src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_single/nb_kernel_ElecEw_VdwCSTab_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_ElecEw_VdwCSTab_GeomP1P1_VF_avx_256_single
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            CubicSplineTable
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEw_VdwCSTab_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     __m256i          vfitab;
 102     __m128i          vfitab_lo,vfitab_hi;
 103     __m128i          ifour       = _mm_set1_epi32(4);
 104     __m256           rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 105     real             *vftab;
 106     __m256i          ewitab;
 107     __m128i          ewitab_lo,ewitab_hi;
 108     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 109     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 110     real             *ewtab;
 111     __m256           dummy_mask,cutoff_mask;
 112     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 113     __m256           one     = _mm256_set1_ps(1.0);
 114     __m256           two     = _mm256_set1_ps(2.0);
 115     x                = xx[0];
 116     f                = ff[0];
 117
 118     nri              = nlist->nri;
 119     iinr             = nlist->iinr;
 120     jindex           = nlist->jindex;
 121     jjnr             = nlist->jjnr;
 122     shiftidx         = nlist->shift;
 123     gid              = nlist->gid;
 124     shiftvec         = fr->shift_vec[0];
 125     fshift           = fr->fshift[0];
 126     facel            = _mm256_set1_ps(fr->epsfac);
 127     charge           = mdatoms->chargeA;
 128     nvdwtype         = fr->ntype;
 129     vdwparam         = fr->nbfp;
 130     vdwtype          = mdatoms->typeA;
 131
 132     vftab            = kernel_data->table_vdw->data;
 133     vftabscale       = _mm256_set1_ps(kernel_data->table_vdw->scale);
 134
 135     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
 136     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
 137     beta2            = _mm256_mul_ps(beta,beta);
 138     beta3            = _mm256_mul_ps(beta,beta2);
 139
 140     ewtab            = fr->ic->tabq_coul_FDV0;
 141     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
 142     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
 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             r00              = _mm256_mul_ps(rsq00,rinv00);
 252
 253             /* Compute parameters for interactions between i and j atoms */
 254             qq00             = _mm256_mul_ps(iq0,jq0);
 255             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 256                                             vdwioffsetptr0+vdwjidx0B,
 257                                             vdwioffsetptr0+vdwjidx0C,
 258                                             vdwioffsetptr0+vdwjidx0D,
 259                                             vdwioffsetptr0+vdwjidx0E,
 260                                             vdwioffsetptr0+vdwjidx0F,
 261                                             vdwioffsetptr0+vdwjidx0G,
 262                                             vdwioffsetptr0+vdwjidx0H,
 263                                             &c6_00,&c12_00);
 264
 265             /* Calculate table index by multiplying r with table scale and truncate to integer */
 266             rt               = _mm256_mul_ps(r00,vftabscale);
 267             vfitab           = _mm256_cvttps_epi32(rt);
 268             vfeps            = _mm256_sub_ps(rt,_mm256_round_ps(rt, _MM_FROUND_FLOOR));
 269             /*         AVX1 does not support 256-bit integer operations, so now we go to 128-bit mode... */
 270             vfitab_lo        = _mm256_extractf128_si256(vfitab,0x0);
 271             vfitab_hi        = _mm256_extractf128_si256(vfitab,0x1);
 272             vfitab_lo        = _mm_slli_epi32(vfitab_lo,3);
 273             vfitab_hi        = _mm_slli_epi32(vfitab_hi,3);
 274
 275             /* EWALD ELECTROSTATICS */
 276
 277             /* Analytical PME correction */
 278             zeta2            = _mm256_mul_ps(beta2,rsq00);
 279             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 280             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 281             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 282             felec            = _mm256_mul_ps(qq00,felec);
 283             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 284             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 285             velec            = _mm256_sub_ps(rinv00,pmecorrV);
 286             velec            = _mm256_mul_ps(qq00,velec);
 287
 288             /* CUBIC SPLINE TABLE DISPERSION */
 289             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 290                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 291             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 292                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 293             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 294                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 295             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 296                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 297             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 298             Heps             = _mm256_mul_ps(vfeps,H);
 299             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 300             VV               = _mm256_add_ps(Y,_mm256_mul_ps(vfeps,Fp));
 301             vvdw6            = _mm256_mul_ps(c6_00,VV);
 302             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 303             fvdw6            = _mm256_mul_ps(c6_00,FF);
 304
 305             /* CUBIC SPLINE TABLE REPULSION */
 306             vfitab_lo        = _mm_add_epi32(vfitab_lo,ifour);
