src/gromacs/gmxlib/nonbonded/nb_kernel_avx_256_double/nb_kernel_ElecEw_VdwNone_GeomW3W3_avx_256_double.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_double 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_double.h"
  50 #include "kernelutil_x86_avx_256_double.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwNone_GeomW3W3_VF_avx_256_double
  54  * Electrostatics interaction: Ewald
  55  * VdW interaction:            None
  56  * Geometry:                   Water3-Water3
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecEw_VdwNone_GeomW3W3_VF_avx_256_double
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB,jnrC,jnrD;
  77     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
  78     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
  79     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
  80     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  81     real             rcutoff_scalar;
  82     real             *shiftvec,*fshift,*x,*f;
  83     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
  84     real             scratch[4*DIM];
  85     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  86     real *           vdwioffsetptr0;
  87     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  88     real *           vdwioffsetptr1;
  89     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  90     real *           vdwioffsetptr2;
  91     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  92     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
  93     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  94     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
  95     __m256d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
  96     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
  97     __m256d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
  98     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  99     __m256d          dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
 100     __m256d          dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
 101     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 102     __m256d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
 103     __m256d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
 104     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 105     __m256d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 106     __m256d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 107     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
 108     real             *charge;
 109     __m128i          ewitab;
 110     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 111     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
 112     real             *ewtab;
 113     __m256d          dummy_mask,cutoff_mask;
 114     __m128           tmpmask0,tmpmask1;
 115     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
 116     __m256d          one     = _mm256_set1_pd(1.0);
 117     __m256d          two     = _mm256_set1_pd(2.0);
 118     x                = xx[0];
 119     f                = ff[0];
 120
 121     nri              = nlist->nri;
 122     iinr             = nlist->iinr;
 123     jindex           = nlist->jindex;
 124     jjnr             = nlist->jjnr;
 125     shiftidx         = nlist->shift;
 126     gid              = nlist->gid;
 127     shiftvec         = fr->shift_vec[0];
 128     fshift           = fr->fshift[0];
 129     facel            = _mm256_set1_pd(fr->epsfac);
 130     charge           = mdatoms->chargeA;
 131
 132     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
 133     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
 134     beta2            = _mm256_mul_pd(beta,beta);
 135     beta3            = _mm256_mul_pd(beta,beta2);
 136
 137     ewtab            = fr->ic->tabq_coul_FDV0;
 138     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
 139     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
 140
 141     /* Setup water-specific parameters */
 142     inr              = nlist->iinr[0];
 143     iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
 144     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
 145     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
 146
 147     jq0              = _mm256_set1_pd(charge[inr+0]);
 148     jq1              = _mm256_set1_pd(charge[inr+1]);
 149     jq2              = _mm256_set1_pd(charge[inr+2]);
 150     qq00             = _mm256_mul_pd(iq0,jq0);
 151     qq01             = _mm256_mul_pd(iq0,jq1);
 152     qq02             = _mm256_mul_pd(iq0,jq2);
 153     qq10             = _mm256_mul_pd(iq1,jq0);
 154     qq11             = _mm256_mul_pd(iq1,jq1);
 155     qq12             = _mm256_mul_pd(iq1,jq2);
 156     qq20             = _mm256_mul_pd(iq2,jq0);
 157     qq21             = _mm256_mul_pd(iq2,jq1);
 158     qq22             = _mm256_mul_pd(iq2,jq2);
 159
 160     /* Avoid stupid compiler warnings */
 161     jnrA = jnrB = jnrC = jnrD = 0;
 162     j_coord_offsetA = 0;
 163     j_coord_offsetB = 0;
 164     j_coord_offsetC = 0;
 165     j_coord_offsetD = 0;
 166
 167     outeriter        = 0;
 168     inneriter        = 0;
 169
 170     for(iidx=0;iidx<4*DIM;iidx++)
 171     {
 172         scratch[iidx] = 0.0;
 173     }
 174
 175     /* Start outer loop over neighborlists */
 176     for(iidx=0; iidx<nri; iidx++)
 177     {
 178         /* Load shift vector for this list */
 179         i_shift_offset   = DIM*shiftidx[iidx];
 180
 181         /* Load limits for loop over neighbors */
 182         j_index_start    = jindex[iidx];
 183         j_index_end      = jindex[iidx+1];
 184
 185         /* Get outer coordinate index */
 186         inr              = iinr[iidx];
 187         i_coord_offset   = DIM*inr;
 188
 189         /* Load i particle coords and add shift vector */
 190         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 191                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
 192
 193         fix0             = _mm256_setzero_pd();
 194         fiy0             = _mm256_setzero_pd();
 195         fiz0             = _mm256_setzero_pd();
 196         fix1             = _mm256_setzero_pd();
 197         fiy1             = _mm256_setzero_pd();
 198         fiz1             = _mm256_setzero_pd();
 199         fix2             = _mm256_setzero_pd();
 200         fiy2             = _mm256_setzero_pd();
 201         fiz2             = _mm256_setzero_pd();
 202
 203         /* Reset potential sums */
 204         velecsum         = _mm256_setzero_pd();
 205
 206         /* Start inner kernel loop */
 207         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
 208         {
 209
 210             /* Get j neighbor index, and coordinate index */
 211             jnrA             = jjnr[jidx];
 212             jnrB             = jjnr[jidx+1];
 213             jnrC             = jjnr[jidx+2];
 214             jnrD             = jjnr[jidx+3];
 215             j_coord_offsetA  = DIM*jnrA;
 216             j_coord_offsetB  = DIM*jnrB;
 217             j_coord_offsetC  = DIM*jnrC;
 218             j_coord_offsetD  = DIM*jnrD;
 219
 220             /* load j atom coordinates */
 221             gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 222                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 223                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 224
 225             /* Calculate displacement vector */
 226             dx00             = _mm256_sub_pd(ix0,jx0);
 227             dy00             = _mm256_sub_pd(iy0,jy0);
 228             dz00             = _mm256_sub_pd(iz0,jz0);
 229             dx01             = _mm256_sub_pd(ix0,jx1);
 230             dy01             = _mm256_sub_pd(iy0,jy1);
 231             dz01             = _mm256_sub_pd(iz0,jz1);
 232             dx02             = _mm256_sub_pd(ix0,jx2);
 233             dy02             = _mm256_sub_pd(iy0,jy2);
 234             dz02             = _mm256_sub_pd(iz0,jz2);
 235             dx10             = _mm256_sub_pd(ix1,jx0);
 236             dy10             = _mm256_sub_pd(iy1,jy0);
 237             dz10             = _mm256_sub_pd(iz1,jz0);
 238             dx11             = _mm256_sub_pd(ix1,jx1);
 239             dy11             = _mm256_sub_pd(iy1,jy1);
 240             dz11             = _mm256_sub_pd(iz1,jz1);
 241             dx12             = _mm256_sub_pd(ix1,jx2);
 242             dy12             = _mm256_sub_pd(iy1,jy2);
 243             dz12             = _mm256_sub_pd(iz1,jz2);
 244             dx20             = _mm256_sub_pd(ix2,jx0);
 245             dy20             = _mm256_sub_pd(iy2,jy0);
 246             dz20             = _mm256_sub_pd(iz2,jz0);
 247             dx21             = _mm256_sub_pd(ix2,jx1);
 248             dy21             = _mm256_sub_pd(iy2,jy1);
 249             dz21             = _mm256_sub_pd(iz2,jz1);
 250             dx22             = _mm256_sub_pd(ix2,jx2);
 251             dy22             = _mm256_sub_pd(iy2,jy2);
 252             dz22             = _mm256_sub_pd(iz2,jz2);
 253
 254             /* Calculate squared distance and things based on it */
 255             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 256             rsq01            = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
 257             rsq02            = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
 258             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 259             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
 260             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
 261             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 262             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
 263             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
 264
 265             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 266             rinv01           = gmx_mm256_invsqrt_pd(rsq01);
 267             rinv02           = gmx_mm256_invsqrt_pd(rsq02);
 268             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 269             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
 270             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
 271             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 272             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
 273             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
 274
 275             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 276             rinvsq01         = _mm256_mul_pd(rinv01,rinv01);
 277             rinvsq02         = _mm256_mul_pd(rinv02,rinv02);
 278             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 279             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
 280             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
 281             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 282             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
 283             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
 284
 285             fjx0             = _mm256_setzero_pd();
 286             fjy0             = _mm256_setzero_pd();
 287             fjz0             = _mm256_setzero_pd();
 288             fjx1             = _mm256_setzero_pd();
 289             fjy1             = _mm256_setzero_pd();
 290             fjz1             = _mm256_setzero_pd();
 291             fjx2             = _mm256_setzero_pd();
 292             fjy2             = _mm256_setzero_pd();
 293             fjz2             = _mm256_setzero_pd();
 294
 295             /**************************
 296              * CALCULATE INTERACTIONS *
 297              **************************/
 298
 299             r00              = _mm256_mul_pd(rsq00,rinv00);
 300
 301             /* EWALD ELECTROSTATICS */
 302
 303             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 304             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 305             ewitab           = _mm256_cvttpd_epi32(ewrt);
 306             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 307             ewitab           = _mm_slli_epi32(ewitab,2);
 308             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 309             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 310             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 311             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 312             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 313             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 314             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 315             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(rinv00,velec));
 316             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 317
 318             /* Update potential sum for this i atom from the interaction with this j atom. */
 319             velecsum         = _mm256_add_pd(velecsum,velec);
 320
 321             fscal            = felec;
 322
 323             /* Calculate temporary vectorial force */
 324             tx               = _mm256_mul_pd(fscal,dx00);
 325             ty               = _mm256_mul_pd(fscal,dy00);
 326             tz               = _mm256_mul_pd(fscal,dz00);
 327
 328             /* Update vectorial force */
 329             fix0             = _mm256_add_pd(fix0,tx);
 330             fiy0             = _mm256_add_pd(fiy0,ty);
 331             fiz0             = _mm256_add_pd(fiz0,tz);
 332
 333             fjx0             = _mm256_add_pd(fjx0,tx);
 334             fjy0             = _mm256_add_pd(fjy0,ty);
 335             fjz0             = _mm256_add_pd(fjz0,tz);
 336
 337             /**************************
 338              * CALCULATE INTERACTIONS *
 339              **************************/
 340
 341             r01              = _mm256_mul_pd(rsq01,rinv01);
 342
 343             /* EWALD ELECTROSTATICS */
 344
 345             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 346             ewrt             = _mm256_mul_pd(r01,ewtabscale);
 347             ewitab           = _mm256_cvttpd_epi32(ewrt);
 348             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 349             ewitab           = _mm_slli_epi32(ewitab,2);
 350             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 351             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 352             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 353             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 354             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 355             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 356             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 357             velec            = _mm256_mul_pd(qq01,_mm256_sub_pd(rinv01,velec));
 358             felec            = _mm256_mul_pd(_mm256_mul_pd(qq01,rinv01),_mm256_sub_pd(rinvsq01,felec));
 359
 360             /* Update potential sum for this i atom from the interaction with this j atom. */
 361             velecsum         = _mm256_add_pd(velecsum,velec);
 362
 363             fscal            = felec;
 364
 365             /* Calculate temporary vectorial force */
 366             tx               = _mm256_mul_pd(fscal,dx01);
 367             ty               = _mm256_mul_pd(fscal,dy01);
 368             tz               = _mm256_mul_pd(fscal,dz01);
 369
 370             /* Update vectorial force */
 371             fix0             = _mm256_add_pd(fix0,tx);
 372             fiy0             = _mm256_add_pd(fiy0,ty);
 373             fiz0             = _mm256_add_pd(fiz0,tz);
 374
 375             fjx1             = _mm256_add_pd(fjx1,tx);
 376             fjy1             = _mm256_add_pd(fjy1,ty);
