src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_ElecEw_VdwLJEw_GeomW4W4_avx_128_fma_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_128_fma_double kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  48 #include "kernelutil_x86_avx_128_fma_double.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwLJEw_GeomW4W4_VF_avx_128_fma_double
  52  * Electrostatics interaction: Ewald
  53  * VdW interaction:            LJEwald
  54  * Geometry:                   Water4-Water4
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecEw_VdwLJEw_GeomW4W4_VF_avx_128_fma_double
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB;
  75     int              j_coord_offsetA,j_coord_offsetB;
  76     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  77     real             rcutoff_scalar;
  78     real             *shiftvec,*fshift,*x,*f;
  79     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  80     int              vdwioffset0;
  81     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  82     int              vdwioffset1;
  83     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  84     int              vdwioffset2;
  85     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  86     int              vdwioffset3;
  87     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  88     int              vdwjidx0A,vdwjidx0B;
  89     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  90     int              vdwjidx1A,vdwjidx1B;
  91     __m128d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
  92     int              vdwjidx2A,vdwjidx2B;
  93     __m128d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
  94     int              vdwjidx3A,vdwjidx3B;
  95     __m128d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
  96     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  97     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
  98     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
  99     __m128d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
 100     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
 101     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
 102     __m128d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
 103     __m128d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
 104     __m128d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
 105     __m128d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
 106     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 107     real             *charge;
 108     int              nvdwtype;
 109     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 110     int              *vdwtype;
 111     real             *vdwparam;
 112     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 113     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 114     __m128d           c6grid_00;
 115     __m128d           c6grid_11;
 116     __m128d           c6grid_12;
 117     __m128d           c6grid_13;
 118     __m128d           c6grid_21;
 119     __m128d           c6grid_22;
 120     __m128d           c6grid_23;
 121     __m128d           c6grid_31;
 122     __m128d           c6grid_32;
 123     __m128d           c6grid_33;
 124     real             *vdwgridparam;
 125     __m128d           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 126     __m128d           one_half  = _mm_set1_pd(0.5);
 127     __m128d           minus_one = _mm_set1_pd(-1.0);
 128     __m128i          ewitab;
 129     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 130     real             *ewtab;
 131     __m128d          dummy_mask,cutoff_mask;
 132     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 133     __m128d          one     = _mm_set1_pd(1.0);
 134     __m128d          two     = _mm_set1_pd(2.0);
 135     x                = xx[0];
 136     f                = ff[0];
 137
 138     nri              = nlist->nri;
 139     iinr             = nlist->iinr;
 140     jindex           = nlist->jindex;
 141     jjnr             = nlist->jjnr;
 142     shiftidx         = nlist->shift;
 143     gid              = nlist->gid;
 144     shiftvec         = fr->shift_vec[0];
 145     fshift           = fr->fshift[0];
 146     facel            = _mm_set1_pd(fr->epsfac);
 147     charge           = mdatoms->chargeA;
 148     nvdwtype         = fr->ntype;
 149     vdwparam         = fr->nbfp;
 150     vdwtype          = mdatoms->typeA;
 151     vdwgridparam     = fr->ljpme_c6grid;
 152     sh_lj_ewald      = _mm_set1_pd(fr->ic->sh_lj_ewald);
 153     ewclj            = _mm_set1_pd(fr->ewaldcoeff_lj);
 154     ewclj2           = _mm_mul_pd(minus_one,_mm_mul_pd(ewclj,ewclj));
 155
 156     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
 157     ewtab            = fr->ic->tabq_coul_FDV0;
 158     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 159     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 160
 161     /* Setup water-specific parameters */
 162     inr              = nlist->iinr[0];
 163     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
 164     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
 165     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
 166     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 167
 168     jq1              = _mm_set1_pd(charge[inr+1]);
 169     jq2              = _mm_set1_pd(charge[inr+2]);
 170     jq3              = _mm_set1_pd(charge[inr+3]);
 171     vdwjidx0A        = 2*vdwtype[inr+0];
 172     c6_00            = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
 173     c12_00           = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
 174     c6grid_00        = _mm_set1_pd(vdwgridparam[vdwioffset0+vdwjidx0A]);
 175     qq11             = _mm_mul_pd(iq1,jq1);
 176     qq12             = _mm_mul_pd(iq1,jq2);
 177     qq13             = _mm_mul_pd(iq1,jq3);
 178     qq21             = _mm_mul_pd(iq2,jq1);
 179     qq22             = _mm_mul_pd(iq2,jq2);
 180     qq23             = _mm_mul_pd(iq2,jq3);
 181     qq31             = _mm_mul_pd(iq3,jq1);
 182     qq32             = _mm_mul_pd(iq3,jq2);
 183     qq33             = _mm_mul_pd(iq3,jq3);
 184
 185     /* Avoid stupid compiler warnings */
 186     jnrA = jnrB = 0;
 187     j_coord_offsetA = 0;
 188     j_coord_offsetB = 0;
 189
 190     outeriter        = 0;
 191     inneriter        = 0;
 192
 193     /* Start outer loop over neighborlists */
 194     for(iidx=0; iidx<nri; iidx++)
 195     {
 196         /* Load shift vector for this list */
 197         i_shift_offset   = DIM*shiftidx[iidx];
 198
 199         /* Load limits for loop over neighbors */
 200         j_index_start    = jindex[iidx];
 201         j_index_end      = jindex[iidx+1];
 202
 203         /* Get outer coordinate index */
 204         inr              = iinr[iidx];
 205         i_coord_offset   = DIM*inr;
 206
 207         /* Load i particle coords and add shift vector */
 208         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 209                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 210
 211         fix0             = _mm_setzero_pd();
 212         fiy0             = _mm_setzero_pd();
 213         fiz0             = _mm_setzero_pd();
 214         fix1             = _mm_setzero_pd();
 215         fiy1             = _mm_setzero_pd();
 216         fiz1             = _mm_setzero_pd();
 217         fix2             = _mm_setzero_pd();
 218         fiy2             = _mm_setzero_pd();
 219         fiz2             = _mm_setzero_pd();
 220         fix3             = _mm_setzero_pd();
 221         fiy3             = _mm_setzero_pd();
 222         fiz3             = _mm_setzero_pd();
 223
 224         /* Reset potential sums */
 225         velecsum         = _mm_setzero_pd();
 226         vvdwsum          = _mm_setzero_pd();
 227
 228         /* Start inner kernel loop */
 229         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 230         {
 231
 232             /* Get j neighbor index, and coordinate index */
 233             jnrA             = jjnr[jidx];
 234             jnrB             = jjnr[jidx+1];
 235             j_coord_offsetA  = DIM*jnrA;
 236             j_coord_offsetB  = DIM*jnrB;
 237
 238             /* load j atom coordinates */
 239             gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 240                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 241                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 242
 243             /* Calculate displacement vector */
 244             dx00             = _mm_sub_pd(ix0,jx0);
 245             dy00             = _mm_sub_pd(iy0,jy0);
 246             dz00             = _mm_sub_pd(iz0,jz0);
 247             dx11             = _mm_sub_pd(ix1,jx1);
 248             dy11             = _mm_sub_pd(iy1,jy1);
 249             dz11             = _mm_sub_pd(iz1,jz1);
 250             dx12             = _mm_sub_pd(ix1,jx2);
 251             dy12             = _mm_sub_pd(iy1,jy2);
 252             dz12             = _mm_sub_pd(iz1,jz2);
 253             dx13             = _mm_sub_pd(ix1,jx3);
 254             dy13             = _mm_sub_pd(iy1,jy3);
 255             dz13             = _mm_sub_pd(iz1,jz3);
 256             dx21             = _mm_sub_pd(ix2,jx1);
 257             dy21             = _mm_sub_pd(iy2,jy1);
 258             dz21             = _mm_sub_pd(iz2,jz1);
 259             dx22             = _mm_sub_pd(ix2,jx2);
 260             dy22             = _mm_sub_pd(iy2,jy2);
 261             dz22             = _mm_sub_pd(iz2,jz2);
 262             dx23             = _mm_sub_pd(ix2,jx3);
 263             dy23             = _mm_sub_pd(iy2,jy3);
 264             dz23             = _mm_sub_pd(iz2,jz3);
 265             dx31             = _mm_sub_pd(ix3,jx1);
 266             dy31             = _mm_sub_pd(iy3,jy1);
 267             dz31             = _mm_sub_pd(iz3,jz1);
 268             dx32             = _mm_sub_pd(ix3,jx2);
 269             dy32             = _mm_sub_pd(iy3,jy2);
 270             dz32             = _mm_sub_pd(iz3,jz2);
 271             dx33             = _mm_sub_pd(ix3,jx3);
 272             dy33             = _mm_sub_pd(iy3,jy3);
 273             dz33             = _mm_sub_pd(iz3,jz3);
 274
 275             /* Calculate squared distance and things based on it */
 276             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 277             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 278             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 279             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
 280             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 281             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 282             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
 283             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
 284             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
 285             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
 286
 287             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 288             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 289             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 290             rinv13           = gmx_mm_invsqrt_pd(rsq13);
 291             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 292             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 293             rinv23           = gmx_mm_invsqrt_pd(rsq23);
 294             rinv31           = gmx_mm_invsqrt_pd(rsq31);
 295             rinv32           = gmx_mm_invsqrt_pd(rsq32);
 296             rinv33           = gmx_mm_invsqrt_pd(rsq33);
 297
 298             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 299             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
 300             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
 301             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
 302             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
 303             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
 304             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
 305             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
 306             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
 307             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
 308
 309             fjx0             = _mm_setzero_pd();
 310             fjy0             = _mm_setzero_pd();
 311             fjz0             = _mm_setzero_pd();
 312             fjx1             = _mm_setzero_pd();
 313             fjy1             = _mm_setzero_pd();
 314             fjz1             = _mm_setzero_pd();
 315             fjx2             = _mm_setzero_pd();
 316             fjy2             = _mm_setzero_pd();