 307             vfitab_hi        = _mm_add_epi32(vfitab_hi,ifour);
 308             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 309                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 310             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 311                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 312             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 313                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 314             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 315                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 316             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 317             Heps             = _mm256_mul_ps(vfeps,H);
 318             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 319             VV               = _mm256_add_ps(Y,_mm256_mul_ps(vfeps,Fp));
 320             vvdw12           = _mm256_mul_ps(c12_00,VV);
 321             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 322             fvdw12           = _mm256_mul_ps(c12_00,FF);
 323             vvdw             = _mm256_add_ps(vvdw12,vvdw6);
 324             fvdw             = _mm256_xor_ps(signbit,_mm256_mul_ps(_mm256_add_ps(fvdw6,fvdw12),_mm256_mul_ps(vftabscale,rinv00)));
 325
 326             /* Update potential sum for this i atom from the interaction with this j atom. */
 327             velecsum         = _mm256_add_ps(velecsum,velec);
 328             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 329
 330             fscal            = _mm256_add_ps(felec,fvdw);
 331
 332             /* Calculate temporary vectorial force */
 333             tx               = _mm256_mul_ps(fscal,dx00);
 334             ty               = _mm256_mul_ps(fscal,dy00);
 335             tz               = _mm256_mul_ps(fscal,dz00);
 336
 337             /* Update vectorial force */
 338             fix0             = _mm256_add_ps(fix0,tx);
 339             fiy0             = _mm256_add_ps(fiy0,ty);
 340             fiz0             = _mm256_add_ps(fiz0,tz);
 341
 342             fjptrA             = f+j_coord_offsetA;
 343             fjptrB             = f+j_coord_offsetB;
 344             fjptrC             = f+j_coord_offsetC;
 345             fjptrD             = f+j_coord_offsetD;
 346             fjptrE             = f+j_coord_offsetE;
 347             fjptrF             = f+j_coord_offsetF;
 348             fjptrG             = f+j_coord_offsetG;
 349             fjptrH             = f+j_coord_offsetH;
 350             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 351
 352             /* Inner loop uses 118 flops */
 353         }
 354
 355         if(jidx<j_index_end)
 356         {
 357
 358             /* Get j neighbor index, and coordinate index */
 359             jnrlistA         = jjnr[jidx];
 360             jnrlistB         = jjnr[jidx+1];
 361             jnrlistC         = jjnr[jidx+2];
 362             jnrlistD         = jjnr[jidx+3];
 363             jnrlistE         = jjnr[jidx+4];
 364             jnrlistF         = jjnr[jidx+5];
 365             jnrlistG         = jjnr[jidx+6];
 366             jnrlistH         = jjnr[jidx+7];
 367             /* Sign of each element will be negative for non-real atoms.
 368              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 369              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 370              */
 371             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 372                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 373
 374             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 375             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 376             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 377             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 378             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 379             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 380             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 381             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 382             j_coord_offsetA  = DIM*jnrA;
 383             j_coord_offsetB  = DIM*jnrB;
 384             j_coord_offsetC  = DIM*jnrC;
 385             j_coord_offsetD  = DIM*jnrD;
 386             j_coord_offsetE  = DIM*jnrE;
 387             j_coord_offsetF  = DIM*jnrF;
 388             j_coord_offsetG  = DIM*jnrG;
 389             j_coord_offsetH  = DIM*jnrH;
 390
 391             /* load j atom coordinates */
 392             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 393                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 394                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 395                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 396                                                  &jx0,&jy0,&jz0);
 397
 398             /* Calculate displacement vector */
 399             dx00             = _mm256_sub_ps(ix0,jx0);
 400             dy00             = _mm256_sub_ps(iy0,jy0);
 401             dz00             = _mm256_sub_ps(iz0,jz0);
 402
 403             /* Calculate squared distance and things based on it */
 404             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 405
 406             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 407
 408             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 409
 410             /* Load parameters for j particles */
 411             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 412                                                                  charge+jnrC+0,charge+jnrD+0,
 413                                                                  charge+jnrE+0,charge+jnrF+0,