 377             fjz1             = _mm256_add_pd(fjz1,tz);
 378
 379             /**************************
 380              * CALCULATE INTERACTIONS *
 381              **************************/
 382
 383             r02              = _mm256_mul_pd(rsq02,rinv02);
 384
 385             /* EWALD ELECTROSTATICS */
 386
 387             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 388             ewrt             = _mm256_mul_pd(r02,ewtabscale);
 389             ewitab           = _mm256_cvttpd_epi32(ewrt);
 390             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 391             ewitab           = _mm_slli_epi32(ewitab,2);
 392             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 393             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 394             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 395             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 396             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 397             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 398             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 399             velec            = _mm256_mul_pd(qq02,_mm256_sub_pd(rinv02,velec));
 400             felec            = _mm256_mul_pd(_mm256_mul_pd(qq02,rinv02),_mm256_sub_pd(rinvsq02,felec));
 401
 402             /* Update potential sum for this i atom from the interaction with this j atom. */
 403             velecsum         = _mm256_add_pd(velecsum,velec);
 404
 405             fscal            = felec;
 406
 407             /* Calculate temporary vectorial force */
 408             tx               = _mm256_mul_pd(fscal,dx02);
 409             ty               = _mm256_mul_pd(fscal,dy02);
 410             tz               = _mm256_mul_pd(fscal,dz02);
 411
 412             /* Update vectorial force */
 413             fix0             = _mm256_add_pd(fix0,tx);
 414             fiy0             = _mm256_add_pd(fiy0,ty);
 415             fiz0             = _mm256_add_pd(fiz0,tz);
 416
 417             fjx2             = _mm256_add_pd(fjx2,tx);
 418             fjy2             = _mm256_add_pd(fjy2,ty);
 419             fjz2             = _mm256_add_pd(fjz2,tz);
 420
 421             /**************************
 422              * CALCULATE INTERACTIONS *
 423              **************************/
 424
 425             r10              = _mm256_mul_pd(rsq10,rinv10);
 426
 427             /* EWALD ELECTROSTATICS */
 428
 429             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 430             ewrt             = _mm256_mul_pd(r10,ewtabscale);
 431             ewitab           = _mm256_cvttpd_epi32(ewrt);
 432             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 433             ewitab           = _mm_slli_epi32(ewitab,2);
 434             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 435             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 436             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 437             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 438             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 439             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 440             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 441             velec            = _mm256_mul_pd(qq10,_mm256_sub_pd(rinv10,velec));
 442             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
 443
 444             /* Update potential sum for this i atom from the interaction with this j atom. */
 445             velecsum         = _mm256_add_pd(velecsum,velec);
 446
 447             fscal            = felec;
 448
 449             /* Calculate temporary vectorial force */
 450             tx               = _mm256_mul_pd(fscal,dx10);
 451             ty               = _mm256_mul_pd(fscal,dy10);
 452             tz               = _mm256_mul_pd(fscal,dz10);
 453
 454             /* Update vectorial force */
 455             fix1             = _mm256_add_pd(fix1,tx);
 456             fiy1             = _mm256_add_pd(fiy1,ty);
 457             fiz1             = _mm256_add_pd(fiz1,tz);
 458
 459             fjx0             = _mm256_add_pd(fjx0,tx);
 460             fjy0             = _mm256_add_pd(fjy0,ty);
 461             fjz0             = _mm256_add_pd(fjz0,tz);
 462
 463             /**************************
 464              * CALCULATE INTERACTIONS *
 465              **************************/
 466
 467             r11              = _mm256_mul_pd(rsq11,rinv11);
 468
 469             /* EWALD ELECTROSTATICS */
 470
 471             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 472             ewrt             = _mm256_mul_pd(r11,ewtabscale);
 473             ewitab           = _mm256_cvttpd_epi32(ewrt);
 474             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 475             ewitab           = _mm_slli_epi32(ewitab,2);
 476             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 477             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 478             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 479             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 480             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 481             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 482             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 483             velec            = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
 484             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
 485
 486             /* Update potential sum for this i atom from the interaction with this j atom. */
 487             velecsum         = _mm256_add_pd(velecsum,velec);
 488
 489             fscal            = felec;
 490
 491             /* Calculate temporary vectorial force */
 492             tx               = _mm256_mul_pd(fscal,dx11);
 493             ty               = _mm256_mul_pd(fscal,dy11);
 494             tz               = _mm256_mul_pd(fscal,dz11);
 495
 496             /* Update vectorial force */
 497             fix1             = _mm256_add_pd(fix1,tx);
 498             fiy1             = _mm256_add_pd(fiy1,ty);
 499             fiz1             = _mm256_add_pd(fiz1,tz);
 500
 501             fjx1             = _mm256_add_pd(fjx1,tx);
 502             fjy1             = _mm256_add_pd(fjy1,ty);
 503             fjz1             = _mm256_add_pd(fjz1,tz);
 504
 505             /**************************
 506              * CALCULATE INTERACTIONS *
 507              **************************/
 508
 509             r12              = _mm256_mul_pd(rsq12,rinv12);
 510
 511             /* EWALD ELECTROSTATICS */
 512
 513             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 514             ewrt             = _mm256_mul_pd(r12,ewtabscale);
 515             ewitab           = _mm256_cvttpd_epi32(ewrt);
 516             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 517             ewitab           = _mm_slli_epi32(ewitab,2);
 518             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 519             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 520             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 521             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 522             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 523             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 524             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 525             velec            = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
 526             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
 527
 528             /* Update potential sum for this i atom from the interaction with this j atom. */
 529             velecsum         = _mm256_add_pd(velecsum,velec);
 530
 531             fscal            = felec;
 532
 533             /* Calculate temporary vectorial force */
 534             tx               = _mm256_mul_pd(fscal,dx12);
 535             ty               = _mm256_mul_pd(fscal,dy12);
 536             tz               = _mm256_mul_pd(fscal,dz12);
 537
 538             /* Update vectorial force */
 539             fix1             = _mm256_add_pd(fix1,tx);
 540             fiy1             = _mm256_add_pd(fiy1,ty);
 541             fiz1             = _mm256_add_pd(fiz1,tz);
 542
 543             fjx2             = _mm256_add_pd(fjx2,tx);
 544             fjy2             = _mm256_add_pd(fjy2,ty);
 545             fjz2             = _mm256_add_pd(fjz2,tz);
 546
 547             /**************************
 548              * CALCULATE INTERACTIONS *
 549              **************************/
 550
 551             r20              = _mm256_mul_pd(rsq20,rinv20);
 552
 553             /* EWALD ELECTROSTATICS */
 554
 555             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 556             ewrt             = _mm256_mul_pd(r20,ewtabscale);
 557             ewitab           = _mm256_cvttpd_epi32(ewrt);
 558             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 559             ewitab           = _mm_slli_epi32(ewitab,2);
 560             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 561             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 562             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 563             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 564             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 565             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 566             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 567             velec            = _mm256_mul_pd(qq20,_mm256_sub_pd(rinv20,velec));
 568             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
 569
 570             /* Update potential sum for this i atom from the interaction with this j atom. */
 571             velecsum         = _mm256_add_pd(velecsum,velec);
 572
 573             fscal            = felec;
 574
 575             /* Calculate temporary vectorial force */
 576             tx               = _mm256_mul_pd(fscal,dx20);
 577             ty               = _mm256_mul_pd(fscal,dy20);
 578             tz               = _mm256_mul_pd(fscal,dz20);
 579
 580             /* Update vectorial force */
 581             fix2             = _mm256_add_pd(fix2,tx);
 582             fiy2             = _mm256_add_pd(fiy2,ty);
 583             fiz2             = _mm256_add_pd(fiz2,tz);
 584
 585             fjx0             = _mm256_add_pd(fjx0,tx);
 586             fjy0             = _mm256_add_pd(fjy0,ty);
 587             fjz0             = _mm256_add_pd(fjz0,tz);
 588
 589             /**************************
 590              * CALCULATE INTERACTIONS *
 591              **************************/
 592
 593             r21              = _mm256_mul_pd(rsq21,rinv21);
 594
 595             /* EWALD ELECTROSTATICS */
 596
 597             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 598             ewrt             = _mm256_mul_pd(r21,ewtabscale);
 599             ewitab           = _mm256_cvttpd_epi32(ewrt);
 600             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 601             ewitab           = _mm_slli_epi32(ewitab,2);
 602             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 603             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 604             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 605             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 606             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 607             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 608             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 609             velec            = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
 610             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
 611
 612             /* Update potential sum for this i atom from the interaction with this j atom. */
 613             velecsum         = _mm256_add_pd(velecsum,velec);
 614
 615             fscal            = felec;
 616
 617             /* Calculate temporary vectorial force */
 618             tx               = _mm256_mul_pd(fscal,dx21);
 619             ty               = _mm256_mul_pd(fscal,dy21);
 620             tz               = _mm256_mul_pd(fscal,dz21);
 621
 622             /* Update vectorial force */
 623             fix2             = _mm256_add_pd(fix2,tx);
 624             fiy2             = _mm256_add_pd(fiy2,ty);
 625             fiz2             = _mm256_add_pd(fiz2,tz);
 626
 627             fjx1             = _mm256_add_pd(fjx1,tx);
 628             fjy1             = _mm256_add_pd(fjy1,ty);
 629             fjz1             = _mm256_add_pd(fjz1,tz);
 630
 631             /**************************
 632              * CALCULATE INTERACTIONS *
 633              **************************/
 634
 635             r22              = _mm256_mul_pd(rsq22,rinv22);
 636
 637             /* EWALD ELECTROSTATICS */
 638
 639             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 640             ewrt             = _mm256_mul_pd(r22,ewtabscale);
 641             ewitab           = _mm256_cvttpd_epi32(ewrt);
 642             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 643             ewitab           = _mm_slli_epi32(ewitab,2);
 644             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 645             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 646             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 647             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 648             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 649             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 650             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 651             velec            = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
 652             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
 653
 654             /* Update potential sum for this i atom from the interaction with this j atom. */
 655             velecsum         = _mm256_add_pd(velecsum,velec);
 656
 657             fscal            = felec;
 658
 659             /* Calculate temporary vectorial force */
 660             tx               = _mm256_mul_pd(fscal,dx22);
 661             ty               = _mm256_mul_pd(fscal,dy22);
 662             tz               = _mm256_mul_pd(fscal,dz22);
 663
 664             /* Update vectorial force */
 665             fix2             = _mm256_add_pd(fix2,tx);
 666             fiy2             = _mm256_add_pd(fiy2,ty);
 667             fiz2             = _mm256_add_pd(fiz2,tz);
 668
 669             fjx2             = _mm256_add_pd(fjx2,tx);
 670             fjy2             = _mm256_add_pd(fjy2,ty);
 671             fjz2             = _mm256_add_pd(fjz2,tz);
 672
 673             fjptrA             = f+j_coord_offsetA;
 674             fjptrB             = f+j_coord_offsetB;
 675             fjptrC             = f+j_coord_offsetC;
 676             fjptrD             = f+j_coord_offsetD;
 677
 678             gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
 679                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
 680
 681             /* Inner loop uses 369 flops */
 682         }
 683
 684         if(jidx<j_index_end)
 685         {
 686
 687             /* Get j neighbor index, and coordinate index */
 688             jnrlistA         = jjnr[jidx];
 689             jnrlistB         = jjnr[jidx+1];
 690             jnrlistC         = jjnr[jidx+2];
 691             jnrlistD         = jjnr[jidx+3];
 692             /* Sign of each element will be negative for non-real atoms.