 317             fjz2             = _mm_setzero_pd();
 318             fjx3             = _mm_setzero_pd();
 319             fjy3             = _mm_setzero_pd();
 320             fjz3             = _mm_setzero_pd();
 321
 322             /**************************
 323              * CALCULATE INTERACTIONS *
 324              **************************/
 325
 326             r00              = _mm_mul_pd(rsq00,rinv00);
 327
 328             /* Analytical LJ-PME */
 329             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 330             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
 331             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
 332             exponent         = gmx_simd_exp_d(ewcljrsq);
 333             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 334             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
 335             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 336             vvdw6            = _mm_mul_pd(_mm_macc_pd(-c6grid_00,_mm_sub_pd(one,poly),c6_00),rinvsix);
 337             vvdw12           = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
 338             vvdw             = _mm_msub_pd(vvdw12,one_twelfth,_mm_mul_pd(vvdw6,one_sixth));
 339             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 340             fvdw             = _mm_mul_pd(_mm_add_pd(vvdw12,_mm_msub_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6),vvdw6)),rinvsq00);
 341
 342             /* Update potential sum for this i atom from the interaction with this j atom. */
 343             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 344
 345             fscal            = fvdw;
 346
 347             /* Update vectorial force */
 348             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 349             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 350             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 351
 352             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 353             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 354             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 355
 356             /**************************
 357              * CALCULATE INTERACTIONS *
 358              **************************/
 359
 360             r11              = _mm_mul_pd(rsq11,rinv11);
 361
 362             /* EWALD ELECTROSTATICS */
 363
 364             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 365             ewrt             = _mm_mul_pd(r11,ewtabscale);
 366             ewitab           = _mm_cvttpd_epi32(ewrt);
 367 #ifdef __XOP__
 368             eweps            = _mm_frcz_pd(ewrt);
 369 #else
 370             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 371 #endif
 372             twoeweps         = _mm_add_pd(eweps,eweps);
 373             ewitab           = _mm_slli_epi32(ewitab,2);
 374             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 375             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 376             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 377             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 378             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 379             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 380             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 381             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 382             velec            = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
 383             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
 384
 385             /* Update potential sum for this i atom from the interaction with this j atom. */
 386             velecsum         = _mm_add_pd(velecsum,velec);
 387
 388             fscal            = felec;
 389
 390             /* Update vectorial force */
 391             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 392             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 393             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 394
 395             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 396             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 397             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 398
 399             /**************************
 400              * CALCULATE INTERACTIONS *
 401              **************************/
 402
 403             r12              = _mm_mul_pd(rsq12,rinv12);
 404
 405             /* EWALD ELECTROSTATICS */
 406
 407             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 408             ewrt             = _mm_mul_pd(r12,ewtabscale);
 409             ewitab           = _mm_cvttpd_epi32(ewrt);
 410 #ifdef __XOP__
 411             eweps            = _mm_frcz_pd(ewrt);
 412 #else
 413             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 414 #endif
 415             twoeweps         = _mm_add_pd(eweps,eweps);
 416             ewitab           = _mm_slli_epi32(ewitab,2);
 417             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 418             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 419             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 420             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 421             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 422             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 423             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 424             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 425             velec            = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
 426             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
 427
 428             /* Update potential sum for this i atom from the interaction with this j atom. */
 429             velecsum         = _mm_add_pd(velecsum,velec);
 430
 431             fscal            = felec;
 432
 433             /* Update vectorial force */
 434             fix1             = _mm_macc_pd(dx12,fscal,fix1);
 435             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
 436             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
 437
 438             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
 439             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
 440             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
 441
 442             /**************************
 443              * CALCULATE INTERACTIONS *
 444              **************************/
 445
 446             r13              = _mm_mul_pd(rsq13,rinv13);
 447
 448             /* EWALD ELECTROSTATICS */
 449
 450             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 451             ewrt             = _mm_mul_pd(r13,ewtabscale);
 452             ewitab           = _mm_cvttpd_epi32(ewrt);
 453 #ifdef __XOP__
 454             eweps            = _mm_frcz_pd(ewrt);
 455 #else
 456             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 457 #endif
 458             twoeweps         = _mm_add_pd(eweps,eweps);
 459             ewitab           = _mm_slli_epi32(ewitab,2);
 460             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 461             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 462             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 463             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 464             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 465             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 466             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 467             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 468             velec            = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
 469             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
 470
 471             /* Update potential sum for this i atom from the interaction with this j atom. */
 472             velecsum         = _mm_add_pd(velecsum,velec);
 473
 474             fscal            = felec;
 475
 476             /* Update vectorial force */
 477             fix1             = _mm_macc_pd(dx13,fscal,fix1);
 478             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
 479             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
 480
 481             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
 482             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
 483             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
 484
 485             /**************************
 486              * CALCULATE INTERACTIONS *
 487              **************************/
 488
 489             r21              = _mm_mul_pd(rsq21,rinv21);
 490
 491             /* EWALD ELECTROSTATICS */
 492
 493             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 494             ewrt             = _mm_mul_pd(r21,ewtabscale);
 495             ewitab           = _mm_cvttpd_epi32(ewrt);
 496 #ifdef __XOP__
 497             eweps            = _mm_frcz_pd(ewrt);
 498 #else
 499             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 500 #endif
 501             twoeweps         = _mm_add_pd(eweps,eweps);
 502             ewitab           = _mm_slli_epi32(ewitab,2);
 503             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 504             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 505             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 506             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 507             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 508             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 509             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 510             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 511             velec            = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
 512             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
 513
 514             /* Update potential sum for this i atom from the interaction with this j atom. */
 515             velecsum         = _mm_add_pd(velecsum,velec);
 516
 517             fscal            = felec;
 518
 519             /* Update vectorial force */
 520             fix2             = _mm_macc_pd(dx21,fscal,fix2);
 521             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
 522             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
 523
 524             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
 525             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
 526             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
 527
 528             /**************************
 529              * CALCULATE INTERACTIONS *
 530              **************************/
 531
 532             r22              = _mm_mul_pd(rsq22,rinv22);
 533
 534             /* EWALD ELECTROSTATICS */
 535
 536             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 537             ewrt             = _mm_mul_pd(r22,ewtabscale);
 538             ewitab           = _mm_cvttpd_epi32(ewrt);
 539 #ifdef __XOP__
 540             eweps            = _mm_frcz_pd(ewrt);
 541 #else
 542             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 543 #endif
 544             twoeweps         = _mm_add_pd(eweps,eweps);
 545             ewitab           = _mm_slli_epi32(ewitab,2);
 546             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 547             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 548             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 549             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 550             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 551             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 552             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 553             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 554             velec            = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
 555             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
 556
 557             /* Update potential sum for this i atom from the interaction with this j atom. */
 558             velecsum         = _mm_add_pd(velecsum,velec);
 559
 560             fscal            = felec;
 561
 562             /* Update vectorial force */
 563             fix2             = _mm_macc_pd(dx22,fscal,fix2);
 564             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
 565             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
 566
 567             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