 414                                                                  charge+jnrG+0,charge+jnrH+0);
 415             vdwjidx0A        = 2*vdwtype[jnrA+0];
 416             vdwjidx0B        = 2*vdwtype[jnrB+0];
 417             vdwjidx0C        = 2*vdwtype[jnrC+0];
 418             vdwjidx0D        = 2*vdwtype[jnrD+0];
 419             vdwjidx0E        = 2*vdwtype[jnrE+0];
 420             vdwjidx0F        = 2*vdwtype[jnrF+0];
 421             vdwjidx0G        = 2*vdwtype[jnrG+0];
 422             vdwjidx0H        = 2*vdwtype[jnrH+0];
 423
 424             /**************************
 425              * CALCULATE INTERACTIONS *
 426              **************************/
 427
 428             r00              = _mm256_mul_ps(rsq00,rinv00);
 429             r00              = _mm256_andnot_ps(dummy_mask,r00);
 430
 431             /* Compute parameters for interactions between i and j atoms */
 432             qq00             = _mm256_mul_ps(iq0,jq0);
 433             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 434                                             vdwioffsetptr0+vdwjidx0B,
 435                                             vdwioffsetptr0+vdwjidx0C,
 436                                             vdwioffsetptr0+vdwjidx0D,
 437                                             vdwioffsetptr0+vdwjidx0E,
 438                                             vdwioffsetptr0+vdwjidx0F,
 439                                             vdwioffsetptr0+vdwjidx0G,
 440                                             vdwioffsetptr0+vdwjidx0H,
 441                                             &c6_00,&c12_00);
 442
 443             /* Calculate table index by multiplying r with table scale and truncate to integer */
 444             rt               = _mm256_mul_ps(r00,vftabscale);
 445             vfitab           = _mm256_cvttps_epi32(rt);
 446             vfeps            = _mm256_sub_ps(rt,_mm256_round_ps(rt, _MM_FROUND_FLOOR));
 447             /*         AVX1 does not support 256-bit integer operations, so now we go to 128-bit mode... */
 448             vfitab_lo        = _mm256_extractf128_si256(vfitab,0x0);
 449             vfitab_hi        = _mm256_extractf128_si256(vfitab,0x1);
 450             vfitab_lo        = _mm_slli_epi32(vfitab_lo,3);
 451             vfitab_hi        = _mm_slli_epi32(vfitab_hi,3);
 452
 453             /* EWALD ELECTROSTATICS */
 454
 455             /* Analytical PME correction */
 456             zeta2            = _mm256_mul_ps(beta2,rsq00);
 457             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 458             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 459             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 460             felec            = _mm256_mul_ps(qq00,felec);
 461             pmecorrV         = gmx_mm256_pmecorrV_ps(zeta2);
 462             pmecorrV         = _mm256_mul_ps(pmecorrV,beta);
 463             velec            = _mm256_sub_ps(rinv00,pmecorrV);
 464             velec            = _mm256_mul_ps(qq00,velec);
 465
 466             /* CUBIC SPLINE TABLE DISPERSION */
 467             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 468                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 469             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 470                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 471             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 472                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 473             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 474                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 475             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 476             Heps             = _mm256_mul_ps(vfeps,H);
 477             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 478             VV               = _mm256_add_ps(Y,_mm256_mul_ps(vfeps,Fp));
 479             vvdw6            = _mm256_mul_ps(c6_00,VV);
 480             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 481             fvdw6            = _mm256_mul_ps(c6_00,FF);
 482
 483             /* CUBIC SPLINE TABLE REPULSION */
 484             vfitab_lo        = _mm_add_epi32(vfitab_lo,ifour);
 485             vfitab_hi        = _mm_add_epi32(vfitab_hi,ifour);
 486             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 487                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 488             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 489                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 490             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 491                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 492             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 493                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 494             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 495             Heps             = _mm256_mul_ps(vfeps,H);
 496             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 497             VV               = _mm256_add_ps(Y,_mm256_mul_ps(vfeps,Fp));
 498             vvdw12           = _mm256_mul_ps(c12_00,VV);
 499             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 500             fvdw12           = _mm256_mul_ps(c12_00,FF);
 501             vvdw             = _mm256_add_ps(vvdw12,vvdw6);
 502             fvdw             = _mm256_xor_ps(signbit,_mm256_mul_ps(_mm256_add_ps(fvdw6,fvdw12),_mm256_mul_ps(vftabscale,rinv00)));
 503
 504             /* Update potential sum for this i atom from the interaction with this j atom. */
 505             velec            = _mm256_andnot_ps(dummy_mask,velec);
 506             velecsum         = _mm256_add_ps(velecsum,velec);
 507             vvdw             = _mm256_andnot_ps(dummy_mask,vvdw);