 693              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
 694              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
 695              */
 696             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
 697
 698             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
 699             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
 700             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
 701
 702             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
 703             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
 704             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
 705             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
 706             j_coord_offsetA  = DIM*jnrA;
 707             j_coord_offsetB  = DIM*jnrB;
 708             j_coord_offsetC  = DIM*jnrC;
 709             j_coord_offsetD  = DIM*jnrD;
 710
 711             /* load j atom coordinates */
 712             gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 713                                                  x+j_coord_offsetC,x+j_coord_offsetD,
 714                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
 715
 716             /* Calculate displacement vector */
 717             dx00             = _mm256_sub_pd(ix0,jx0);
 718             dy00             = _mm256_sub_pd(iy0,jy0);
 719             dz00             = _mm256_sub_pd(iz0,jz0);
 720             dx01             = _mm256_sub_pd(ix0,jx1);
 721             dy01             = _mm256_sub_pd(iy0,jy1);
 722             dz01             = _mm256_sub_pd(iz0,jz1);
 723             dx02             = _mm256_sub_pd(ix0,jx2);
 724             dy02             = _mm256_sub_pd(iy0,jy2);
 725             dz02             = _mm256_sub_pd(iz0,jz2);
 726             dx10             = _mm256_sub_pd(ix1,jx0);
 727             dy10             = _mm256_sub_pd(iy1,jy0);
 728             dz10             = _mm256_sub_pd(iz1,jz0);
 729             dx11             = _mm256_sub_pd(ix1,jx1);
 730             dy11             = _mm256_sub_pd(iy1,jy1);
 731             dz11             = _mm256_sub_pd(iz1,jz1);
 732             dx12             = _mm256_sub_pd(ix1,jx2);
 733             dy12             = _mm256_sub_pd(iy1,jy2);
 734             dz12             = _mm256_sub_pd(iz1,jz2);
 735             dx20             = _mm256_sub_pd(ix2,jx0);
 736             dy20             = _mm256_sub_pd(iy2,jy0);
 737             dz20             = _mm256_sub_pd(iz2,jz0);
 738             dx21             = _mm256_sub_pd(ix2,jx1);
 739             dy21             = _mm256_sub_pd(iy2,jy1);
 740             dz21             = _mm256_sub_pd(iz2,jz1);
 741             dx22             = _mm256_sub_pd(ix2,jx2);
 742             dy22             = _mm256_sub_pd(iy2,jy2);
 743             dz22             = _mm256_sub_pd(iz2,jz2);
 744
 745             /* Calculate squared distance and things based on it */
 746             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
 747             rsq01            = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
 748             rsq02            = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
 749             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
 750             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
 751             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
 752             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
 753             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
 754             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
 755
 756             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
 757             rinv01           = gmx_mm256_invsqrt_pd(rsq01);
 758             rinv02           = gmx_mm256_invsqrt_pd(rsq02);
 759             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
 760             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
 761             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
 762             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
 763             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
 764             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
 765
 766             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
 767             rinvsq01         = _mm256_mul_pd(rinv01,rinv01);
 768             rinvsq02         = _mm256_mul_pd(rinv02,rinv02);
 769             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
 770             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
 771             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
 772             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
 773             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
 774             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
 775
 776             fjx0             = _mm256_setzero_pd();
 777             fjy0             = _mm256_setzero_pd();
 778             fjz0             = _mm256_setzero_pd();
 779             fjx1             = _mm256_setzero_pd();
 780             fjy1             = _mm256_setzero_pd();
 781             fjz1             = _mm256_setzero_pd();
 782             fjx2             = _mm256_setzero_pd();
 783             fjy2             = _mm256_setzero_pd();
 784             fjz2             = _mm256_setzero_pd();
 785
 786             /**************************
 787              * CALCULATE INTERACTIONS *
 788              **************************/
 789
 790             r00              = _mm256_mul_pd(rsq00,rinv00);
 791             r00              = _mm256_andnot_pd(dummy_mask,r00);
 792
 793             /* EWALD ELECTROSTATICS */
 794
 795             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 796             ewrt             = _mm256_mul_pd(r00,ewtabscale);
 797             ewitab           = _mm256_cvttpd_epi32(ewrt);
 798             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 799             ewitab           = _mm_slli_epi32(ewitab,2);
 800             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 801             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 802             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 803             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 804             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 805             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 806             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 807             velec            = _mm256_mul_pd(qq00,_mm256_sub_pd(rinv00,velec));
 808             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
 809
 810             /* Update potential sum for this i atom from the interaction with this j atom. */
 811             velec            = _mm256_andnot_pd(dummy_mask,velec);
 812             velecsum         = _mm256_add_pd(velecsum,velec);
 813
 814             fscal            = felec;
 815
 816             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 817
 818             /* Calculate temporary vectorial force */
 819             tx               = _mm256_mul_pd(fscal,dx00);
 820             ty               = _mm256_mul_pd(fscal,dy00);
 821             tz               = _mm256_mul_pd(fscal,dz00);
 822
 823             /* Update vectorial force */
 824             fix0             = _mm256_add_pd(fix0,tx);
 825             fiy0             = _mm256_add_pd(fiy0,ty);
 826             fiz0             = _mm256_add_pd(fiz0,tz);
 827
 828             fjx0             = _mm256_add_pd(fjx0,tx);
 829             fjy0             = _mm256_add_pd(fjy0,ty);
 830             fjz0             = _mm256_add_pd(fjz0,tz);
 831
 832             /**************************
 833              * CALCULATE INTERACTIONS *
 834              **************************/
 835
 836             r01              = _mm256_mul_pd(rsq01,rinv01);
 837             r01              = _mm256_andnot_pd(dummy_mask,r01);
 838
 839             /* EWALD ELECTROSTATICS */
 840
 841             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 842             ewrt             = _mm256_mul_pd(r01,ewtabscale);
 843             ewitab           = _mm256_cvttpd_epi32(ewrt);
 844             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 845             ewitab           = _mm_slli_epi32(ewitab,2);
 846             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 847             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 848             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 849             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 850             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 851             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 852             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 853             velec            = _mm256_mul_pd(qq01,_mm256_sub_pd(rinv01,velec));
 854             felec            = _mm256_mul_pd(_mm256_mul_pd(qq01,rinv01),_mm256_sub_pd(rinvsq01,felec));
 855
 856             /* Update potential sum for this i atom from the interaction with this j atom. */
 857             velec            = _mm256_andnot_pd(dummy_mask,velec);
 858             velecsum         = _mm256_add_pd(velecsum,velec);
 859
 860             fscal            = felec;
 861
 862             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 863
 864             /* Calculate temporary vectorial force */
 865             tx               = _mm256_mul_pd(fscal,dx01);
 866             ty               = _mm256_mul_pd(fscal,dy01);
 867             tz               = _mm256_mul_pd(fscal,dz01);
 868
 869             /* Update vectorial force */
 870             fix0             = _mm256_add_pd(fix0,tx);
 871             fiy0             = _mm256_add_pd(fiy0,ty);
 872             fiz0             = _mm256_add_pd(fiz0,tz);
 873
 874             fjx1             = _mm256_add_pd(fjx1,tx);
 875             fjy1             = _mm256_add_pd(fjy1,ty);
 876             fjz1             = _mm256_add_pd(fjz1,tz);
 877
 878             /**************************
 879              * CALCULATE INTERACTIONS *
 880              **************************/
 881
 882             r02              = _mm256_mul_pd(rsq02,rinv02);
 883             r02              = _mm256_andnot_pd(dummy_mask,r02);
 884
 885             /* EWALD ELECTROSTATICS */
 886
 887             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 888             ewrt             = _mm256_mul_pd(r02,ewtabscale);
 889             ewitab           = _mm256_cvttpd_epi32(ewrt);
 890             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 891             ewitab           = _mm_slli_epi32(ewitab,2);
 892             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 893             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 894             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 895             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 896             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 897             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 898             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 899             velec            = _mm256_mul_pd(qq02,_mm256_sub_pd(rinv02,velec));
 900             felec            = _mm256_mul_pd(_mm256_mul_pd(qq02,rinv02),_mm256_sub_pd(rinvsq02,felec));
 901
 902             /* Update potential sum for this i atom from the interaction with this j atom. */
 903             velec            = _mm256_andnot_pd(dummy_mask,velec);
 904             velecsum         = _mm256_add_pd(velecsum,velec);
 905
 906             fscal            = felec;
 907
 908             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 909
 910             /* Calculate temporary vectorial force */
 911             tx               = _mm256_mul_pd(fscal,dx02);
 912             ty               = _mm256_mul_pd(fscal,dy02);
 913             tz               = _mm256_mul_pd(fscal,dz02);
 914
 915             /* Update vectorial force */
 916             fix0             = _mm256_add_pd(fix0,tx);
 917             fiy0             = _mm256_add_pd(fiy0,ty);
 918             fiz0             = _mm256_add_pd(fiz0,tz);
 919
 920             fjx2             = _mm256_add_pd(fjx2,tx);
 921             fjy2             = _mm256_add_pd(fjy2,ty);
 922             fjz2             = _mm256_add_pd(fjz2,tz);
 923
 924             /**************************
 925              * CALCULATE INTERACTIONS *
 926              **************************/
 927
 928             r10              = _mm256_mul_pd(rsq10,rinv10);
 929             r10              = _mm256_andnot_pd(dummy_mask,r10);
 930
 931             /* EWALD ELECTROSTATICS */
 932
 933             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 934             ewrt             = _mm256_mul_pd(r10,ewtabscale);
 935             ewitab           = _mm256_cvttpd_epi32(ewrt);
 936             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 937             ewitab           = _mm_slli_epi32(ewitab,2);
 938             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 939             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 940             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 941             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 942             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 943             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 944             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 945             velec            = _mm256_mul_pd(qq10,_mm256_sub_pd(rinv10,velec));
 946             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
 947
 948             /* Update potential sum for this i atom from the interaction with this j atom. */
 949             velec            = _mm256_andnot_pd(dummy_mask,velec);
 950             velecsum         = _mm256_add_pd(velecsum,velec);
 951
 952             fscal            = felec;
 953
 954             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
 955
 956             /* Calculate temporary vectorial force */
 957             tx               = _mm256_mul_pd(fscal,dx10);
 958             ty               = _mm256_mul_pd(fscal,dy10);
 959             tz               = _mm256_mul_pd(fscal,dz10);
 960
 961             /* Update vectorial force */
