummy_mask,velec);
 459             velecsum         = _mm256_add_ps(velecsum,velec);
 460             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
 461             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 462
 463             fscal            = _mm256_add_ps(felec,fvdw);
 464
 465             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 466
 467             /* Calculate temporary vectorial force */
 468             tx               = _mm256_mul_ps(fscal,dx00);
 469             ty               = _mm256_mul_ps(fscal,dy00);
 470             tz               = _mm256_mul_ps(fscal,dz00);
 471
 472             /* Update vectorial force */
 473             fix0             = _mm256_add_ps(fix0,tx);
 474             fiy0             = _mm256_add_ps(fiy0,ty);
 475             fiz0             = _mm256_add_ps(fiz0,tz);
 476
 477             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 478             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 479             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 480             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 481             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 482             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 483             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 484             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 485             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 486
 487             /* Inner loop uses 113 flops */
 488         }
 489
 490         /* End of innermost loop */
 491
 492         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 493                                                  f+i_coord_offset,fshift+i_shift_offset);
 494
 495         ggid                        = gid[iidx];
 496         /* Update potential energies */
 497         gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 498         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 499
 500         /* Increment number of inner iterations */
 501         inneriter                  += j_index_end - j_index_start;
 502
 503         /* Outer loop uses 9 flops */
 504     }
 505
 506     /* Increment number of outer iterations */
 507     outeriter        += nri;
 508
 509     /* Update outer/inner flops */
 510
 511     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*113);
 512 }
 513 /*
 514  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwLJEw_GeomP1P1_F_avx_256_single
 515  * Electrostatics interaction: Ewald
 516  * VdW interaction:            LJEwald
 517  * Geometry:                   Particle-Particle
 518  * Calculate force/pot:        Force
 519  */
 520 void
 521 nb_kernel_ElecEw_VdwLJEw_GeomP1P1_F_avx_256_single
 522                     (t_nblist                    * gmx_restrict       nlist,
 523                      rvec                        * gmx_restrict          xx,
 524                      rvec                        * gmx_restrict          ff,
 525                      t_forcerec                  * gmx_restrict          fr,
 526                      t_mdatoms                   * gmx_restrict     mdatoms,
 527                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 528                      t_nrnb                      * gmx_restrict        nrnb)
 529 {
 530     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 531      * just 0 for non-waters.
 532      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
 533      * jnr indices corresponding to data put in the four positions in the SIMD register.
 534      */
 535     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 536     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 537     int              jnrA,jnrB,jnrC,jnrD;
 538     int              jnrE,jnrF,jnrG,jnrH;
 539     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 540     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 541     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 542     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
 543     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 544     real             rcutoff_scalar;
 545     real             *shiftvec,*fshift,*x,*f;
 546     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
 547     real             scratch[4*DIM];
 548     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 549     real *           vdwioffsetptr0;
 550     real *           vdwgridioffsetptr0;
 551     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 552     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
 553     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 554     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 555     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
 556     real             *charge;
 557     int              nvdwtype;
 558     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 559     int              *vdwtype;
 560     real             *vdwparam;
 561     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 562     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 563     __m256           c6grid_00;
 564     real             *vdwgridparam;
 565     __m256           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 566     __m256           one_half  = _mm256_set1_ps(0.5);
 567     __m256           minus_one = _mm256_set1_ps(-1.0);
 568     __m256i          ewitab;
 569     __m128i          ewitab_lo,ewitab_hi;