 508             vvdwsum          = _mm256_add_ps(vvdwsum,vvdw);
 509
 510             fscal            = _mm256_add_ps(felec,fvdw);
 511
 512             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 513
 514             /* Calculate temporary vectorial force */
 515             tx               = _mm256_mul_ps(fscal,dx00);
 516             ty               = _mm256_mul_ps(fscal,dy00);
 517             tz               = _mm256_mul_ps(fscal,dz00);
 518
 519             /* Update vectorial force */
 520             fix0             = _mm256_add_ps(fix0,tx);
 521             fiy0             = _mm256_add_ps(fiy0,ty);
 522             fiz0             = _mm256_add_ps(fiz0,tz);
 523
 524             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
 525             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
 526             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
 527             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
 528             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
 529             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
 530             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
 531             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
 532             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 533
 534             /* Inner loop uses 119 flops */
 535         }
 536
 537         /* End of innermost loop */
 538
 539         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
 540                                                  f+i_coord_offset,fshift+i_shift_offset);
 541
 542         ggid                        = gid[iidx];
 543         /* Update potential energies */
 544         gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
 545         gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
 546
 547         /* Increment number of inner iterations */
 548         inneriter                  += j_index_end - j_index_start;
 549
 550         /* Outer loop uses 9 flops */
 551     }
 552
 553     /* Increment number of outer iterations */
 554     outeriter        += nri;
 555
 556     /* Update outer/inner flops */
 557
 558     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*119);
 559 }
 560 /*
 561  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwCSTab_GeomP1P1_F_avx_256_single
 562  * Electrostatics interaction: Ewald
 563  * VdW interaction:            CubicSplineTable
 564  * Geometry:                   Particle-Particle
 565  * Calculate force/pot:        Force
 566  */
 567 void
 568 nb_kernel_ElecEw_VdwCSTab_GeomP1P1_F_avx_256_single
 569                     (t_nblist                    * gmx_restrict       nlist,
 570                      rvec                        * gmx_restrict          xx,
 571                      rvec                        * gmx_restrict          ff,
 572                      t_forcerec                  * gmx_restrict          fr,
 573                      t_mdatoms                   * gmx_restrict     mdatoms,
 574                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 575                      t_nrnb                      * gmx_restrict        nrnb)
 576 {
 577     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 578      * just 0 for non-waters.
 579      * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
 580      * jnr indices corresponding to data put in the four positions in the SIMD register.
 581      */
 582     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 583     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 584     int              jnrA,jnrB,jnrC,jnrD;
 585     int              jnrE,jnrF,jnrG,jnrH;
 586     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
 587     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
 588     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
 589     int              j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
 590     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 591     real             rcutoff_scalar;
 592     real             *shiftvec,*fshift,*x,*f;
 593     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
 594     real             scratch[4*DIM];
 595     __m256           tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 596     real *           vdwioffsetptr0;
 597     __m256           ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 598     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
 599     __m256           jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 600     __m256           dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 601     __m256           velec,felec,velecsum,facel,crf,krf,krf2;
 602     real             *charge;
 603     int              nvdwtype;
 604     __m256           rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 605     int              *vdwtype;
 606     real             *vdwparam;
 607     __m256           one_sixth   = _mm256_set1_ps(1.0/6.0);
 608     __m256           one_twelfth = _mm256_set1_ps(1.0/12.0);
 609     __m256i          vfitab;
 610     __m128i          vfitab_lo,vfitab_hi;
 611     __m128i          ifour       = _mm_set1_epi32(4);
 612     __m256           rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
 613     real             *vftab;
 614     __m256i          ewitab;
 615     __m128i          ewitab_lo,ewitab_hi;
 616     __m256           ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 617     __m256           beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 618     real             *ewtab;
 619     __m256           dummy_mask,cutoff_mask;
 620     __m256           signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
 621     __m256           one     = _mm256_set1_ps(1.0);
 622     __m256           two     = _mm256_set1_ps(2.0);