 962             fix1             = _mm256_add_pd(fix1,tx);
 963             fiy1             = _mm256_add_pd(fiy1,ty);
 964             fiz1             = _mm256_add_pd(fiz1,tz);
 965
 966             fjx0             = _mm256_add_pd(fjx0,tx);
 967             fjy0             = _mm256_add_pd(fjy0,ty);
 968             fjz0             = _mm256_add_pd(fjz0,tz);
 969
 970             /**************************
 971              * CALCULATE INTERACTIONS *
 972              **************************/
 973
 974             r11              = _mm256_mul_pd(rsq11,rinv11);
 975             r11              = _mm256_andnot_pd(dummy_mask,r11);
 976
 977             /* EWALD ELECTROSTATICS */
 978
 979             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 980             ewrt             = _mm256_mul_pd(r11,ewtabscale);
 981             ewitab           = _mm256_cvttpd_epi32(ewrt);
 982             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
 983             ewitab           = _mm_slli_epi32(ewitab,2);
 984             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 985             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 986             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
 987             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
 988             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
 989             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
 990             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
 991             velec            = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
 992             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
 993
 994             /* Update potential sum for this i atom from the interaction with this j atom. */
 995             velec            = _mm256_andnot_pd(dummy_mask,velec);
 996             velecsum         = _mm256_add_pd(velecsum,velec);
 997
 998             fscal            = felec;
 999
1000             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1001
1002             /* Calculate temporary vectorial force */
1003             tx               = _mm256_mul_pd(fscal,dx11);
1004             ty               = _mm256_mul_pd(fscal,dy11);
1005             tz               = _mm256_mul_pd(fscal,dz11);
1006
1007             /* Update vectorial force */
1008             fix1             = _mm256_add_pd(fix1,tx);
1009             fiy1             = _mm256_add_pd(fiy1,ty);
1010             fiz1             = _mm256_add_pd(fiz1,tz);
1011
1012             fjx1             = _mm256_add_pd(fjx1,tx);
1013             fjy1             = _mm256_add_pd(fjy1,ty);
1014             fjz1             = _mm256_add_pd(fjz1,tz);
1015
1016             /**************************
1017              * CALCULATE INTERACTIONS *
1018              **************************/
1019
1020             r12              = _mm256_mul_pd(rsq12,rinv12);
1021             r12              = _mm256_andnot_pd(dummy_mask,r12);
1022
1023             /* EWALD ELECTROSTATICS */
1024
1025             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1026             ewrt             = _mm256_mul_pd(r12,ewtabscale);
1027             ewitab           = _mm256_cvttpd_epi32(ewrt);
1028             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1029             ewitab           = _mm_slli_epi32(ewitab,2);
1030             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1031             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1032             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1033             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1034             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1035             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1036             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1037             velec            = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
1038             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
1039
1040             /* Update potential sum for this i atom from the interaction with this j atom. */
1041             velec            = _mm256_andnot_pd(dummy_mask,velec);
1042             velecsum         = _mm256_add_pd(velecsum,velec);
1043
1044             fscal            = felec;
1045
1046             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1047
1048             /* Calculate temporary vectorial force */
1049             tx               = _mm256_mul_pd(fscal,dx12);
1050             ty               = _mm256_mul_pd(fscal,dy12);
1051             tz               = _mm256_mul_pd(fscal,dz12);
1052
1053             /* Update vectorial force */
1054             fix1             = _mm256_add_pd(fix1,tx);
1055             fiy1             = _mm256_add_pd(fiy1,ty);
1056             fiz1             = _mm256_add_pd(fiz1,tz);
1057
1058             fjx2             = _mm256_add_pd(fjx2,tx);
1059             fjy2             = _mm256_add_pd(fjy2,ty);
1060             fjz2             = _mm256_add_pd(fjz2,tz);
1061
1062             /**************************
1063              * CALCULATE INTERACTIONS *
1064              **************************/
1065
1066             r20              = _mm256_mul_pd(rsq20,rinv20);
1067             r20              = _mm256_andnot_pd(dummy_mask,r20);
1068
1069             /* EWALD ELECTROSTATICS */
1070
1071             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1072             ewrt             = _mm256_mul_pd(r20,ewtabscale);
1073             ewitab           = _mm256_cvttpd_epi32(ewrt);
1074             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1075             ewitab           = _mm_slli_epi32(ewitab,2);
1076             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1077             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1078             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1079             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1080             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1081             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1082             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1083             velec            = _mm256_mul_pd(qq20,_mm256_sub_pd(rinv20,velec));
1084             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
1085
1086             /* Update potential sum for this i atom from the interaction with this j atom. */
1087             velec            = _mm256_andnot_pd(dummy_mask,velec);
1088             velecsum         = _mm256_add_pd(velecsum,velec);
1089
1090             fscal            = felec;
1091
1092             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1093
1094             /* Calculate temporary vectorial force */
1095             tx               = _mm256_mul_pd(fscal,dx20);
1096             ty               = _mm256_mul_pd(fscal,dy20);
1097             tz               = _mm256_mul_pd(fscal,dz20);
1098
1099             /* Update vectorial force */
1100             fix2             = _mm256_add_pd(fix2,tx);
1101             fiy2             = _mm256_add_pd(fiy2,ty);
1102             fiz2             = _mm256_add_pd(fiz2,tz);
1103
1104             fjx0             = _mm256_add_pd(fjx0,tx);
1105             fjy0             = _mm256_add_pd(fjy0,ty);
1106             fjz0             = _mm256_add_pd(fjz0,tz);
1107
1108             /**************************
1109              * CALCULATE INTERACTIONS *
1110              **************************/
1111
1112             r21              = _mm256_mul_pd(rsq21,rinv21);
1113             r21              = _mm256_andnot_pd(dummy_mask,r21);
1114
1115             /* EWALD ELECTROSTATICS */
1116
1117             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1118             ewrt             = _mm256_mul_pd(r21,ewtabscale);
1119             ewitab           = _mm256_cvttpd_epi32(ewrt);
1120             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1121             ewitab           = _mm_slli_epi32(ewitab,2);
1122             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1123             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1124             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1125             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1126             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1127             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1128             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1129             velec            = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
1130             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1131
1132             /* Update potential sum for this i atom from the interaction with this j atom. */
1133             velec            = _mm256_andnot_pd(dummy_mask,velec);
1134             velecsum         = _mm256_add_pd(velecsum,velec);
1135
1136             fscal            = felec;
1137
1138             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1139
1140             /* Calculate temporary vectorial force */
1141             tx               = _mm256_mul_pd(fscal,dx21);
1142             ty               = _mm256_mul_pd(fscal,dy21);
1143             tz               = _mm256_mul_pd(fscal,dz21);
1144
1145             /* Update vectorial force */
1146             fix2             = _mm256_add_pd(fix2,tx);
1147             fiy2             = _mm256_add_pd(fiy2,ty);
1148             fiz2             = _mm256_add_pd(fiz2,tz);
1149
1150             fjx1             = _mm256_add_pd(fjx1,tx);
1151             fjy1             = _mm256_add_pd(fjy1,ty);
1152             fjz1             = _mm256_add_pd(fjz1,tz);
1153
1154             /**************************
1155              * CALCULATE INTERACTIONS *
1156              **************************/
1157
1158             r22              = _mm256_mul_pd(rsq22,rinv22);
1159             r22              = _mm256_andnot_pd(dummy_mask,r22);
1160
1161             /* EWALD ELECTROSTATICS */
1162
1163             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1164             ewrt             = _mm256_mul_pd(r22,ewtabscale);
1165             ewitab           = _mm256_cvttpd_epi32(ewrt);
1166             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1167             ewitab           = _mm_slli_epi32(ewitab,2);
1168             ewtabF           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1169             ewtabD           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1170             ewtabV           = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1171             ewtabFn          = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1172             GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1173             felec            = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1174             velec            = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1175             velec            = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
1176             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1177
1178             /* Update potential sum for this i atom from the interaction with this j atom. */
1179             velec            = _mm256_andnot_pd(dummy_mask,velec);
1180             velecsum         = _mm256_add_pd(velecsum,velec);
1181
1182             fscal            = felec;
1183
1184             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1185
1186             /* Calculate temporary vectorial force */
1187             tx               = _mm256_mul_pd(fscal,dx22);
1188             ty               = _mm256_mul_pd(fscal,dy22);
1189             tz               = _mm256_mul_pd(fscal,dz22);
1190
1191             /* Update vectorial force */
1192             fix2             = _mm256_add_pd(fix2,tx);
1193             fiy2             = _mm256_add_pd(fiy2,ty);
1194             fiz2             = _mm256_add_pd(fiz2,tz);
1195
1196             fjx2             = _mm256_add_pd(fjx2,tx);
1197             fjy2             = _mm256_add_pd(fjy2,ty);
1198             fjz2             = _mm256_add_pd(fjz2,tz);
1199
1200             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1201             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1202             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1203             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1204
1205             gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1206                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1207
1208             /* Inner loop uses 378 flops */
1209         }
1210
1211         /* End of innermost loop */
1212
1213         gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1214                                                  f+i_coord_offset,fshift+i_shift_offset);
1215
1216         ggid                        = gid[iidx];
1217         /* Update potential energies */
1218         gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1219
1220         /* Increment number of inner iterations */
1221         inneriter                  += j_index_end - j_index_start;
1222
1223         /* Outer loop uses 19 flops */
1224     }
1225
1226     /* Increment number of outer iterations */
1227     outeriter        += nri;
1228
1229     /* Update outer/inner flops */
1230
1231     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W3W3_VF,outeriter*19 + inneriter*378);
1232 }
1233 /*
1234  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwNone_GeomW3W3_F_avx_256_double
1235  * Electrostatics interaction: Ewald
1236  * VdW interaction:            None
1237  * Geometry:                   Water3-Water3
1238  * Calculate force/pot:        Force
1239  */
1240 void
1241 nb_kernel_ElecEw_VdwNone_GeomW3W3_F_avx_256_double
1242                     (t_nblist                    * gmx_restrict       nlist,
1243                      rvec                        * gmx_restrict          xx,
1244                      rvec                        * gmx_restrict          ff,
1245                      t_forcerec                  * gmx_restrict          fr,
1246                      t_mdatoms                   * gmx_restrict     mdatoms,
1247                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1248                      t_nrnb                      * gmx_restrict        nrnb)
1249 {
1250     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1251      * just 0 for non-waters.