 568             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
 569             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
 570
 571             /**************************
 572              * CALCULATE INTERACTIONS *
 573              **************************/
 574
 575             r23              = _mm_mul_pd(rsq23,rinv23);
 576
 577             /* EWALD ELECTROSTATICS */
 578
 579             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 580             ewrt             = _mm_mul_pd(r23,ewtabscale);
 581             ewitab           = _mm_cvttpd_epi32(ewrt);
 582 #ifdef __XOP__
 583             eweps            = _mm_frcz_pd(ewrt);
 584 #else
 585             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 586 #endif
 587             twoeweps         = _mm_add_pd(eweps,eweps);
 588             ewitab           = _mm_slli_epi32(ewitab,2);
 589             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 590             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 591             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 592             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 593             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 594             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 595             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 596             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 597             velec            = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
 598             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
 599
 600             /* Update potential sum for this i atom from the interaction with this j atom. */
 601             velecsum         = _mm_add_pd(velecsum,velec);
 602
 603             fscal            = felec;
 604
 605             /* Update vectorial force */
 606             fix2             = _mm_macc_pd(dx23,fscal,fix2);
 607             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
 608             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
 609
 610             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
 611             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
 612             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
 613
 614             /**************************
 615              * CALCULATE INTERACTIONS *
 616              **************************/
 617
 618             r31              = _mm_mul_pd(rsq31,rinv31);
 619
 620             /* EWALD ELECTROSTATICS */
 621
 622             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 623             ewrt             = _mm_mul_pd(r31,ewtabscale);
 624             ewitab           = _mm_cvttpd_epi32(ewrt);
 625 #ifdef __XOP__
 626             eweps            = _mm_frcz_pd(ewrt);
 627 #else
 628             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 629 #endif
 630             twoeweps         = _mm_add_pd(eweps,eweps);
 631             ewitab           = _mm_slli_epi32(ewitab,2);
 632             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 633             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 634             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 635             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 636             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 637             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 638             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 639             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 640             velec            = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
 641             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
 642
 643             /* Update potential sum for this i atom from the interaction with this j atom. */
 644             velecsum         = _mm_add_pd(velecsum,velec);
 645
 646             fscal            = felec;
 647
 648             /* Update vectorial force */
 649             fix3             = _mm_macc_pd(dx31,fscal,fix3);
 650             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
 651             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
 652
 653             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
 654             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
 655             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
 656
 657             /**************************
 658              * CALCULATE INTERACTIONS *
 659              **************************/
 660
 661             r32              = _mm_mul_pd(rsq32,rinv32);
 662
 663             /* EWALD ELECTROSTATICS */
 664
 665             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 666             ewrt             = _mm_mul_pd(r32,ewtabscale);
 667             ewitab           = _mm_cvttpd_epi32(ewrt);
 668 #ifdef __XOP__
 669             eweps            = _mm_frcz_pd(ewrt);
 670 #else
 671             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 672 #endif
 673             twoeweps         = _mm_add_pd(eweps,eweps);
 674             ewitab           = _mm_slli_epi32(ewitab,2);
 675             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 676             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 677             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 678             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 679             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 680             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 681             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 682             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 683             velec            = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
 684             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
 685
 686             /* Update potential sum for this i atom from the interaction with this j atom. */
 687             velecsum         = _mm_add_pd(velecsum,velec);
 688
 689             fscal            = felec;
 690
 691             /* Update vectorial force */
 692             fix3             = _mm_macc_pd(dx32,fscal,fix3);
 693             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
 694             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
 695
 696             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
 697             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
 698             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
 699
 700             /**************************
 701              * CALCULATE INTERACTIONS *
 702              **************************/
 703
 704             r33              = _mm_mul_pd(rsq33,rinv33);
 705
 706             /* EWALD ELECTROSTATICS */
 707
 708             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 709             ewrt             = _mm_mul_pd(r33,ewtabscale);
 710             ewitab           = _mm_cvttpd_epi32(ewrt);
 711 #ifdef __XOP__
 712             eweps            = _mm_frcz_pd(ewrt);
 713 #else
 714             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 715 #endif
 716             twoeweps         = _mm_add_pd(eweps,eweps);
 717             ewitab           = _mm_slli_epi32(ewitab,2);
 718             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 719             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 720             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 721             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 722             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 723             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 724             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 725             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 726             velec            = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
 727             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
 728
 729             /* Update potential sum for this i atom from the interaction with this j atom. */
 730             velecsum         = _mm_add_pd(velecsum,velec);
 731
 732             fscal            = felec;
 733
 734             /* Update vectorial force */
 735             fix3             = _mm_macc_pd(dx33,fscal,fix3);
 736             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
 737             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
 738
 739             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
 740             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
 741             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
 742
 743             gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
 744
 745             /* Inner loop uses 449 flops */
 746         }
 747
 748         if(jidx<j_index_end)
 749         {
 750
 751             jnrA             = jjnr[jidx];
 752             j_coord_offsetA  = DIM*jnrA;
 753
 754             /* load j atom coordinates */
 755             gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 756                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
 757                                               &jy2,&jz2,&jx3,&jy3,&jz3);
 758
 759             /* Calculate displacement vector */
 760             dx00             = _mm_sub_pd(ix0,jx0);
 761             dy00             = _mm_sub_pd(iy0,jy0);
 762             dz00             = _mm_sub_pd(iz0,jz0);
 763             dx11             = _mm_sub_pd(ix1,jx1);
 764             dy11             = _mm_sub_pd(iy1,jy1);
 765             dz11             = _mm_sub_pd(iz1,jz1);
 766             dx12             = _mm_sub_pd(ix1,jx2);
 767             dy12             = _mm_sub_pd(iy1,jy2);
 768             dz12             = _mm_sub_pd(iz1,jz2);
 769             dx13             = _mm_sub_pd(ix1,jx3);
 770             dy13             = _mm_sub_pd(iy1,jy3);
 771             dz13             = _mm_sub_pd(iz1,jz3);
 772             dx21             = _mm_sub_pd(ix2,jx1);
 773             dy21             = _mm_sub_pd(iy2,jy1);
 774             dz21             = _mm_sub_pd(iz2,jz1);
 775             dx22             = _mm_sub_pd(ix2,jx2);
 776             dy22             = _mm_sub_pd(iy2,jy2);
 777             dz22             = _mm_sub_pd(iz2,jz2);
 778             dx23             = _mm_sub_pd(ix2,jx3);
 779             dy23             = _mm_sub_pd(iy2,jy3);
 780             dz23             = _mm_sub_pd(iz2,jz3);
 781             dx31             = _mm_sub_pd(ix3,jx1);
 782             dy31             = _mm_sub_pd(iy3,jy1);
 783             dz31             = _mm_sub_pd(iz3,jz1);
 784             dx32             = _mm_sub_pd(ix3,jx2);
 785             dy32             = _mm_sub_pd(iy3,jy2);
 786             dz32             = _mm_sub_pd(iz3,jz2);
 787             dx33             = _mm_sub_pd(ix3,jx3);
 788             dy33             = _mm_sub_pd(iy3,jy3);
 789             dz33             = _mm_sub_pd(iz3,jz3);
 790
 791             /* Calculate squared distance and things based on it */
 792             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 793             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
 794             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
 795             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
 796             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
 797             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
 798             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
 799             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
 800             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
 801             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
 802
 803             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 804             rinv11           = gmx_mm_invsqrt_pd(rsq11);
 805             rinv12           = gmx_mm_invsqrt_pd(rsq12);
 806             rinv13           = gmx_mm_invsqrt_pd(rsq13);
 807             rinv21           = gmx_mm_invsqrt_pd(rsq21);
 808             rinv22           = gmx_mm_invsqrt_pd(rsq22);
 809             rinv23           = gmx_mm_invsqrt_pd(rsq23);
 810             rinv31           = gmx_mm_invsqrt_pd(rsq31);
 811             rinv32           = gmx_mm_invsqrt_pd(rsq32);
 812             rinv33           = gmx_mm_invsqrt_pd(rsq33);
 813
 814             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 815             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
 816             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
 817             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