 623     x                = xx[0];
 624     f                = ff[0];
 625
 626     nri              = nlist->nri;
 627     iinr             = nlist->iinr;
 628     jindex           = nlist->jindex;
 629     jjnr             = nlist->jjnr;
 630     shiftidx         = nlist->shift;
 631     gid              = nlist->gid;
 632     shiftvec         = fr->shift_vec[0];
 633     fshift           = fr->fshift[0];
 634     facel            = _mm256_set1_ps(fr->epsfac);
 635     charge           = mdatoms->chargeA;
 636     nvdwtype         = fr->ntype;
 637     vdwparam         = fr->nbfp;
 638     vdwtype          = mdatoms->typeA;
 639
 640     vftab            = kernel_data->table_vdw->data;
 641     vftabscale       = _mm256_set1_ps(kernel_data->table_vdw->scale);
 642
 643     sh_ewald         = _mm256_set1_ps(fr->ic->sh_ewald);
 644     beta             = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
 645     beta2            = _mm256_mul_ps(beta,beta);
 646     beta3            = _mm256_mul_ps(beta,beta2);
 647
 648     ewtab            = fr->ic->tabq_coul_F;
 649     ewtabscale       = _mm256_set1_ps(fr->ic->tabq_scale);
 650     ewtabhalfspace   = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
 651
 652     /* Avoid stupid compiler warnings */
 653     jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
 654     j_coord_offsetA = 0;
 655     j_coord_offsetB = 0;
 656     j_coord_offsetC = 0;
 657     j_coord_offsetD = 0;
 658     j_coord_offsetE = 0;
 659     j_coord_offsetF = 0;
 660     j_coord_offsetG = 0;
 661     j_coord_offsetH = 0;
 662
 663     outeriter        = 0;
 664     inneriter        = 0;
 665
 666     for(iidx=0;iidx<4*DIM;iidx++)
 667     {
 668         scratch[iidx] = 0.0;
 669     }
 670
 671     /* Start outer loop over neighborlists */
 672     for(iidx=0; iidx<nri; iidx++)
 673     {
 674         /* Load shift vector for this list */
 675         i_shift_offset   = DIM*shiftidx[iidx];
 676
 677         /* Load limits for loop over neighbors */
 678         j_index_start    = jindex[iidx];
 679         j_index_end      = jindex[iidx+1];
 680
 681         /* Get outer coordinate index */
 682         inr              = iinr[iidx];
 683         i_coord_offset   = DIM*inr;
 684
 685         /* Load i particle coords and add shift vector */
 686         gmx_mm256_load_shift_and_1rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 687
 688         fix0             = _mm256_setzero_ps();
 689         fiy0             = _mm256_setzero_ps();
 690         fiz0             = _mm256_setzero_ps();
 691
 692         /* Load parameters for i particles */
 693         iq0              = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+0]));
 694         vdwioffsetptr0   = vdwparam+2*nvdwtype*vdwtype[inr+0];
 695
 696         /* Start inner kernel loop */
 697         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
 698         {
 699
 700             /* Get j neighbor index, and coordinate index */
 701             jnrA             = jjnr[jidx];
 702             jnrB             = jjnr[jidx+1];
 703             jnrC             = jjnr[jidx+2];
 704             jnrD             = jjnr[jidx+3];
 705             jnrE             = jjnr[jidx+4];
 706             jnrF             = jjnr[jidx+5];
 707             jnrG             = jjnr[jidx+6];
 708             jnrH             = jjnr[jidx+7];
 709             j_coord_offsetA  = DIM*jnrA;
 710             j_coord_offsetB  = DIM*jnrB;
 711             j_coord_offsetC  = DIM*jnrC;
 712             j_coord_offsetD  = DIM*jnrD;
 713             j_coord_offsetE  = DIM*jnrE;
 714             j_coord_offsetF  = DIM*jnrF;
 715             j_coord_offsetG  = DIM*jnrG;
 716             j_coord_offsetH  = DIM*jnrH;
 717
 718             /* load j atom coordinates */
 719             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 720                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 721                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 722                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 723                                                  &jx0,&jy0,&jz0);
 724
 725             /* Calculate displacement vector */
 726             dx00             = _mm256_sub_ps(ix0,jx0);
 727             dy00             = _mm256_sub_ps(iy0,jy0);
 728             dz00             = _mm256_sub_ps(iz0,jz0);
 729
 730             /* Calculate squared distance and things based on it */
 731             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 732
 733             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 734
 735             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 736
 737             /* Load parameters for j particles */
 738             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 739                                                                  charge+jnrC+0,charge+jnrD+0,
 740                                                                  charge+jnrE+0,charge+jnrF+0,
 741                                                                  charge+jnrG+0,charge+jnrH+0);
 742             vdwjidx0A        = 2*vdwtype[jnrA+0];
 743             vdwjidx0B        = 2*vdwtype[jnrB+0];
 744             vdwjidx0C        = 2*vdwtype[jnrC+0];
 745             vdwjidx0D        = 2*vdwtype[jnrD+0];
 746             vdwjidx0E        = 2*vdwtype[jnrE+0];
 747             vdwjidx0F        = 2*vdwtype[jnrF+0];
 748             vdwjidx0G        = 2*vdwtype[jnrG+0];
 749             vdwjidx0H        = 2*vdwtype[jnrH+0];
 750
 751             /**************************
 752              * CALCULATE INTERACTIONS *
 753              **************************/
 754
 755             r00              = _mm256_mul_ps(rsq00,rinv00);
 756
 757             /* Compute parameters for interactions between i and j atoms */