1252      * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
1253      * jnr indices corresponding to data put in the four positions in the SIMD register.
1254      */
1255     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1256     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1257     int              jnrA,jnrB,jnrC,jnrD;
1258     int              jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1259     int              jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1260     int              j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1261     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1262     real             rcutoff_scalar;
1263     real             *shiftvec,*fshift,*x,*f;
1264     real             *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1265     real             scratch[4*DIM];
1266     __m256d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1267     real *           vdwioffsetptr0;
1268     __m256d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1269     real *           vdwioffsetptr1;
1270     __m256d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1271     real *           vdwioffsetptr2;
1272     __m256d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1273     int              vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1274     __m256d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1275     int              vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1276     __m256d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1277     int              vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1278     __m256d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1279     __m256d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1280     __m256d          dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
1281     __m256d          dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
1282     __m256d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
1283     __m256d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1284     __m256d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1285     __m256d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
1286     __m256d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1287     __m256d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1288     __m256d          velec,felec,velecsum,facel,crf,krf,krf2;
1289     real             *charge;
1290     __m128i          ewitab;
1291     __m256d          ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1292     __m256d          beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
1293     real             *ewtab;
1294     __m256d          dummy_mask,cutoff_mask;
1295     __m128           tmpmask0,tmpmask1;
1296     __m256d          signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
1297     __m256d          one     = _mm256_set1_pd(1.0);
1298     __m256d          two     = _mm256_set1_pd(2.0);
1299     x                = xx[0];
1300     f                = ff[0];
1301
1302     nri              = nlist->nri;
1303     iinr             = nlist->iinr;
1304     jindex           = nlist->jindex;
1305     jjnr             = nlist->jjnr;
1306     shiftidx         = nlist->shift;
1307     gid              = nlist->gid;
1308     shiftvec         = fr->shift_vec[0];
1309     fshift           = fr->fshift[0];
1310     facel            = _mm256_set1_pd(fr->epsfac);
1311     charge           = mdatoms->chargeA;
1312
1313     sh_ewald         = _mm256_set1_pd(fr->ic->sh_ewald);
1314     beta             = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
1315     beta2            = _mm256_mul_pd(beta,beta);
1316     beta3            = _mm256_mul_pd(beta,beta2);
1317
1318     ewtab            = fr->ic->tabq_coul_F;
1319     ewtabscale       = _mm256_set1_pd(fr->ic->tabq_scale);
1320     ewtabhalfspace   = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
1321
1322     /* Setup water-specific parameters */
1323     inr              = nlist->iinr[0];
1324     iq0              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
1325     iq1              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
1326     iq2              = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
1327
1328     jq0              = _mm256_set1_pd(charge[inr+0]);
1329     jq1              = _mm256_set1_pd(charge[inr+1]);
1330     jq2              = _mm256_set1_pd(charge[inr+2]);
1331     qq00             = _mm256_mul_pd(iq0,jq0);
1332     qq01             = _mm256_mul_pd(iq0,jq1);
1333     qq02             = _mm256_mul_pd(iq0,jq2);
1334     qq10             = _mm256_mul_pd(iq1,jq0);
1335     qq11             = _mm256_mul_pd(iq1,jq1);
1336     qq12             = _mm256_mul_pd(iq1,jq2);
1337     qq20             = _mm256_mul_pd(iq2,jq0);
1338     qq21             = _mm256_mul_pd(iq2,jq1);
1339     qq22             = _mm256_mul_pd(iq2,jq2);
1340
1341     /* Avoid stupid compiler warnings */
1342     jnrA = jnrB = jnrC = jnrD = 0;
1343     j_coord_offsetA = 0;
1344     j_coord_offsetB = 0;
1345     j_coord_offsetC = 0;
1346     j_coord_offsetD = 0;
1347
1348     outeriter        = 0;
1349     inneriter        = 0;
1350
1351     for(iidx=0;iidx<4*DIM;iidx++)
1352     {
1353         scratch[iidx] = 0.0;
1354     }
1355
1356     /* Start outer loop over neighborlists */
1357     for(iidx=0; iidx<nri; iidx++)
1358     {
1359         /* Load shift vector for this list */
1360         i_shift_offset   = DIM*shiftidx[iidx];
1361
1362         /* Load limits for loop over neighbors */
1363         j_index_start    = jindex[iidx];
1364         j_index_end      = jindex[iidx+1];
1365
1366         /* Get outer coordinate index */
1367         inr              = iinr[iidx];
1368         i_coord_offset   = DIM*inr;
1369
1370         /* Load i particle coords and add shift vector */
1371         gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1372                                                     &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
1373
1374         fix0             = _mm256_setzero_pd();
1375         fiy0             = _mm256_setzero_pd();
1376         fiz0             = _mm256_setzero_pd();
1377         fix1             = _mm256_setzero_pd();
1378         fiy1             = _mm256_setzero_pd();
1379         fiz1             = _mm256_setzero_pd();
1380         fix2             = _mm256_setzero_pd();
1381         fiy2             = _mm256_setzero_pd();
1382         fiz2             = _mm256_setzero_pd();
1383
1384         /* Start inner kernel loop */
1385         for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1386         {
1387
1388             /* Get j neighbor index, and coordinate index */
1389             jnrA             = jjnr[jidx];
1390             jnrB             = jjnr[jidx+1];
1391             jnrC             = jjnr[jidx+2];
1392             jnrD             = jjnr[jidx+3];
1393             j_coord_offsetA  = DIM*jnrA;
1394             j_coord_offsetB  = DIM*jnrB;
1395             j_coord_offsetC  = DIM*jnrC;
1396             j_coord_offsetD  = DIM*jnrD;
1397
1398             /* load j atom coordinates */
1399             gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1400                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1401                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1402
1403             /* Calculate displacement vector */
1404             dx00             = _mm256_sub_pd(ix0,jx0);
1405             dy00             = _mm256_sub_pd(iy0,jy0);
1406             dz00             = _mm256_sub_pd(iz0,jz0);
1407             dx01             = _mm256_sub_pd(ix0,jx1);
1408             dy01             = _mm256_sub_pd(iy0,jy1);
1409             dz01             = _mm256_sub_pd(iz0,jz1);
1410             dx02             = _mm256_sub_pd(ix0,jx2);
1411             dy02             = _mm256_sub_pd(iy0,jy2);
1412             dz02             = _mm256_sub_pd(iz0,jz2);
1413             dx10             = _mm256_sub_pd(ix1,jx0);
1414             dy10             = _mm256_sub_pd(iy1,jy0);
1415             dz10             = _mm256_sub_pd(iz1,jz0);
1416             dx11             = _mm256_sub_pd(ix1,jx1);
1417             dy11             = _mm256_sub_pd(iy1,jy1);
1418             dz11             = _mm256_sub_pd(iz1,jz1);
1419             dx12             = _mm256_sub_pd(ix1,jx2);
1420             dy12             = _mm256_sub_pd(iy1,jy2);
1421             dz12             = _mm256_sub_pd(iz1,jz2);
1422             dx20             = _mm256_sub_pd(ix2,jx0);
1423             dy20             = _mm256_sub_pd(iy2,jy0);
1424             dz20             = _mm256_sub_pd(iz2,jz0);
1425             dx21             = _mm256_sub_pd(ix2,jx1);
1426             dy21             = _mm256_sub_pd(iy2,jy1);
1427             dz21             = _mm256_sub_pd(iz2,jz1);
1428             dx22             = _mm256_sub_pd(ix2,jx2);
1429             dy22             = _mm256_sub_pd(iy2,jy2);
1430             dz22             = _mm256_sub_pd(iz2,jz2);
1431
1432             /* Calculate squared distance and things based on it */
1433             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1434             rsq01            = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
1435             rsq02            = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
1436             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1437             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1438             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1439             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1440             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1441             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1442
1443             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
1444             rinv01           = gmx_mm256_invsqrt_pd(rsq01);
1445             rinv02           = gmx_mm256_invsqrt_pd(rsq02);
1446             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
1447             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
1448             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
1449             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
1450             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
1451             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
1452
1453             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
1454             rinvsq01         = _mm256_mul_pd(rinv01,rinv01);
1455             rinvsq02         = _mm256_mul_pd(rinv02,rinv02);
1456             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
1457             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
1458             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
1459             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
1460             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
1461             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
1462
1463             fjx0             = _mm256_setzero_pd();
1464             fjy0             = _mm256_setzero_pd();
1465             fjz0             = _mm256_setzero_pd();
1466             fjx1             = _mm256_setzero_pd();
1467             fjy1             = _mm256_setzero_pd();
1468             fjz1             = _mm256_setzero_pd();
1469             fjx2             = _mm256_setzero_pd();
1470             fjy2             = _mm256_setzero_pd();
1471             fjz2             = _mm256_setzero_pd();
1472
1473             /**************************
1474              * CALCULATE INTERACTIONS *
1475              **************************/
1476
1477             r00              = _mm256_mul_pd(rsq00,rinv00);
1478
1479             /* EWALD ELECTROSTATICS */
1480
1481             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1482             ewrt             = _mm256_mul_pd(r00,ewtabscale);
1483             ewitab           = _mm256_cvttpd_epi32(ewrt);
1484             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1485             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1486                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1487                                             &ewtabF,&ewtabFn);
1488             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1489             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
1490
1491             fscal            = felec;
1492
1493             /* Calculate temporary vectorial force */
1494             tx               = _mm256_mul_pd(fscal,dx00);
1495             ty               = _mm256_mul_pd(fscal,dy00);
1496             tz               = _mm256_mul_pd(fscal,dz00);
1497
1498             /* Update vectorial force */
1499             fix0             = _mm256_add_pd(fix0,tx);
1500             fiy0             = _mm256_add_pd(fiy0,ty);
1501             fiz0             = _mm256_add_pd(fiz0,tz);
1502
1503             fjx0             = _mm256_add_pd(fjx0,tx);
1504             fjy0             = _mm256_add_pd(fjy0,ty);
1505             fjz0             = _mm256_add_pd(fjz0,tz);
1506
1507             /**************************
1508              * CALCULATE INTERACTIONS *
1509              **************************/
1510
1511             r01              = _mm256_mul_pd(rsq01,rinv01);
1512
1513             /* EWALD ELECTROSTATICS */
1514