 818             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
 819             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
 820             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
 821             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
 822             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
 823             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
 824
 825             fjx0             = _mm_setzero_pd();
 826             fjy0             = _mm_setzero_pd();
 827             fjz0             = _mm_setzero_pd();
 828             fjx1             = _mm_setzero_pd();
 829             fjy1             = _mm_setzero_pd();
 830             fjz1             = _mm_setzero_pd();
 831             fjx2             = _mm_setzero_pd();
 832             fjy2             = _mm_setzero_pd();
 833             fjz2             = _mm_setzero_pd();
 834             fjx3             = _mm_setzero_pd();
 835             fjy3             = _mm_setzero_pd();
 836             fjz3             = _mm_setzero_pd();
 837
 838             /**************************
 839              * CALCULATE INTERACTIONS *
 840              **************************/
 841
 842             r00              = _mm_mul_pd(rsq00,rinv00);
 843
 844             /* Analytical LJ-PME */
 845             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 846             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
 847             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
 848             exponent         = gmx_simd_exp_d(ewcljrsq);
 849             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 850             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
 851             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 852             vvdw6            = _mm_mul_pd(_mm_macc_pd(-c6grid_00,_mm_sub_pd(one,poly),c6_00),rinvsix);
 853             vvdw12           = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
 854             vvdw             = _mm_msub_pd(vvdw12,one_twelfth,_mm_mul_pd(vvdw6,one_sixth));
 855             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 856             fvdw             = _mm_mul_pd(_mm_add_pd(vvdw12,_mm_msub_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6),vvdw6)),rinvsq00);
 857
 858             /* Update potential sum for this i atom from the interaction with this j atom. */
 859             vvdw             = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
 860             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 861
 862             fscal            = fvdw;
 863
 864             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 865
 866             /* Update vectorial force */
 867             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 868             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 869             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 870
 871             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 872             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 873             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 874
 875             /**************************
 876              * CALCULATE INTERACTIONS *
 877              **************************/
 878
 879             r11              = _mm_mul_pd(rsq11,rinv11);
 880
 881             /* EWALD ELECTROSTATICS */
 882
 883             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 884             ewrt             = _mm_mul_pd(r11,ewtabscale);
 885             ewitab           = _mm_cvttpd_epi32(ewrt);
 886 #ifdef __XOP__
 887             eweps            = _mm_frcz_pd(ewrt);
 888 #else
 889             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 890 #endif
 891             twoeweps         = _mm_add_pd(eweps,eweps);
 892             ewitab           = _mm_slli_epi32(ewitab,2);
 893             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 894             ewtabD           = _mm_setzero_pd();
 895             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 896             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 897             ewtabFn          = _mm_setzero_pd();
 898             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 899             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 900             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 901             velec            = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
 902             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
 903
 904             /* Update potential sum for this i atom from the interaction with this j atom. */
 905             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 906             velecsum         = _mm_add_pd(velecsum,velec);
 907
 908             fscal            = felec;
 909
 910             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 911
 912             /* Update vectorial force */
 913             fix1             = _mm_macc_pd(dx11,fscal,fix1);
 914             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
 915             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
 916
 917             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
 918             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
 919             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
 920
 921             /**************************
 922              * CALCULATE INTERACTIONS *
 923              **************************/
 924
 925             r12              = _mm_mul_pd(rsq12,rinv12);
 926
 927             /* EWALD ELECTROSTATICS */
 928
 929             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 930             ewrt             = _mm_mul_pd(r12,ewtabscale);
 931             ewitab           = _mm_cvttpd_epi32(ewrt);
 932 #ifdef __XOP__
 933             eweps            = _mm_frcz_pd(ewrt);
 934 #else
 935             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 936 #endif
 937             twoeweps         = _mm_add_pd(eweps,eweps);
 938             ewitab           = _mm_slli_epi32(ewitab,2);
 939             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 940             ewtabD           = _mm_setzero_pd();
 941             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 942             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 943             ewtabFn          = _mm_setzero_pd();
 944             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 945             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 946             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 947             velec            = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
 948             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
 949
 950             /* Update potential sum for this i atom from the interaction with this j atom. */
 951             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 952             velecsum         = _mm_add_pd(velecsum,velec);
 953
 954             fscal            = felec;
 955
 956             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 957
 958             /* Update vectorial force */
 959             fix1             = _mm_macc_pd(dx12,fscal,fix1);
 960             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
 961             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
 962
 963             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
 964             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
 965             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
 966
 967             /**************************
 968              * CALCULATE INTERACTIONS *
 969              **************************/
 970
 971             r13              = _mm_mul_pd(rsq13,rinv13);
 972
 973             /* EWALD ELECTROSTATICS */
 974
 975             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 976             ewrt             = _mm_mul_pd(r13,ewtabscale);
 977             ewitab           = _mm_cvttpd_epi32(ewrt);
 978 #ifdef __XOP__
 979             eweps            = _mm_frcz_pd(ewrt);
 980 #else
 981             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 982 #endif
 983             twoeweps         = _mm_add_pd(eweps,eweps);
 984             ewitab           = _mm_slli_epi32(ewitab,2);
 985             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 986             ewtabD           = _mm_setzero_pd();
 987             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 988             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 989             ewtabFn          = _mm_setzero_pd();
 990             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 991             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 992             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 993             velec            = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
 994             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
 995
 996             /* Update potential sum for this i atom from the interaction with this j atom. */
 997             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 998             velecsum         = _mm_add_pd(velecsum,velec);
 999
1000             fscal            = felec;
1001
1002             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1003
1004             /* Update vectorial force */
1005             fix1             = _mm_macc_pd(dx13,fscal,fix1);
1006             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
1007             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
1008
1009             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
1010             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
1011             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
1012
1013             /**************************
1014              * CALCULATE INTERACTIONS *
1015              **************************/
1016
1017             r21              = _mm_mul_pd(rsq21,rinv21);
1018
1019             /* EWALD ELECTROSTATICS */
1020
1021             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1022             ewrt             = _mm_mul_pd(r21,ewtabscale);
1023             ewitab           = _mm_cvttpd_epi32(ewrt);
1024 #ifdef __XOP__
1025             eweps            = _mm_frcz_pd(ewrt);
1026 #else
1027             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1028 #endif
1029             twoeweps         = _mm_add_pd(eweps,eweps);
1030             ewitab           = _mm_slli_epi32(ewitab,2);
1031             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1032             ewtabD           = _mm_setzero_pd();
1033             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1034             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1035             ewtabFn          = _mm_setzero_pd();
1036             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1037             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1038             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1039             velec            = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
1040             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
1041
1042             /* Update potential sum for this i atom from the interaction with this j atom. */
1043             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1044             velecsum         = _mm_add_pd(velecsum,velec);
1045
1046             fscal            = felec;
1047
1048             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1049
1050             /* Update vectorial force */
1051             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1052             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1053             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1054
1055             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1056             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1057             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1058
1059             /**************************
1060              * CALCULATE INTERACTIONS *
1061              **************************/
1062
1063             r22              = _mm_mul_pd(rsq22,rinv22);
1064
1065             /* EWALD ELECTROSTATICS */
1066
1067             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1068             ewrt             = _mm_mul_pd(r22,ewtabscale);
1069             ewitab           = _mm_cvttpd_epi32(ewrt);
1070 #ifdef __XOP__
1071             eweps            = _mm_frcz_pd(ewrt);
1072 #else