 758             qq00             = _mm256_mul_ps(iq0,jq0);
 759             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 760                                             vdwioffsetptr0+vdwjidx0B,
 761                                             vdwioffsetptr0+vdwjidx0C,
 762                                             vdwioffsetptr0+vdwjidx0D,
 763                                             vdwioffsetptr0+vdwjidx0E,
 764                                             vdwioffsetptr0+vdwjidx0F,
 765                                             vdwioffsetptr0+vdwjidx0G,
 766                                             vdwioffsetptr0+vdwjidx0H,
 767                                             &c6_00,&c12_00);
 768
 769             /* Calculate table index by multiplying r with table scale and truncate to integer */
 770             rt               = _mm256_mul_ps(r00,vftabscale);
 771             vfitab           = _mm256_cvttps_epi32(rt);
 772             vfeps            = _mm256_sub_ps(rt,_mm256_round_ps(rt, _MM_FROUND_FLOOR));
 773             /*         AVX1 does not support 256-bit integer operations, so now we go to 128-bit mode... */
 774             vfitab_lo        = _mm256_extractf128_si256(vfitab,0x0);
 775             vfitab_hi        = _mm256_extractf128_si256(vfitab,0x1);
 776             vfitab_lo        = _mm_slli_epi32(vfitab_lo,3);
 777             vfitab_hi        = _mm_slli_epi32(vfitab_hi,3);
 778
 779             /* EWALD ELECTROSTATICS */
 780
 781             /* Analytical PME correction */
 782             zeta2            = _mm256_mul_ps(beta2,rsq00);
 783             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 784             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 785             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 786             felec            = _mm256_mul_ps(qq00,felec);
 787
 788             /* CUBIC SPLINE TABLE DISPERSION */
 789             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 790                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 791             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 792                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 793             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 794                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 795             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 796                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 797             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 798             Heps             = _mm256_mul_ps(vfeps,H);
 799             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 800             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 801             fvdw6            = _mm256_mul_ps(c6_00,FF);
 802
 803             /* CUBIC SPLINE TABLE REPULSION */
 804             vfitab_lo        = _mm_add_epi32(vfitab_lo,ifour);
 805             vfitab_hi        = _mm_add_epi32(vfitab_hi,ifour);
 806             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 807                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 808             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 809                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 810             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 811                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 812             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 813                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 814             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 815             Heps             = _mm256_mul_ps(vfeps,H);
 816             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 817             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 818             fvdw12           = _mm256_mul_ps(c12_00,FF);
 819             fvdw             = _mm256_xor_ps(signbit,_mm256_mul_ps(_mm256_add_ps(fvdw6,fvdw12),_mm256_mul_ps(vftabscale,rinv00)));
 820
 821             fscal            = _mm256_add_ps(felec,fvdw);
 822
 823             /* Calculate temporary vectorial force */
 824             tx               = _mm256_mul_ps(fscal,dx00);
 825             ty               = _mm256_mul_ps(fscal,dy00);
 826             tz               = _mm256_mul_ps(fscal,dz00);
 827
 828             /* Update vectorial force */
 829             fix0             = _mm256_add_ps(fix0,tx);
 830             fiy0             = _mm256_add_ps(fiy0,ty);
 831             fiz0             = _mm256_add_ps(fiz0,tz);
 832
 833             fjptrA             = f+j_coord_offsetA;
 834             fjptrB             = f+j_coord_offsetB;
 835             fjptrC             = f+j_coord_offsetC;
 836             fjptrD             = f+j_coord_offsetD;
 837             fjptrE             = f+j_coord_offsetE;
 838             fjptrF             = f+j_coord_offsetF;
 839             fjptrG             = f+j_coord_offsetG;
 840             fjptrH             = f+j_coord_offsetH;
 841             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
 842
 843             /* Inner loop uses 82 flops */
 844         }
 845
 846         if(jidx<j_index_end)
 847         {
 848
 849             /* Get j neighbor index, and coordinate index */
 850             jnrlistA         = jjnr[jidx];
 851             jnrlistB         = jjnr[jidx+1];
 852             jnrlistC         = jjnr[jidx+2];
 853             jnrlistD         = jjnr[jidx+3];
 854             jnrlistE         = jjnr[jidx+4];
 855             jnrlistF         = jjnr[jidx+5];
 856             jnrlistG         = jjnr[jidx+6];
 857             jnrlistH         = jjnr[jidx+7];
 858             /* Sign of each element will be negative for non-real atoms.