 570     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 571     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 572     real             *ewtab;
 573     __m256           dummy_mask,cutoff_mask;
 574     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 575     __m256           one     = _mm256_set1_ps(1.0);
 576     __m256           two     = _mm256_set1_ps(2.0);
 577     x                = xx[0];
 578     f                = ff[0];
 579
 580     nri              = nlist->nri;
 581     iinr             = nlist->iinr;
 582     jindex           = nlist->jindex;
 583     jjnr             = nlist->jjnr;
 584     shiftidx         = nlist->shift;
 585     gid              = nlist->gid;
 586     shiftvec         = fr->shift_vec[0];
 587     fshift           = fr->fshift[0];
 588     facel            = _mm256_set1_ps(fr->epsfac);
 589     charge           = mdatoms->chargeA;
 590     nvdwtype         = fr->ntype;
 591     vdwparam         = fr->nbfp;
 592     vdwtype          = mdatoms->typeA;
 593     vdwgridparam     = fr->ljpme_c6grid;
 594     sh_lj_ewald      = _mm256_set1_ps(fr->ic->sh_lj_ewald);
 595     ewclj            = _mm256_set1_ps(fr->ewaldcoeff_lj);
 596     ewclj2           = _mm256_mul_ps(minus_one,_mm256_mul_ps(ewclj,ewclj));
 597
 598     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
 599     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
 600     beta2            = _mm256_mul_ps(beta,beta);
 601     beta3            = _mm256_mul_ps(beta,beta2);
 602
 603     ewtab            = fr->ic->tabq_coul_F;
 604     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
 605     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
 606
 607     /* Avoid stupid compiler warnings */
 608     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 609     j_coord_offsetA = 0;
 610     j_coord_offsetB = 0;
 611     j_coord_offsetC = 0;
 612     j_coord_offsetD = 0;
 613     j_coord_offsetE = 0;
 614     j_coord_offsetF = 0;
 615     j_coord_offsetG = 0;
 616     j_coord_offsetH = 0;
 617
 618     outeriter        = 0;
 619     inneriter        = 0;
 620
 621     for(iidx=0;iidx<4*DIM;iidx++)
 622     {
 623         scratch[iidx] = 0.0;
 624     }
 625
 626     /* Start outer loop over neighborlists */
 627     for(iidx=0; iidx<nri; iidx++)
 628     {
 629         /* Load shift vector for this list */
 630         i_shift_offset   = DIM*shiftidx[iidx];
 631
 632         /* Load limits for loop over neighbors */
 633         j_index_start    = jindex[iidx];
 634         j_index_end      = jindex[iidx+1];
 635
 636         /* Get outer coordinate index */
 637         inr              = iinr[iidx];
 638         i_coord_offset   = DIM*inr;
 639
 640         /* Load i particle coords and add shift vector */
 641         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 642
 643         fix0             = _mm256_setzero_ps();
 644         fiy0             = _mm256_setzero_ps();
 645         fiz0             = _mm256_setzero_ps();
 646
 647         /* Load parameters for i particles */
 648         iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 649         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 650         vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
 651
 652         /* Start inner kernel loop */
 653         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 654         {
 655
 656             /* Get j neighbor index, and coordinate index */
 657             jnrA             = jjnr[jidx];
 658             jnrB             = jjnr[jidx+1];
 659             jnrC             = jjnr[jidx+2];
 660             jnrD             = jjnr[jidx+3];
 661             jnrE             = jjnr[jidx+4];
 662             jnrF             = jjnr[jidx+5];
 663             jnrG             = jjnr[jidx+6];
 664             jnrH             = jjnr[jidx+7];
 665             j_coord_offsetA  = DIM*jnrA;
 666             j_coord_offsetB  = DIM*jnrB;
 667             j_coord_offsetC  = DIM*jnrC;
 668             j_coord_offsetD  = DIM*jnrD;
 669             j_coord_offsetE  = DIM*jnrE;
 670             j_coord_offsetF  = DIM*jnrF;
 671             j_coord_offsetG  = DIM*jnrG;
 672             j_coord_offsetH  = DIM*jnrH;
 673
 674             /* load j atom coordinates */
 675             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 676                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 677                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 678                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 679                                                  &jx0,&jy0,&jz0);
 680
 681             /* Calculate displacement vector */
 682             dx00             = _mm256_sub_ps(ix0,jx0);
 683             dy00             = _mm256_sub_ps(iy0,jy0);
 684             dz00             = _mm256_sub_ps(iz0,jz0);
 685
 686             /* Calculate squared distance and things based on it */
 687             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 688
 689             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 690
 691             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 692
 693             /* Load parameters for j particles */
 694             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 695                                                                  charge+jnrC+0,charge+jnrD+0,
 696                                                                  charge+jnrE+0,charge+jnrF+0,
 697                                                                  charge+jnrG+0,charge+jnrH+0);