1073             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1074 #endif
1075             twoeweps         = _mm_add_pd(eweps,eweps);
1076             ewitab           = _mm_slli_epi32(ewitab,2);
1077             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1078             ewtabD           = _mm_setzero_pd();
1079             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1080             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1081             ewtabFn          = _mm_setzero_pd();
1082             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1083             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1084             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1085             velec            = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
1086             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
1087
1088             /* Update potential sum for this i atom from the interaction with this j atom. */
1089             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1090             velecsum         = _mm_add_pd(velecsum,velec);
1091
1092             fscal            = felec;
1093
1094             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1095
1096             /* Update vectorial force */
1097             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1098             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1099             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1100
1101             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1102             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1103             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1104
1105             /**************************
1106              * CALCULATE INTERACTIONS *
1107              **************************/
1108
1109             r23              = _mm_mul_pd(rsq23,rinv23);
1110
1111             /* EWALD ELECTROSTATICS */
1112
1113             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1114             ewrt             = _mm_mul_pd(r23,ewtabscale);
1115             ewitab           = _mm_cvttpd_epi32(ewrt);
1116 #ifdef __XOP__
1117             eweps            = _mm_frcz_pd(ewrt);
1118 #else
1119             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1120 #endif
1121             twoeweps         = _mm_add_pd(eweps,eweps);
1122             ewitab           = _mm_slli_epi32(ewitab,2);
1123             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1124             ewtabD           = _mm_setzero_pd();
1125             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1126             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1127             ewtabFn          = _mm_setzero_pd();
1128             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1129             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1130             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1131             velec            = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
1132             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
1133
1134             /* Update potential sum for this i atom from the interaction with this j atom. */
1135             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1136             velecsum         = _mm_add_pd(velecsum,velec);
1137
1138             fscal            = felec;
1139
1140             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1141
1142             /* Update vectorial force */
1143             fix2             = _mm_macc_pd(dx23,fscal,fix2);
1144             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
1145             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
1146
1147             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
1148             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
1149             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
1150
1151             /**************************
1152              * CALCULATE INTERACTIONS *
1153              **************************/
1154
1155             r31              = _mm_mul_pd(rsq31,rinv31);
1156
1157             /* EWALD ELECTROSTATICS */
1158
1159             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1160             ewrt             = _mm_mul_pd(r31,ewtabscale);
1161             ewitab           = _mm_cvttpd_epi32(ewrt);
1162 #ifdef __XOP__
1163             eweps            = _mm_frcz_pd(ewrt);
1164 #else
1165             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1166 #endif
1167             twoeweps         = _mm_add_pd(eweps,eweps);
1168             ewitab           = _mm_slli_epi32(ewitab,2);
1169             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1170             ewtabD           = _mm_setzero_pd();
1171             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1172             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1173             ewtabFn          = _mm_setzero_pd();
1174             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1175             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1176             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1177             velec            = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
1178             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
1179
1180             /* Update potential sum for this i atom from the interaction with this j atom. */
1181             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1182             velecsum         = _mm_add_pd(velecsum,velec);
1183
1184             fscal            = felec;
1185
1186             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1187
1188             /* Update vectorial force */
1189             fix3             = _mm_macc_pd(dx31,fscal,fix3);
1190             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
1191             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
1192
1193             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
1194             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
1195             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
1196
1197             /**************************
1198              * CALCULATE INTERACTIONS *
1199              **************************/
1200
1201             r32              = _mm_mul_pd(rsq32,rinv32);
1202
1203             /* EWALD ELECTROSTATICS */
1204
1205             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1206             ewrt             = _mm_mul_pd(r32,ewtabscale);
1207             ewitab           = _mm_cvttpd_epi32(ewrt);
1208 #ifdef __XOP__
1209             eweps            = _mm_frcz_pd(ewrt);
1210 #else
1211             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1212 #endif
1213             twoeweps         = _mm_add_pd(eweps,eweps);
1214             ewitab           = _mm_slli_epi32(ewitab,2);
1215             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1216             ewtabD           = _mm_setzero_pd();
1217             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1218             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1219             ewtabFn          = _mm_setzero_pd();
1220             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1221             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1222             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1223             velec            = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
1224             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
1225
1226             /* Update potential sum for this i atom from the interaction with this j atom. */
1227             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1228             velecsum         = _mm_add_pd(velecsum,velec);
1229
1230             fscal            = felec;
1231
1232             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1233
1234             /* Update vectorial force */
1235             fix3             = _mm_macc_pd(dx32,fscal,fix3);
1236             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
1237             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
1238
1239             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
1240             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
1241             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
1242
1243             /**************************
1244              * CALCULATE INTERACTIONS *
1245              **************************/
1246
1247             r33              = _mm_mul_pd(rsq33,rinv33);
1248
1249             /* EWALD ELECTROSTATICS */
1250
1251             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1252             ewrt             = _mm_mul_pd(r33,ewtabscale);
1253             ewitab           = _mm_cvttpd_epi32(ewrt);
1254 #ifdef __XOP__
1255             eweps            = _mm_frcz_pd(ewrt);
1256 #else
1257             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1258 #endif
1259             twoeweps         = _mm_add_pd(eweps,eweps);
1260             ewitab           = _mm_slli_epi32(ewitab,2);
1261             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1262             ewtabD           = _mm_setzero_pd();
1263             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1264             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
1265             ewtabFn          = _mm_setzero_pd();
1266             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1267             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
1268             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
1269             velec            = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
1270             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
1271
1272             /* Update potential sum for this i atom from the interaction with this j atom. */
1273             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1274             velecsum         = _mm_add_pd(velecsum,velec);
1275
1276             fscal            = felec;
1277
1278             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1279
1280             /* Update vectorial force */
1281             fix3             = _mm_macc_pd(dx33,fscal,fix3);
1282             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
1283             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
1284
1285             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
1286             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
1287             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
1288
1289             gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1290
1291             /* Inner loop uses 449 flops */
1292         }
1293
1294         /* End of innermost loop */
1295
1296         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1297                                               f+i_coord_offset,fshift+i_shift_offset);
1298
1299         ggid                        = gid[iidx];
1300         /* Update potential energies */
1301         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1302         gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1303
1304         /* Increment number of inner iterations */
1305         inneriter                  += j_index_end - j_index_start;
1306
1307         /* Outer loop uses 26 flops */
1308     }
1309
1310     /* Increment number of outer iterations */
1311     outeriter        += nri;
1312
1313     /* Update outer/inner flops */
1314
1315     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*449);
1316 }
1317 /*
1318  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwLJEw_GeomW4W4_F_avx_128_fma_double
1319  * Electrostatics interaction: Ewald
1320  * VdW interaction:            LJEwald
1321  * Geometry:                   Water4-Water4
1322  * Calculate force/pot:        Force
1323  */
1324 void
1325 nb_kernel_ElecEw_VdwLJEw_GeomW4W4_F_avx_128_fma_double
1326                     (t_nblist                    * gmx_restrict       nlist,
1327                      rvec                        * gmx_restrict          xx,
1328                      rvec                        * gmx_restrict          ff,
1329                      t_forcerec                  * gmx_restrict          fr,
1330                      t_mdatoms                   * gmx_restrict     mdatoms,
1331                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1332                      t_nrnb                      * gmx_restrict        nrnb)
1333 {
1334     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1335      * just 0 for non-waters.