 859              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 860              * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
 861              */
 862             dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
 863                                             gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
 864
 865             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 866             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 867             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 868             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 869             jnrE       = (jnrlistE>=0) ? jnrlistE : 0;
 870             jnrF       = (jnrlistF>=0) ? jnrlistF : 0;
 871             jnrG       = (jnrlistG>=0) ? jnrlistG : 0;
 872             jnrH       = (jnrlistH>=0) ? jnrlistH : 0;
 873             j_coord_offsetA  = DIM*jnrA;
 874             j_coord_offsetB  = DIM*jnrB;
 875             j_coord_offsetC  = DIM*jnrC;
 876             j_coord_offsetD  = DIM*jnrD;
 877             j_coord_offsetE  = DIM*jnrE;
 878             j_coord_offsetF  = DIM*jnrF;
 879             j_coord_offsetG  = DIM*jnrG;
 880             j_coord_offsetH  = DIM*jnrH;
 881
 882             /* load j atom coordinates */
 883             gmx_mm256_load_1rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
 884                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 885                                                  x+j_coord_offsetE,x+j_coord_offsetF,
 886                                                  x+j_coord_offsetG,x+j_coord_offsetH,
 887                                                  &jx0,&jy0,&jz0);
 888
 889             /* Calculate displacement vector */
 890             dx00             = _mm256_sub_ps(ix0,jx0);
 891             dy00             = _mm256_sub_ps(iy0,jy0);
 892             dz00             = _mm256_sub_ps(iz0,jz0);
 893
 894             /* Calculate squared distance and things based on it */
 895             rsq00            = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
 896
 897             rinv00           = gmx_mm256_invsqrt_ps(rsq00);
 898
 899             rinvsq00         = _mm256_mul_ps(rinv00,rinv00);
 900
 901             /* Load parameters for j particles */
 902             jq0              = gmx_mm256_load_8real_swizzle_ps(charge+jnrA+0,charge+jnrB+0,
 903                                                                  charge+jnrC+0,charge+jnrD+0,
 904                                                                  charge+jnrE+0,charge+jnrF+0,
 905                                                                  charge+jnrG+0,charge+jnrH+0);
 906             vdwjidx0A        = 2*vdwtype[jnrA+0];
 907             vdwjidx0B        = 2*vdwtype[jnrB+0];
 908             vdwjidx0C        = 2*vdwtype[jnrC+0];
 909             vdwjidx0D        = 2*vdwtype[jnrD+0];
 910             vdwjidx0E        = 2*vdwtype[jnrE+0];
 911             vdwjidx0F        = 2*vdwtype[jnrF+0];
 912             vdwjidx0G        = 2*vdwtype[jnrG+0];
 913             vdwjidx0H        = 2*vdwtype[jnrH+0];
 914
 915             /**************************
 916              * CALCULATE INTERACTIONS *
 917              **************************/
 918
 919             r00              = _mm256_mul_ps(rsq00,rinv00);
 920             r00              = _mm256_andnot_ps(dummy_mask,r00);
 921
 922             /* Compute parameters for interactions between i and j atoms */
 923             qq00             = _mm256_mul_ps(iq0,jq0);
 924             gmx_mm256_load_8pair_swizzle_ps(vdwioffsetptr0+vdwjidx0A,
 925                                             vdwioffsetptr0+vdwjidx0B,
 926                                             vdwioffsetptr0+vdwjidx0C,
 927                                             vdwioffsetptr0+vdwjidx0D,
 928                                             vdwioffsetptr0+vdwjidx0E,
 929                                             vdwioffsetptr0+vdwjidx0F,
 930                                             vdwioffsetptr0+vdwjidx0G,
 931                                             vdwioffsetptr0+vdwjidx0H,
 932                                             &c6_00,&c12_00);
 933
 934             /* Calculate table index by multiplying r with table scale and truncate to integer */
 935             rt               = _mm256_mul_ps(r00,vftabscale);
 936             vfitab           = _mm256_cvttps_epi32(rt);
 937             vfeps            = _mm256_sub_ps(rt,_mm256_round_ps(rt, _MM_FROUND_FLOOR));
 938             /*         AVX1 does not support 256-bit integer operations, so now we go to 128-bit mode... */
 939             vfitab_lo        = _mm256_extractf128_si256(vfitab,0x0);
 940             vfitab_hi        = _mm256_extractf128_si256(vfitab,0x1);
 941             vfitab_lo        = _mm_slli_epi32(vfitab_lo,3);
 942             vfitab_hi        = _mm_slli_epi32(vfitab_hi,3);