 698             vdwjidx0A        = 2*vdwtype[jnrA+0];
 699             vdwjidx0B        = 2*vdwtype[jnrB+0];
 700             vdwjidx0C        = 2*vdwtype[jnrC+0];
 701             vdwjidx0D        = 2*vdwtype[jnrD+0];
 702             vdwjidx0E        = 2*vdwtype[jnrE+0];
 703             vdwjidx0F        = 2*vdwtype[jnrF+0];
 704             vdwjidx0G        = 2*vdwtype[jnrG+0];
 705             vdwjidx0H        = 2*vdwtype[jnrH+0];
 706
 707             /**************************
 708              * CALCULATE INTERACTIONS *
 709              **************************/
 710
 711             r00              = _mm256_mul_ps(rsq00,rinv00);
 712
 713             /* Compute parameters for interactions between i and j atoms */
 714             qq00             = _mm256_mul_ps(iq0,jq0);
 715             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 716                                             vdwioffsetptr0+vdwjidx0B,
 717                                             vdwioffsetptr0+vdwjidx0C,
 718                                             vdwioffsetptr0+vdwjidx0D,
 719                                             vdwioffsetptr0+vdwjidx0E,
 720                                             vdwioffsetptr0+vdwjidx0F,
 721                                             vdwioffsetptr0+vdwjidx0G,
 722                                             vdwioffsetptr0+vdwjidx0H,
 723                                             &c6_00,&c12_00);
 724
 725             c6grid_00       = gmx_mm256_load_8real_swizzle_ps(vdwgridioffsetptr0+vdwjidx0A,
 726                                                                   vdwgridioffsetptr0+vdwjidx0B,
 727                                                                   vdwgridioffsetptr0+vdwjidx0C,
 728                                                                   vdwgridioffsetptr0+vdwjidx0D,
 729                                                                   vdwgridioffsetptr0+vdwjidx0E,
 730                                                                   vdwgridioffsetptr0+vdwjidx0F,
 731                                                                   vdwgridioffsetptr0+vdwjidx0G,
 732                                                                   vdwgridioffsetptr0+vdwjidx0H);
 733
 734             /* EWALD ELECTROSTATICS */
 735
 736             /* Analytical PME correction */
 737             zeta2            = _mm256_mul_ps(beta2,rsq00);
 738             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 739             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 740             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 741             felec            = _mm256_mul_ps(qq00,felec);
 742
 743             /* Analytical LJ-PME */
 744             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 745             ewcljrsq         = _mm256_mul_ps(ewclj2,rsq00);
 746             ewclj6           = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
 747             exponent         = gmx_simd_exp_r(ewcljrsq);
 748             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 749             poly             = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
 750             /* f6A = 6 * C6grid * (1 - poly) */
 751             f6A              = _mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly));
 752             /* f6B = C6grid * exponent * beta^6 */
 753             f6B              = _mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6));
 754             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 755             fvdw              = _mm256_mul_ps(_mm256_add_ps(_mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),_mm256_sub_ps(c6_00,f6A)),rinvsix),f6B),rinvsq00);
 756
 757             fscal            = _mm256_add_ps(felec,fvdw);
 758
 759             /* Calculate temporary vectorial force */
 760             tx               = _mm256_mul_ps(fscal,dx00);
 761             ty               = _mm256_mul_ps(fscal,dy00);
 762             tz               = _mm256_mul_ps(fscal,dz00);
 763
 764             /* Update vectorial force */
 765             fix0             = _mm256_add_ps(fix0,tx);
 766             fiy0             = _mm256_add_ps(fiy0,ty);
 767             fiz0             = _mm256_add_ps(fiz0,tz);
 768
 769             fjptrA             = f+j_coord_offsetA;
 770             fjptrB             = f+j_coord_offsetB;
 771             fjptrC             = f+j_coord_offsetC;
 772             fjptrD             = f+j_coord_offsetD;
 773             fjptrE             = f+j_coord_offsetE;
 774             fjptrF             = f+j_coord_offsetF;
 775             fjptrG             = f+j_coord_offsetG;
 776             fjptrH             = f+j_coord_offsetH;
 777             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 778
 779             /* Inner loop uses 79 flops */
 780         }
 781
 782         if(jidx<j_index_end)
 783         {
 784
 785             /* Get j neighbor index, and coordinate index */
 786             jnrlistA         = jjnr[jidx];
 787             jnrlistB         = jjnr[jidx+1];
 788             jnrlistC         = jjnr[jidx+2];
 789             jnrlistD         = jjnr[jidx+3];
 790             jnrlistE         = jjnr[jidx+4];
 791             jnrlistF         = jjnr[jidx+5];
 792             jnrlistG         = jjnr[jidx+6];
 793             jnrlistH         = jjnr[jidx+7];
 794             /* Sign of each element will be negative for non-real atoms.