1336      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
1337      * jnr indices corresponding to data put in the four positions in the SIMD register.
1338      */
1339     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
1340     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1341     int              jnrA,jnrB;
1342     int              j_coord_offsetA,j_coord_offsetB;
1343     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
1344     real             rcutoff_scalar;
1345     real             *shiftvec,*fshift,*x,*f;
1346     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1347     int              vdwioffset0;
1348     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1349     int              vdwioffset1;
1350     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1351     int              vdwioffset2;
1352     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1353     int              vdwioffset3;
1354     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1355     int              vdwjidx0A,vdwjidx0B;
1356     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1357     int              vdwjidx1A,vdwjidx1B;
1358     __m128d          jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1359     int              vdwjidx2A,vdwjidx2B;
1360     __m128d          jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1361     int              vdwjidx3A,vdwjidx3B;
1362     __m128d          jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1363     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1364     __m128d          dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1365     __m128d          dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1366     __m128d          dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1367     __m128d          dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1368     __m128d          dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1369     __m128d          dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1370     __m128d          dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1371     __m128d          dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1372     __m128d          dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1373     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
1374     real             *charge;
1375     int              nvdwtype;
1376     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1377     int              *vdwtype;
1378     real             *vdwparam;
1379     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
1380     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
1381     __m128d           c6grid_00;
1382     __m128d           c6grid_11;
1383     __m128d           c6grid_12;
1384     __m128d           c6grid_13;
1385     __m128d           c6grid_21;
1386     __m128d           c6grid_22;
1387     __m128d           c6grid_23;
1388     __m128d           c6grid_31;
1389     __m128d           c6grid_32;
1390     __m128d           c6grid_33;
1391     real             *vdwgridparam;
1392     __m128d           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
1393     __m128d           one_half  = _mm_set1_pd(0.5);
1394     __m128d           minus_one = _mm_set1_pd(-1.0);
1395     __m128i          ewitab;
1396     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1397     real             *ewtab;
1398     __m128d          dummy_mask,cutoff_mask;
1399     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
1400     __m128d          one     = _mm_set1_pd(1.0);
1401     __m128d          two     = _mm_set1_pd(2.0);
1402     x                = xx[0];
1403     f                = ff[0];
1404
1405     nri              = nlist->nri;
1406     iinr             = nlist->iinr;
1407     jindex           = nlist->jindex;
1408     jjnr             = nlist->jjnr;
1409     shiftidx         = nlist->shift;
1410     gid              = nlist->gid;
1411     shiftvec         = fr->shift_vec[0];
1412     fshift           = fr->fshift[0];
1413     facel            = _mm_set1_pd(fr->epsfac);
1414     charge           = mdatoms->chargeA;
1415     nvdwtype         = fr->ntype;
1416     vdwparam         = fr->nbfp;
1417     vdwtype          = mdatoms->typeA;
1418     vdwgridparam     = fr->ljpme_c6grid;
1419     sh_lj_ewald      = _mm_set1_pd(fr->ic->sh_lj_ewald);
1420     ewclj            = _mm_set1_pd(fr->ewaldcoeff_lj);
1421     ewclj2           = _mm_mul_pd(minus_one,_mm_mul_pd(ewclj,ewclj));
1422
1423     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
1424     ewtab            = fr->ic->tabq_coul_F;
1425     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
1426     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
1427
1428     /* Setup water-specific parameters */
1429     inr              = nlist->iinr[0];
1430     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
1431     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
1432     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
1433     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
1434
1435     jq1              = _mm_set1_pd(charge[inr+1]);
1436     jq2              = _mm_set1_pd(charge[inr+2]);
1437     jq3              = _mm_set1_pd(charge[inr+3]);
1438     vdwjidx0A        = 2*vdwtype[inr+0];
1439     c6_00            = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
1440     c12_00           = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
1441     c6grid_00        = _mm_set1_pd(vdwgridparam[vdwioffset0+vdwjidx0A]);
1442     qq11             = _mm_mul_pd(iq1,jq1);
1443     qq12             = _mm_mul_pd(iq1,jq2);
1444     qq13             = _mm_mul_pd(iq1,jq3);
1445     qq21             = _mm_mul_pd(iq2,jq1);
1446     qq22             = _mm_mul_pd(iq2,jq2);
1447     qq23             = _mm_mul_pd(iq2,jq3);
1448     qq31             = _mm_mul_pd(iq3,jq1);
1449     qq32             = _mm_mul_pd(iq3,jq2);
1450     qq33             = _mm_mul_pd(iq3,jq3);
1451
1452     /* Avoid stupid compiler warnings */
1453     jnrA = jnrB = 0;
1454     j_coord_offsetA = 0;
1455     j_coord_offsetB = 0;
1456
1457     outeriter        = 0;
1458     inneriter        = 0;
1459
1460     /* Start outer loop over neighborlists */
1461     for(iidx=0; iidx<nri; iidx++)
1462     {
1463         /* Load shift vector for this list */
1464         i_shift_offset   = DIM*shiftidx[iidx];
1465
1466         /* Load limits for loop over neighbors */
1467         j_index_start    = jindex[iidx];
1468         j_index_end      = jindex[iidx+1];
1469
1470         /* Get outer coordinate index */
1471         inr              = iinr[iidx];
1472         i_coord_offset   = DIM*inr;
1473
1474         /* Load i particle coords and add shift vector */
1475         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1476                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1477
1478         fix0             = _mm_setzero_pd();
1479         fiy0             = _mm_setzero_pd();
1480         fiz0             = _mm_setzero_pd();
1481         fix1             = _mm_setzero_pd();
1482         fiy1             = _mm_setzero_pd();
1483         fiz1             = _mm_setzero_pd();
1484         fix2             = _mm_setzero_pd();
1485         fiy2             = _mm_setzero_pd();
1486         fiz2             = _mm_setzero_pd();
1487         fix3             = _mm_setzero_pd();
1488         fiy3             = _mm_setzero_pd();
1489         fiz3             = _mm_setzero_pd();
1490
1491         /* Start inner kernel loop */
1492         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
1493         {
1494
1495             /* Get j neighbor index, and coordinate index */
1496             jnrA             = jjnr[jidx];
1497             jnrB             = jjnr[jidx+1];
1498             j_coord_offsetA  = DIM*jnrA;
1499             j_coord_offsetB  = DIM*jnrB;
1500
1501             /* load j atom coordinates */
1502             gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1503                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1504                                               &jy2,&jz2,&jx3,&jy3,&jz3);
1505
1506             /* Calculate displacement vector */
1507             dx00             = _mm_sub_pd(ix0,jx0);
1508             dy00             = _mm_sub_pd(iy0,jy0);
1509             dz00             = _mm_sub_pd(iz0,jz0);
1510             dx11             = _mm_sub_pd(ix1,jx1);
1511             dy11             = _mm_sub_pd(iy1,jy1);
1512             dz11             = _mm_sub_pd(iz1,jz1);
1513             dx12             = _mm_sub_pd(ix1,jx2);
1514             dy12             = _mm_sub_pd(iy1,jy2);
1515             dz12             = _mm_sub_pd(iz1,jz2);
1516             dx13             = _mm_sub_pd(ix1,jx3);
1517             dy13             = _mm_sub_pd(iy1,jy3);
1518             dz13             = _mm_sub_pd(iz1,jz3);
1519             dx21             = _mm_sub_pd(ix2,jx1);
1520             dy21             = _mm_sub_pd(iy2,jy1);
1521             dz21             = _mm_sub_pd(iz2,jz1);
1522             dx22             = _mm_sub_pd(ix2,jx2);
1523             dy22             = _mm_sub_pd(iy2,jy2);
1524             dz22             = _mm_sub_pd(iz2,jz2);
1525             dx23             = _mm_sub_pd(ix2,jx3);
1526             dy23             = _mm_sub_pd(iy2,jy3);
1527             dz23             = _mm_sub_pd(iz2,jz3);
1528             dx31             = _mm_sub_pd(ix3,jx1);
1529             dy31             = _mm_sub_pd(iy3,jy1);
1530             dz31             = _mm_sub_pd(iz3,jz1);
1531             dx32             = _mm_sub_pd(ix3,jx2);
1532             dy32             = _mm_sub_pd(iy3,jy2);