1515             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1516             ewrt             = _mm256_mul_pd(r01,ewtabscale);
1517             ewitab           = _mm256_cvttpd_epi32(ewrt);
1518             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1519             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1520                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1521                                             &ewtabF,&ewtabFn);
1522             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1523             felec            = _mm256_mul_pd(_mm256_mul_pd(qq01,rinv01),_mm256_sub_pd(rinvsq01,felec));
1524
1525             fscal            = felec;
1526
1527             /* Calculate temporary vectorial force */
1528             tx               = _mm256_mul_pd(fscal,dx01);
1529             ty               = _mm256_mul_pd(fscal,dy01);
1530             tz               = _mm256_mul_pd(fscal,dz01);
1531
1532             /* Update vectorial force */
1533             fix0             = _mm256_add_pd(fix0,tx);
1534             fiy0             = _mm256_add_pd(fiy0,ty);
1535             fiz0             = _mm256_add_pd(fiz0,tz);
1536
1537             fjx1             = _mm256_add_pd(fjx1,tx);
1538             fjy1             = _mm256_add_pd(fjy1,ty);
1539             fjz1             = _mm256_add_pd(fjz1,tz);
1540
1541             /**************************
1542              * CALCULATE INTERACTIONS *
1543              **************************/
1544
1545             r02              = _mm256_mul_pd(rsq02,rinv02);
1546
1547             /* EWALD ELECTROSTATICS */
1548
1549             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1550             ewrt             = _mm256_mul_pd(r02,ewtabscale);
1551             ewitab           = _mm256_cvttpd_epi32(ewrt);
1552             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1553             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1554                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1555                                             &ewtabF,&ewtabFn);
1556             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1557             felec            = _mm256_mul_pd(_mm256_mul_pd(qq02,rinv02),_mm256_sub_pd(rinvsq02,felec));
1558
1559             fscal            = felec;
1560
1561             /* Calculate temporary vectorial force */
1562             tx               = _mm256_mul_pd(fscal,dx02);
1563             ty               = _mm256_mul_pd(fscal,dy02);
1564             tz               = _mm256_mul_pd(fscal,dz02);
1565
1566             /* Update vectorial force */
1567             fix0             = _mm256_add_pd(fix0,tx);
1568             fiy0             = _mm256_add_pd(fiy0,ty);
1569             fiz0             = _mm256_add_pd(fiz0,tz);
1570
1571             fjx2             = _mm256_add_pd(fjx2,tx);
1572             fjy2             = _mm256_add_pd(fjy2,ty);
1573             fjz2             = _mm256_add_pd(fjz2,tz);
1574
1575             /**************************
1576              * CALCULATE INTERACTIONS *
1577              **************************/
1578
1579             r10              = _mm256_mul_pd(rsq10,rinv10);
1580
1581             /* EWALD ELECTROSTATICS */
1582
1583             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1584             ewrt             = _mm256_mul_pd(r10,ewtabscale);
1585             ewitab           = _mm256_cvttpd_epi32(ewrt);
1586             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1587             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1588                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1589                                             &ewtabF,&ewtabFn);
1590             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1591             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
1592
1593             fscal            = felec;
1594
1595             /* Calculate temporary vectorial force */
1596             tx               = _mm256_mul_pd(fscal,dx10);
1597             ty               = _mm256_mul_pd(fscal,dy10);
1598             tz               = _mm256_mul_pd(fscal,dz10);
1599
1600             /* Update vectorial force */
1601             fix1             = _mm256_add_pd(fix1,tx);
1602             fiy1             = _mm256_add_pd(fiy1,ty);
1603             fiz1             = _mm256_add_pd(fiz1,tz);
1604
1605             fjx0             = _mm256_add_pd(fjx0,tx);
1606             fjy0             = _mm256_add_pd(fjy0,ty);
1607             fjz0             = _mm256_add_pd(fjz0,tz);
1608
1609             /**************************
1610              * CALCULATE INTERACTIONS *
1611              **************************/
1612
1613             r11              = _mm256_mul_pd(rsq11,rinv11);
1614
1615             /* EWALD ELECTROSTATICS */
1616
1617             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1618             ewrt             = _mm256_mul_pd(r11,ewtabscale);
1619             ewitab           = _mm256_cvttpd_epi32(ewrt);
1620             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1621             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1622                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1623                                             &ewtabF,&ewtabFn);
1624             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1625             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
1626
1627             fscal            = felec;
1628
1629             /* Calculate temporary vectorial force */
1630             tx               = _mm256_mul_pd(fscal,dx11);
1631             ty               = _mm256_mul_pd(fscal,dy11);
1632             tz               = _mm256_mul_pd(fscal,dz11);
1633
1634             /* Update vectorial force */
1635             fix1             = _mm256_add_pd(fix1,tx);
1636             fiy1             = _mm256_add_pd(fiy1,ty);
1637             fiz1             = _mm256_add_pd(fiz1,tz);
1638
1639             fjx1             = _mm256_add_pd(fjx1,tx);
1640             fjy1             = _mm256_add_pd(fjy1,ty);
1641             fjz1             = _mm256_add_pd(fjz1,tz);
1642
1643             /**************************
1644              * CALCULATE INTERACTIONS *
1645              **************************/
1646
1647             r12              = _mm256_mul_pd(rsq12,rinv12);
1648
1649             /* EWALD ELECTROSTATICS */
1650
1651             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1652             ewrt             = _mm256_mul_pd(r12,ewtabscale);
1653             ewitab           = _mm256_cvttpd_epi32(ewrt);
1654             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1655             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1656                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1657                                             &ewtabF,&ewtabFn);
1658             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1659             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
1660
1661             fscal            = felec;
1662
1663             /* Calculate temporary vectorial force */
1664             tx               = _mm256_mul_pd(fscal,dx12);
1665             ty               = _mm256_mul_pd(fscal,dy12);
1666             tz               = _mm256_mul_pd(fscal,dz12);
1667
1668             /* Update vectorial force */
1669             fix1             = _mm256_add_pd(fix1,tx);
1670             fiy1             = _mm256_add_pd(fiy1,ty);
1671             fiz1             = _mm256_add_pd(fiz1,tz);
1672
1673             fjx2             = _mm256_add_pd(fjx2,tx);
1674             fjy2             = _mm256_add_pd(fjy2,ty);
1675             fjz2             = _mm256_add_pd(fjz2,tz);
1676
1677             /**************************
1678              * CALCULATE INTERACTIONS *
1679              **************************/
1680
1681             r20              = _mm256_mul_pd(rsq20,rinv20);
1682
1683             /* EWALD ELECTROSTATICS */
1684
1685             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1686             ewrt             = _mm256_mul_pd(r20,ewtabscale);
1687             ewitab           = _mm256_cvttpd_epi32(ewrt);
1688             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1689             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1690                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1691                                             &ewtabF,&ewtabFn);
1692             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1693             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
1694
1695             fscal            = felec;
1696
1697             /* Calculate temporary vectorial force */
1698             tx               = _mm256_mul_pd(fscal,dx20);
1699             ty               = _mm256_mul_pd(fscal,dy20);
1700             tz               = _mm256_mul_pd(fscal,dz20);
1701
1702             /* Update vectorial force */
1703             fix2             = _mm256_add_pd(fix2,tx);
1704             fiy2             = _mm256_add_pd(fiy2,ty);
1705             fiz2             = _mm256_add_pd(fiz2,tz);
1706
1707             fjx0             = _mm256_add_pd(fjx0,tx);
1708             fjy0             = _mm256_add_pd(fjy0,ty);
1709             fjz0             = _mm256_add_pd(fjz0,tz);
1710
1711             /**************************
1712              * CALCULATE INTERACTIONS *
1713              **************************/
1714
1715             r21              = _mm256_mul_pd(rsq21,rinv21);
1716
1717             /* EWALD ELECTROSTATICS */
1718
1719             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1720             ewrt             = _mm256_mul_pd(r21,ewtabscale);
1721             ewitab           = _mm256_cvttpd_epi32(ewrt);
1722             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1723             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1724                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1725                                             &ewtabF,&ewtabFn);
1726             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1727             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1728
1729             fscal            = felec;
1730
1731             /* Calculate temporary vectorial force */
1732             tx               = _mm256_mul_pd(fscal,dx21);
1733             ty               = _mm256_mul_pd(fscal,dy21);
1734             tz               = _mm256_mul_pd(fscal,dz21);
1735
1736             /* Update vectorial force */
1737             fix2             = _mm256_add_pd(fix2,tx);
1738             fiy2             = _mm256_add_pd(fiy2,ty);
1739             fiz2             = _mm256_add_pd(fiz2,tz);
1740
1741             fjx1             = _mm256_add_pd(fjx1,tx);
1742             fjy1             = _mm256_add_pd(fjy1,ty);
1743             fjz1             = _mm256_add_pd(fjz1,tz);
1744
1745             /**************************
1746              * CALCULATE INTERACTIONS *
1747              **************************/
1748
1749             r22              = _mm256_mul_pd(rsq22,rinv22);
1750
1751             /* EWALD ELECTROSTATICS */
1752
1753             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1754             ewrt             = _mm256_mul_pd(r22,ewtabscale);
1755             ewitab           = _mm256_cvttpd_epi32(ewrt);
1756             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1757             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1758                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1759                                             &ewtabF,&ewtabFn);
1760             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1761             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1762
1763             fscal            = felec;
1764
1765             /* Calculate temporary vectorial force */
1766             tx               = _mm256_mul_pd(fscal,dx22);
1767             ty               = _mm256_mul_pd(fscal,dy22);
1768             tz               = _mm256_mul_pd(fscal,dz22);
1769
1770             /* Update vectorial force */
1771             fix2             = _mm256_add_pd(fix2,tx);
1772             fiy2             = _mm256_add_pd(fiy2,ty);
1773             fiz2             = _mm256_add_pd(fiz2,tz);
1774
1775             fjx2             = _mm256_add_pd(fjx2,tx);
1776             fjy2             = _mm256_add_pd(fjy2,ty);
1777             fjz2             = _mm256_add_pd(fjz2,tz);
1778
1779             fjptrA             = f+j_coord_offsetA;
1780             fjptrB             = f+j_coord_offsetB;
1781             fjptrC             = f+j_coord_offsetC;
1782             fjptrD             = f+j_coord_offsetD;
1783
1784             gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1785                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1786
1787             /* Inner loop uses 324 flops */
1788         }
1789
1790         if(jidx<j_index_end)
1791         {
1792
1793             /* Get j neighbor index, and coordinate index */
1794             jnrlistA         = jjnr[jidx];
1795             jnrlistB         = jjnr[jidx+1];
1796             jnrlistC         = jjnr[jidx+2];
1797             jnrlistD         = jjnr[jidx+3];
1798             /* Sign of each element will be negative for non-real atoms.