 943
 944             /* EWALD ELECTROSTATICS */
 945
 946             /* Analytical PME correction */
 947             zeta2            = _mm256_mul_ps(beta2,rsq00);
 948             rinv3            = _mm256_mul_ps(rinvsq00,rinv00);
 949             pmecorrF         = gmx_mm256_pmecorrF_ps(zeta2);
 950             felec            = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
 951             felec            = _mm256_mul_ps(qq00,felec);
 952
 953             /* CUBIC SPLINE TABLE DISPERSION */
 954             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 955                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 956             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 957                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 958             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 959                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 960             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 961                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 962             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 963             Heps             = _mm256_mul_ps(vfeps,H);
 964             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 965             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 966             fvdw6            = _mm256_mul_ps(c6_00,FF);
 967
 968             /* CUBIC SPLINE TABLE REPULSION */
 969             vfitab_lo        = _mm_add_epi32(vfitab_lo,ifour);
 970             vfitab_hi        = _mm_add_epi32(vfitab_hi,ifour);
 971             Y                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,0)),
 972                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,0)));
 973             F                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,1)),
 974                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,1)));
 975             G                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,2)),
 976                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,2)));
 977             H                = gmx_mm256_set_m128(_mm_load_ps(vftab + _mm_extract_epi32(vfitab_hi,3)),
 978                                                   _mm_load_ps(vftab + _mm_extract_epi32(vfitab_lo,3)));
 979             GMX_MM256_HALFTRANSPOSE4_PS(Y,F,G,H);
 980             Heps             = _mm256_mul_ps(vfeps,H);
 981             Fp               = _mm256_add_ps(F,_mm256_mul_ps(vfeps,_mm256_add_ps(G,Heps)));
 982             FF               = _mm256_add_ps(Fp,_mm256_mul_ps(vfeps,_mm256_add_ps(G,_mm256_add_ps(Heps,Heps))));
 983             fvdw12           = _mm256_mul_ps(c12_00,FF);
 984             fvdw             = _mm256_xor_ps(signbit,_mm256_mul_ps(_mm256_add_ps(fvdw6,fvdw12),_mm256_mul_ps(vftabscale,rinv00)));
 985
 986             fscal            = _mm256_add_ps(felec,fvdw);
 987
 988             fscal            = _mm256_andnot_ps(dummy_mask,fscal);
 989
 990             /* Calculate temporary vectorial force */
 991             tx               = _mm256_mul_ps(fscal,dx00);
 992             ty               = _mm256_mul_ps(fscal,dy00);
 993             tz               = _mm256_mul_ps(fscal,dz00);
 994
 995             /* Update vectorial force */
 996             fix0             = _mm256_add_ps(fix0,tx);
 997             fiy0             = _mm256_add_ps(fiy0,ty);
 998             fiz0             = _mm256_add_ps(fiz0,tz);
 999
1000             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1001             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1002             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1003             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1004             fjptrE             = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
1005             fjptrF             = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
1006             fjptrG             = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
1007             fjptrH             = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
1008             gmx_mm256_decrement_1rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,tx,ty,tz);
1009
1010             /* Inner loop uses 83 flops */
1011         }
1012
1013         /* End of innermost loop */
1014
1015         gmx_mm256_update_iforce_1atom_swizzle_ps(fix0,fiy0,fiz0,
1016                                                  f+i_coord_offset,fshift+i_shift_offset);
1017
1018         /* Increment number of inner iterations */
1019         inneriter                  += j_index_end - j_index_start;
1020
1021         /* Outer loop uses 7 flops */
1022     }
1023
1024     /* Increment number of outer iterations */
1025     outeriter        += nri;
1026
1027     /* Update outer/inner flops */
1028
1029     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*83);
1030 }