1533             dz32             = _mm_sub_pd(iz3,jz2);
1534             dx33             = _mm_sub_pd(ix3,jx3);
1535             dy33             = _mm_sub_pd(iy3,jy3);
1536             dz33             = _mm_sub_pd(iz3,jz3);
1537
1538             /* Calculate squared distance and things based on it */
1539             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1540             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1541             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1542             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
1543             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1544             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1545             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
1546             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
1547             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
1548             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
1549
1550             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1551             rinv11           = gmx_mm_invsqrt_pd(rsq11);
1552             rinv12           = gmx_mm_invsqrt_pd(rsq12);
1553             rinv13           = gmx_mm_invsqrt_pd(rsq13);
1554             rinv21           = gmx_mm_invsqrt_pd(rsq21);
1555             rinv22           = gmx_mm_invsqrt_pd(rsq22);
1556             rinv23           = gmx_mm_invsqrt_pd(rsq23);
1557             rinv31           = gmx_mm_invsqrt_pd(rsq31);
1558             rinv32           = gmx_mm_invsqrt_pd(rsq32);
1559             rinv33           = gmx_mm_invsqrt_pd(rsq33);
1560
1561             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
1562             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
1563             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
1564             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
1565             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
1566             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
1567             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
1568             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
1569             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
1570             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
1571
1572             fjx0             = _mm_setzero_pd();
1573             fjy0             = _mm_setzero_pd();
1574             fjz0             = _mm_setzero_pd();
1575             fjx1             = _mm_setzero_pd();
1576             fjy1             = _mm_setzero_pd();
1577             fjz1             = _mm_setzero_pd();
1578             fjx2             = _mm_setzero_pd();
1579             fjy2             = _mm_setzero_pd();
1580             fjz2             = _mm_setzero_pd();
1581             fjx3             = _mm_setzero_pd();
1582             fjy3             = _mm_setzero_pd();
1583             fjz3             = _mm_setzero_pd();
1584
1585             /**************************
1586              * CALCULATE INTERACTIONS *
1587              **************************/
1588
1589             r00              = _mm_mul_pd(rsq00,rinv00);
1590
1591             /* Analytical LJ-PME */
1592             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1593             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
1594             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
1595             exponent         = gmx_simd_exp_d(ewcljrsq);
1596             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
1597             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
1598             /* f6A = 6 * C6grid * (1 - poly) */
1599             f6A              = _mm_mul_pd(c6grid_00,_mm_sub_pd(one,poly));
1600             /* f6B = C6grid * exponent * beta^6 */
1601             f6B              = _mm_mul_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6));
1602             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
1603             fvdw              = _mm_mul_pd(_mm_macc_pd(_mm_msub_pd(c12_00,rinvsix,_mm_sub_pd(c6_00,f6A)),rinvsix,f6B),rinvsq00);
1604
1605             fscal            = fvdw;
1606
1607             /* Update vectorial force */
1608             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1609             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1610             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1611
1612             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1613             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1614             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1615
1616             /**************************
1617              * CALCULATE INTERACTIONS *
1618              **************************/
1619
1620             r11              = _mm_mul_pd(rsq11,rinv11);
1621
1622             /* EWALD ELECTROSTATICS */
1623
1624             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1625             ewrt             = _mm_mul_pd(r11,ewtabscale);
1626             ewitab           = _mm_cvttpd_epi32(ewrt);
1627 #ifdef __XOP__
1628             eweps            = _mm_frcz_pd(ewrt);
1629 #else
1630             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1631 #endif
1632             twoeweps         = _mm_add_pd(eweps,eweps);
1633             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1634                                          &ewtabF,&ewtabFn);
1635             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1636             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
1637
1638             fscal            = felec;
1639
1640             /* Update vectorial force */
1641             fix1             = _mm_macc_pd(dx11,fscal,fix1);
1642             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
1643             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
1644
1645             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
1646             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
1647             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
1648
1649             /**************************
1650              * CALCULATE INTERACTIONS *
1651              **************************/
1652
1653             r12              = _mm_mul_pd(rsq12,rinv12);
1654
1655             /* EWALD ELECTROSTATICS */
1656
1657             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1658             ewrt             = _mm_mul_pd(r12,ewtabscale);
1659             ewitab           = _mm_cvttpd_epi32(ewrt);
1660 #ifdef __XOP__
1661             eweps            = _mm_frcz_pd(ewrt);
1662 #else
1663             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1664 #endif
1665             twoeweps         = _mm_add_pd(eweps,eweps);
1666             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1667                                          &ewtabF,&ewtabFn);
1668             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1669             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
1670
1671             fscal            = felec;
1672
1673             /* Update vectorial force */
1674             fix1             = _mm_macc_pd(dx12,fscal,fix1);
1675             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
1676             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
1677
1678             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
1679             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
1680             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
1681
1682             /**************************
1683              * CALCULATE INTERACTIONS *
1684              **************************/
1685
1686             r13              = _mm_mul_pd(rsq13,rinv13);
1687
1688             /* EWALD ELECTROSTATICS */
1689
1690             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1691             ewrt             = _mm_mul_pd(r13,ewtabscale);
1692             ewitab           = _mm_cvttpd_epi32(ewrt);
1693 #ifdef __XOP__
1694             eweps            = _mm_frcz_pd(ewrt);
1695 #else
1696             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1697 #endif
1698             twoeweps         = _mm_add_pd(eweps,eweps);
1699             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1700                                          &ewtabF,&ewtabFn);
1701             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1702             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
1703
1704             fscal            = felec;
1705
1706             /* Update vectorial force */
1707             fix1             = _mm_macc_pd(dx13,fscal,fix1);
1708             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
1709             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
1710
1711             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
1712             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
1713             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
1714
1715             /**************************
1716              * CALCULATE INTERACTIONS *
1717              **************************/
1718
1719             r21              = _mm_mul_pd(rsq21,rinv21);
1720
1721             /* EWALD ELECTROSTATICS */
1722
1723             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1724             ewrt             = _mm_mul_pd(r21,ewtabscale);
1725             ewitab           = _mm_cvttpd_epi32(ewrt);
1726 #ifdef __XOP__
1727             eweps            = _mm_frcz_pd(ewrt);
1728 #else
1729             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1730 #endif