1799              * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1800              * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1801              */
1802             tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1803
1804             tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1805             tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1806             dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1807
1808             jnrA       = (jnrlistA>=0) ? jnrlistA : 0;
1809             jnrB       = (jnrlistB>=0) ? jnrlistB : 0;
1810             jnrC       = (jnrlistC>=0) ? jnrlistC : 0;
1811             jnrD       = (jnrlistD>=0) ? jnrlistD : 0;
1812             j_coord_offsetA  = DIM*jnrA;
1813             j_coord_offsetB  = DIM*jnrB;
1814             j_coord_offsetC  = DIM*jnrC;
1815             j_coord_offsetD  = DIM*jnrD;
1816
1817             /* load j atom coordinates */
1818             gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1819                                                  x+j_coord_offsetC,x+j_coord_offsetD,
1820                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1821
1822             /* Calculate displacement vector */
1823             dx00             = _mm256_sub_pd(ix0,jx0);
1824             dy00             = _mm256_sub_pd(iy0,jy0);
1825             dz00             = _mm256_sub_pd(iz0,jz0);
1826             dx01             = _mm256_sub_pd(ix0,jx1);
1827             dy01             = _mm256_sub_pd(iy0,jy1);
1828             dz01             = _mm256_sub_pd(iz0,jz1);
1829             dx02             = _mm256_sub_pd(ix0,jx2);
1830             dy02             = _mm256_sub_pd(iy0,jy2);
1831             dz02             = _mm256_sub_pd(iz0,jz2);
1832             dx10             = _mm256_sub_pd(ix1,jx0);
1833             dy10             = _mm256_sub_pd(iy1,jy0);
1834             dz10             = _mm256_sub_pd(iz1,jz0);
1835             dx11             = _mm256_sub_pd(ix1,jx1);
1836             dy11             = _mm256_sub_pd(iy1,jy1);
1837             dz11             = _mm256_sub_pd(iz1,jz1);
1838             dx12             = _mm256_sub_pd(ix1,jx2);
1839             dy12             = _mm256_sub_pd(iy1,jy2);
1840             dz12             = _mm256_sub_pd(iz1,jz2);
1841             dx20             = _mm256_sub_pd(ix2,jx0);
1842             dy20             = _mm256_sub_pd(iy2,jy0);
1843             dz20             = _mm256_sub_pd(iz2,jz0);
1844             dx21             = _mm256_sub_pd(ix2,jx1);
1845             dy21             = _mm256_sub_pd(iy2,jy1);
1846             dz21             = _mm256_sub_pd(iz2,jz1);
1847             dx22             = _mm256_sub_pd(ix2,jx2);
1848             dy22             = _mm256_sub_pd(iy2,jy2);
1849             dz22             = _mm256_sub_pd(iz2,jz2);
1850
1851             /* Calculate squared distance and things based on it */
1852             rsq00            = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1853             rsq01            = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
1854             rsq02            = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
1855             rsq10            = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1856             rsq11            = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1857             rsq12            = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1858             rsq20            = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1859             rsq21            = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1860             rsq22            = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1861
1862             rinv00           = gmx_mm256_invsqrt_pd(rsq00);
1863             rinv01           = gmx_mm256_invsqrt_pd(rsq01);
1864             rinv02           = gmx_mm256_invsqrt_pd(rsq02);
1865             rinv10           = gmx_mm256_invsqrt_pd(rsq10);
1866             rinv11           = gmx_mm256_invsqrt_pd(rsq11);
1867             rinv12           = gmx_mm256_invsqrt_pd(rsq12);
1868             rinv20           = gmx_mm256_invsqrt_pd(rsq20);
1869             rinv21           = gmx_mm256_invsqrt_pd(rsq21);
1870             rinv22           = gmx_mm256_invsqrt_pd(rsq22);
1871
1872             rinvsq00         = _mm256_mul_pd(rinv00,rinv00);
1873             rinvsq01         = _mm256_mul_pd(rinv01,rinv01);
1874             rinvsq02         = _mm256_mul_pd(rinv02,rinv02);
1875             rinvsq10         = _mm256_mul_pd(rinv10,rinv10);
1876             rinvsq11         = _mm256_mul_pd(rinv11,rinv11);
1877             rinvsq12         = _mm256_mul_pd(rinv12,rinv12);
1878             rinvsq20         = _mm256_mul_pd(rinv20,rinv20);
1879             rinvsq21         = _mm256_mul_pd(rinv21,rinv21);
1880             rinvsq22         = _mm256_mul_pd(rinv22,rinv22);
1881
1882             fjx0             = _mm256_setzero_pd();
1883             fjy0             = _mm256_setzero_pd();
1884             fjz0             = _mm256_setzero_pd();
1885             fjx1             = _mm256_setzero_pd();
1886             fjy1             = _mm256_setzero_pd();
1887             fjz1             = _mm256_setzero_pd();
1888             fjx2             = _mm256_setzero_pd();
1889             fjy2             = _mm256_setzero_pd();
1890             fjz2             = _mm256_setzero_pd();
1891
1892             /**************************
1893              * CALCULATE INTERACTIONS *
1894              **************************/
1895
1896             r00              = _mm256_mul_pd(rsq00,rinv00);
1897             r00              = _mm256_andnot_pd(dummy_mask,r00);
1898
1899             /* EWALD ELECTROSTATICS */
1900
1901             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1902             ewrt             = _mm256_mul_pd(r00,ewtabscale);
1903             ewitab           = _mm256_cvttpd_epi32(ewrt);
1904             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1905             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1906                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1907                                             &ewtabF,&ewtabFn);
1908             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1909             felec            = _mm256_mul_pd(_mm256_mul_pd(qq00,rinv00),_mm256_sub_pd(rinvsq00,felec));
1910
1911             fscal            = felec;
1912
1913             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1914
1915             /* Calculate temporary vectorial force */
1916             tx               = _mm256_mul_pd(fscal,dx00);
1917             ty               = _mm256_mul_pd(fscal,dy00);
1918             tz               = _mm256_mul_pd(fscal,dz00);
1919
1920             /* Update vectorial force */
1921             fix0             = _mm256_add_pd(fix0,tx);
1922             fiy0             = _mm256_add_pd(fiy0,ty);
1923             fiz0             = _mm256_add_pd(fiz0,tz);
1924
1925             fjx0             = _mm256_add_pd(fjx0,tx);
1926             fjy0             = _mm256_add_pd(fjy0,ty);
1927             fjz0             = _mm256_add_pd(fjz0,tz);
1928
1929             /**************************
1930              * CALCULATE INTERACTIONS *
1931              **************************/
1932
1933             r01              = _mm256_mul_pd(rsq01,rinv01);
1934             r01              = _mm256_andnot_pd(dummy_mask,r01);
1935
1936             /* EWALD ELECTROSTATICS */
1937
1938             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1939             ewrt             = _mm256_mul_pd(r01,ewtabscale);
1940             ewitab           = _mm256_cvttpd_epi32(ewrt);
1941             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1942             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1943                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1944                                             &ewtabF,&ewtabFn);
1945             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1946             felec            = _mm256_mul_pd(_mm256_mul_pd(qq01,rinv01),_mm256_sub_pd(rinvsq01,felec));
1947
1948             fscal            = felec;
1949
1950             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1951
1952             /* Calculate temporary vectorial force */
1953             tx               = _mm256_mul_pd(fscal,dx01);
1954             ty               = _mm256_mul_pd(fscal,dy01);
1955             tz               = _mm256_mul_pd(fscal,dz01);
1956
1957             /* Update vectorial force */
1958             fix0             = _mm256_add_pd(fix0,tx);
1959             fiy0             = _mm256_add_pd(fiy0,ty);
1960             fiz0             = _mm256_add_pd(fiz0,tz);
1961
1962             fjx1             = _mm256_add_pd(fjx1,tx);
1963             fjy1             = _mm256_add_pd(fjy1,ty);
1964             fjz1             = _mm256_add_pd(fjz1,tz);
1965
1966             /**************************
1967              * CALCULATE INTERACTIONS *
1968              **************************/
1969
1970             r02              = _mm256_mul_pd(rsq02,rinv02);
1971             r02              = _mm256_andnot_pd(dummy_mask,r02);
1972
1973             /* EWALD ELECTROSTATICS */
1974
1975             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1976             ewrt             = _mm256_mul_pd(r02,ewtabscale);
1977             ewitab           = _mm256_cvttpd_epi32(ewrt);
1978             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1979             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1980                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1981                                             &ewtabF,&ewtabFn);
1982             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1983             felec            = _mm256_mul_pd(_mm256_mul_pd(qq02,rinv02),_mm256_sub_pd(rinvsq02,felec));
1984
1985             fscal            = felec;
1986
1987             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
1988
1989             /* Calculate temporary vectorial force */
1990             tx               = _mm256_mul_pd(fscal,dx02);
1991             ty               = _mm256_mul_pd(fscal,dy02);
1992             tz               = _mm256_mul_pd(fscal,dz02);
1993
1994             /* Update vectorial force */
1995             fix0             = _mm256_add_pd(fix0,tx);
1996             fiy0             = _mm256_add_pd(fiy0,ty);
1997             fiz0             = _mm256_add_pd(fiz0,tz);
1998
1999             fjx2             = _mm256_add_pd(fjx2,tx);
2000             fjy2             = _mm256_add_pd(fjy2,ty);
2001             fjz2             = _mm256_add_pd(fjz2,tz);
2002
2003             /**************************
2004              * CALCULATE INTERACTIONS *
2005              **************************/
2006
2007             r10              = _mm256_mul_pd(rsq10,rinv10);
2008             r10              = _mm256_andnot_pd(dummy_mask,r10);
2009
2010             /* EWALD ELECTROSTATICS */
2011
2012             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2013             ewrt             = _mm256_mul_pd(r10,ewtabscale);
2014             ewitab           = _mm256_cvttpd_epi32(ewrt);