 795              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 796              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 797              */
 798             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 799                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 800
 801             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 802             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 803             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 804             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 805             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 806             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 807             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 808             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 809             j_coord_offsetA  = DIM*jnrA;
 810             j_coord_offsetB  = DIM*jnrB;
 811             j_coord_offsetC  = DIM*jnrC;
 812             j_coord_offsetD  = DIM*jnrD;
 813             j_coord_offsetE  = DIM*jnrE;
 814             j_coord_offsetF  = DIM*jnrF;
 815             j_coord_offsetG  = DIM*jnrG;
 816             j_coord_offsetH  = DIM*jnrH;
 817
 818             /* load j atom coordinates */
 819             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 820                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 821                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 822                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 823                                                  &jx0,&jy0,&jz0);
 824
 825             /* Calculate displacement vector */
 826             dx00             = _mm256_sub_ps(ix0,jx0);
 827             dy00             = _mm256_sub_ps(iy0,jy0);
 828             dz00             = _mm256_sub_ps(iz0,jz0);
 829
 830             /* Calculate squared distance and things based on it */
 831             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 832
 833             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 834
 835             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 836
 837             /* Load parameters for j particles */
 838             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 839                                                                  charge+jnrC+0,charge+jnrD+0,
 840                                                                  charge+jnrE+0,charge+jnrF+0,
 841                                                                  charge+jnrG+0,charge+jnrH+0);
 842             vdwjidx0A        = 2*vdwtype[jnrA+0];
 843             vdwjidx0B        = 2*vdwtype[jnrB+0];
 844             vdwjidx0C        = 2*vdwtype[jnrC+0];
 845             vdwjidx0D        = 2*vdwtype[jnrD+0];
 846             vdwjidx0E        = 2*vdwtype[jnrE+0];
 847             vdwjidx0F        = 2*vdwtype[jnrF+0];
 848             vdwjidx0G        = 2*vdwtype[jnrG+0];
 849             vdwjidx0H        = 2*vdwtype[jnrH+0];
 850
 851             /**************************
 852              * CALCULATE INTERACTIONS *
 853              **************************/
 854
 855             r00              = _mm256_mul_ps(rsq00,rinv00);
 856             r00              = _mm256_andnot_ps(dummy_mask,r00);
 857
 858             /* Compute parameters for interactions between i and j atoms */
 859             qq00             = _mm256_mul_ps(iq0,jq0);
 860             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 861                                             vdwioffsetptr0+vdwjidx0B,
 862                                             vdwioffsetptr0+vdwjidx0C,
 863                                             vdwioffsetptr0+vdwjidx0D,
 864                                             vdwioffsetptr0+vdwjidx0E,
 865                                             vdwioffsetptr0+vdwjidx0F,
 866                                             vdwioffsetptr0+vdwjidx0G,
 867                                             vdwioffsetptr0+vdwjidx0H,
 868                                             &c6_00,&c12_00);
 869
 870             c6grid_00       = gmx_mm256_load_8real_swizzle_ps(vdwgridioffsetptr0+vdwjidx0A,
 871                                                                   vdwgridioffsetptr0+vdwjidx0B,
 872                                                                   vdwgridioffsetptr0+vdwjidx0C,
 873                                                                   vdwgridioffsetptr0+vdwjidx0D,
 874                                                                   vdwgridioffsetptr0+vdwjidx0E,
 875                                                                   vdwgridioffsetptr0+vdwjidx0F,
 876                                                                   vdwgridioffsetptr0+vdwjidx0G,
 877                                                                   vdwgridioffsetptr0+vdwjidx0H);
 878
 879             /* EWALD ELECTROSTATICS */
 880
 881             /* Analytical PME correction */
 882             zeta2            = _mm256_mul_ps(beta2,rsq00);
 883             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 884             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 885             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 886             felec            = _mm256_mul_ps(qq00,felec);
 887
 888             /* Analytical LJ-PME */
 889             rinvsix          = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
 890             ewcljrsq         = _mm256_mul_ps(ewclj2,rsq00);
 891             ewclj6           = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
 892             exponent         = gmx_simd_exp_r(ewcljrsq);
 893             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 894             poly             = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
 895             /* f6A = 6 * C6grid * (1 - poly) */
 896             f6A              = _mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly));
 897             /* f6B = C6grid * exponent * beta^6 */
 898             f6B              = _mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6));
 899             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
 900             fvdw              = _mm256_mul_ps(_mm256_add_ps(_mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),_mm256_sub_ps(c6_00,f6A)),rinvsix),f6B),rinvsq00);
 901
 902             fscal            = _mm256_add_ps(felec,fvdw);
 903
 904             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 905
 906             /* Calculate temporary vectorial force */
 907             tx               = _mm256_mul_ps(fscal,dx00);
 908             ty               = _mm256_mul_ps(fscal,dy00);
 909             tz               = _mm256_mul_ps(fscal,dz00);
 910
 911             /* Update vectorial force */
 912             fix0             = _mm256_add_ps(fix0,tx);
 913             fiy0             = _mm256_add_ps(fiy0,ty);
 914             fiz0             = _mm256_add_ps(fiz0,tz);
 915
 916             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 917             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 918             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 919             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 920             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 921             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 922             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 923             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 924             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 925
 926             /* Inner loop uses 80 flops */
 927         }
 928
 929         /* End of innermost loop */
 930
 931         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 932                                                  f+i_coord_offset,fshift+i_shift_offset);
 933
 934         /* Increment number of inner iterations */
 935         inneriter                  += j_index_end - j_index_start;
 936
 937         /* Outer loop uses 7 flops */
 938     }
 939
 940     /* Increment number of outer iterations */
 941     outeriter        += nri;
 942
 943     /* Update outer/inner flops */
 944
 945     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*80);
 946 }