2015             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2016             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2017                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2018                                             &ewtabF,&ewtabFn);
2019             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2020             felec            = _mm256_mul_pd(_mm256_mul_pd(qq10,rinv10),_mm256_sub_pd(rinvsq10,felec));
2021
2022             fscal            = felec;
2023
2024             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2025
2026             /* Calculate temporary vectorial force */
2027             tx               = _mm256_mul_pd(fscal,dx10);
2028             ty               = _mm256_mul_pd(fscal,dy10);
2029             tz               = _mm256_mul_pd(fscal,dz10);
2030
2031             /* Update vectorial force */
2032             fix1             = _mm256_add_pd(fix1,tx);
2033             fiy1             = _mm256_add_pd(fiy1,ty);
2034             fiz1             = _mm256_add_pd(fiz1,tz);
2035
2036             fjx0             = _mm256_add_pd(fjx0,tx);
2037             fjy0             = _mm256_add_pd(fjy0,ty);
2038             fjz0             = _mm256_add_pd(fjz0,tz);
2039
2040             /**************************
2041              * CALCULATE INTERACTIONS *
2042              **************************/
2043
2044             r11              = _mm256_mul_pd(rsq11,rinv11);
2045             r11              = _mm256_andnot_pd(dummy_mask,r11);
2046
2047             /* EWALD ELECTROSTATICS */
2048
2049             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2050             ewrt             = _mm256_mul_pd(r11,ewtabscale);
2051             ewitab           = _mm256_cvttpd_epi32(ewrt);
2052             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2053             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2054                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2055                                             &ewtabF,&ewtabFn);
2056             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2057             felec            = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
2058
2059             fscal            = felec;
2060
2061             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2062
2063             /* Calculate temporary vectorial force */
2064             tx               = _mm256_mul_pd(fscal,dx11);
2065             ty               = _mm256_mul_pd(fscal,dy11);
2066             tz               = _mm256_mul_pd(fscal,dz11);
2067
2068             /* Update vectorial force */
2069             fix1             = _mm256_add_pd(fix1,tx);
2070             fiy1             = _mm256_add_pd(fiy1,ty);
2071             fiz1             = _mm256_add_pd(fiz1,tz);
2072
2073             fjx1             = _mm256_add_pd(fjx1,tx);
2074             fjy1             = _mm256_add_pd(fjy1,ty);
2075             fjz1             = _mm256_add_pd(fjz1,tz);
2076
2077             /**************************
2078              * CALCULATE INTERACTIONS *
2079              **************************/
2080
2081             r12              = _mm256_mul_pd(rsq12,rinv12);
2082             r12              = _mm256_andnot_pd(dummy_mask,r12);
2083
2084             /* EWALD ELECTROSTATICS */
2085
2086             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2087             ewrt             = _mm256_mul_pd(r12,ewtabscale);
2088             ewitab           = _mm256_cvttpd_epi32(ewrt);
2089             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2090             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2091                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2092                                             &ewtabF,&ewtabFn);
2093             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2094             felec            = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
2095
2096             fscal            = felec;
2097
2098             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2099
2100             /* Calculate temporary vectorial force */
2101             tx               = _mm256_mul_pd(fscal,dx12);
2102             ty               = _mm256_mul_pd(fscal,dy12);
2103             tz               = _mm256_mul_pd(fscal,dz12);
2104
2105             /* Update vectorial force */
2106             fix1             = _mm256_add_pd(fix1,tx);
2107             fiy1             = _mm256_add_pd(fiy1,ty);
2108             fiz1             = _mm256_add_pd(fiz1,tz);
2109
2110             fjx2             = _mm256_add_pd(fjx2,tx);
2111             fjy2             = _mm256_add_pd(fjy2,ty);
2112             fjz2             = _mm256_add_pd(fjz2,tz);
2113
2114             /**************************
2115              * CALCULATE INTERACTIONS *
2116              **************************/
2117
2118             r20              = _mm256_mul_pd(rsq20,rinv20);
2119             r20              = _mm256_andnot_pd(dummy_mask,r20);
2120
2121             /* EWALD ELECTROSTATICS */
2122
2123             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2124             ewrt             = _mm256_mul_pd(r20,ewtabscale);
2125             ewitab           = _mm256_cvttpd_epi32(ewrt);
2126             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2127             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2128                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2129                                             &ewtabF,&ewtabFn);
2130             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2131             felec            = _mm256_mul_pd(_mm256_mul_pd(qq20,rinv20),_mm256_sub_pd(rinvsq20,felec));
2132
2133             fscal            = felec;
2134
2135             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2136
2137             /* Calculate temporary vectorial force */
2138             tx               = _mm256_mul_pd(fscal,dx20);
2139             ty               = _mm256_mul_pd(fscal,dy20);
2140             tz               = _mm256_mul_pd(fscal,dz20);
2141
2142             /* Update vectorial force */
2143             fix2             = _mm256_add_pd(fix2,tx);
2144             fiy2             = _mm256_add_pd(fiy2,ty);
2145             fiz2             = _mm256_add_pd(fiz2,tz);
2146
2147             fjx0             = _mm256_add_pd(fjx0,tx);
2148             fjy0             = _mm256_add_pd(fjy0,ty);
2149             fjz0             = _mm256_add_pd(fjz0,tz);
2150
2151             /**************************
2152              * CALCULATE INTERACTIONS *
2153              **************************/
2154
2155             r21              = _mm256_mul_pd(rsq21,rinv21);
2156             r21              = _mm256_andnot_pd(dummy_mask,r21);
2157
2158             /* EWALD ELECTROSTATICS */
2159
2160             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2161             ewrt             = _mm256_mul_pd(r21,ewtabscale);
2162             ewitab           = _mm256_cvttpd_epi32(ewrt);
2163             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2164             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2165                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2166                                             &ewtabF,&ewtabFn);
2167             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2168             felec            = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
2169
2170             fscal            = felec;
2171
2172             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2173
2174             /* Calculate temporary vectorial force */
2175             tx               = _mm256_mul_pd(fscal,dx21);
2176             ty               = _mm256_mul_pd(fscal,dy21);
2177             tz               = _mm256_mul_pd(fscal,dz21);
2178
2179             /* Update vectorial force */
2180             fix2             = _mm256_add_pd(fix2,tx);
2181             fiy2             = _mm256_add_pd(fiy2,ty);
2182             fiz2             = _mm256_add_pd(fiz2,tz);
2183
2184             fjx1             = _mm256_add_pd(fjx1,tx);
2185             fjy1             = _mm256_add_pd(fjy1,ty);
2186             fjz1             = _mm256_add_pd(fjz1,tz);
2187
2188             /**************************
2189              * CALCULATE INTERACTIONS *
2190              **************************/
2191
2192             r22              = _mm256_mul_pd(rsq22,rinv22);
2193             r22              = _mm256_andnot_pd(dummy_mask,r22);
2194
2195             /* EWALD ELECTROSTATICS */
2196
2197             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2198             ewrt             = _mm256_mul_pd(r22,ewtabscale);
2199             ewitab           = _mm256_cvttpd_epi32(ewrt);
2200             eweps            = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2201             gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2202                                             ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2203                                             &ewtabF,&ewtabFn);
2204             felec            = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2205             felec            = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
2206
2207             fscal            = felec;
2208
2209             fscal            = _mm256_andnot_pd(dummy_mask,fscal);
2210
2211             /* Calculate temporary vectorial force */
2212             tx               = _mm256_mul_pd(fscal,dx22);
2213             ty               = _mm256_mul_pd(fscal,dy22);
2214             tz               = _mm256_mul_pd(fscal,dz22);
2215
2216             /* Update vectorial force */
2217             fix2             = _mm256_add_pd(fix2,tx);
2218             fiy2             = _mm256_add_pd(fiy2,ty);
2219             fiz2             = _mm256_add_pd(fiz2,tz);
2220
2221             fjx2             = _mm256_add_pd(fjx2,tx);
2222             fjy2             = _mm256_add_pd(fjy2,ty);
2223             fjz2             = _mm256_add_pd(fjz2,tz);
2224
2225             fjptrA             = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
2226             fjptrB             = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
2227             fjptrC             = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
2228             fjptrD             = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
2229
2230             gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
2231                                                       fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
2232
2233             /* Inner loop uses 333 flops */
2234         }
2235
2236         /* End of innermost loop */
2237
2238         gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
2239                                                  f+i_coord_offset,fshift+i_shift_offset);
2240
2241         /* Increment number of inner iterations */
2242         inneriter                  += j_index_end - j_index_start;
2243
2244         /* Outer loop uses 18 flops */
2245     }
2246
2247     /* Increment number of outer iterations */
2248     outeriter        += nri;
2249
2250     /* Update outer/inner flops */
2251
2252     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W3W3_F,outeriter*18 + inneriter*333);
2253 }