1731             twoeweps         = _mm_add_pd(eweps,eweps);
1732             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1733                                          &ewtabF,&ewtabFn);
1734             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1735             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
1736
1737             fscal            = felec;
1738
1739             /* Update vectorial force */
1740             fix2             = _mm_macc_pd(dx21,fscal,fix2);
1741             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
1742             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
1743
1744             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
1745             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
1746             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
1747
1748             /**************************
1749              * CALCULATE INTERACTIONS *
1750              **************************/
1751
1752             r22              = _mm_mul_pd(rsq22,rinv22);
1753
1754             /* EWALD ELECTROSTATICS */
1755
1756             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1757             ewrt             = _mm_mul_pd(r22,ewtabscale);
1758             ewitab           = _mm_cvttpd_epi32(ewrt);
1759 #ifdef __XOP__
1760             eweps            = _mm_frcz_pd(ewrt);
1761 #else
1762             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1763 #endif
1764             twoeweps         = _mm_add_pd(eweps,eweps);
1765             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1766                                          &ewtabF,&ewtabFn);
1767             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1768             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
1769
1770             fscal            = felec;
1771
1772             /* Update vectorial force */
1773             fix2             = _mm_macc_pd(dx22,fscal,fix2);
1774             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
1775             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
1776
1777             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
1778             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
1779             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
1780
1781             /**************************
1782              * CALCULATE INTERACTIONS *
1783              **************************/
1784
1785             r23              = _mm_mul_pd(rsq23,rinv23);
1786
1787             /* EWALD ELECTROSTATICS */
1788
1789             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1790             ewrt             = _mm_mul_pd(r23,ewtabscale);
1791             ewitab           = _mm_cvttpd_epi32(ewrt);
1792 #ifdef __XOP__
1793             eweps            = _mm_frcz_pd(ewrt);
1794 #else
1795             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1796 #endif
1797             twoeweps         = _mm_add_pd(eweps,eweps);
1798             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1799                                          &ewtabF,&ewtabFn);
1800             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1801             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
1802
1803             fscal            = felec;
1804
1805             /* Update vectorial force */
1806             fix2             = _mm_macc_pd(dx23,fscal,fix2);
1807             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
1808             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
1809
1810             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
1811             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
1812             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
1813
1814             /**************************
1815              * CALCULATE INTERACTIONS *
1816              **************************/
1817
1818             r31              = _mm_mul_pd(rsq31,rinv31);
1819
1820             /* EWALD ELECTROSTATICS */
1821
1822             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1823             ewrt             = _mm_mul_pd(r31,ewtabscale);
1824             ewitab           = _mm_cvttpd_epi32(ewrt);
1825 #ifdef __XOP__
1826             eweps            = _mm_frcz_pd(ewrt);
1827 #else
1828             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1829 #endif
1830             twoeweps         = _mm_add_pd(eweps,eweps);
1831             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1832                                          &ewtabF,&ewtabFn);
1833             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1834             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
1835
1836             fscal            = felec;
1837
1838             /* Update vectorial force */
1839             fix3             = _mm_macc_pd(dx31,fscal,fix3);
1840             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
1841             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
1842
1843             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
1844             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
1845             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
1846
1847             /**************************
1848              * CALCULATE INTERACTIONS *
1849              **************************/
1850
1851             r32              = _mm_mul_pd(rsq32,rinv32);
1852
1853             /* EWALD ELECTROSTATICS */
1854
1855             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1856             ewrt             = _mm_mul_pd(r32,ewtabscale);
1857             ewitab           = _mm_cvttpd_epi32(ewrt);
1858 #ifdef __XOP__
1859             eweps            = _mm_frcz_pd(ewrt);
1860 #else
1861             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1862 #endif
1863             twoeweps         = _mm_add_pd(eweps,eweps);
1864             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1865                                          &ewtabF,&ewtabFn);
1866             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1867             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
1868
1869             fscal            = felec;
1870
1871             /* Update vectorial force */
1872             fix3             = _mm_macc_pd(dx32,fscal,fix3);
1873             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
1874             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
1875
1876             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
1877             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
1878             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
1879
1880             /**************************
1881              * CALCULATE INTERACTIONS *
1882              **************************/
1883
1884             r33              = _mm_mul_pd(rsq33,rinv33);
1885
1886             /* EWALD ELECTROSTATICS */
1887
1888             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1889             ewrt             = _mm_mul_pd(r33,ewtabscale);
1890             ewitab           = _mm_cvttpd_epi32(ewrt);
1891 #ifdef __XOP__
1892             eweps            = _mm_frcz_pd(ewrt);
1893 #else
1894             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1895 #endif
1896             twoeweps         = _mm_add_pd(eweps,eweps);
1897             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1898                                          &ewtabF,&ewtabFn);
1899             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1900             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
1901
1902             fscal            = felec;
1903
1904             /* Update vectorial force */
1905             fix3             = _mm_macc_pd(dx33,fscal,fix3);
1906             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
1907             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
1908
1909             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
1910             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
1911             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
1912
1913             gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1914
1915             /* Inner loop uses 401 flops */
1916         }
1917
1918         if(jidx<j_index_end)
1919         {
1920
1921             jnrA             = jjnr[jidx];
1922             j_coord_offsetA  = DIM*jnrA;
1923
1924             /* load j atom coordinates */
1925             gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
1926                                               &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1927                                               &jy2,&jz2,&jx3,&jy3,&jz3);
1928
1929             /* Calculate displacement vector */
1930             dx00             = _mm_sub_pd(ix0,jx0);
1931             dy00             = _mm_sub_pd(iy0,jy0);
1932             dz00             = _mm_sub_pd(iz0,jz0);
1933             dx11             = _mm_sub_pd(ix1,jx1);
1934             dy11             = _mm_sub_pd(iy1,jy1);
1935             dz11             = _mm_sub_pd(iz1,jz1);
1936             dx12             = _mm_sub_pd(ix1,jx2);
1937             dy12             = _mm_sub_pd(iy1,jy2);
1938             dz12             = _mm_sub_pd(iz1,jz2);
1939             dx13             = _mm_sub_pd(ix1,jx3);
1940             dy13             = _mm_sub_pd(iy1,jy3);
1941             dz13             = _mm_sub_pd(iz1,jz3);
1942             dx21             = _mm_sub_pd(ix2,jx1);
1943             dy21             = _mm_sub_pd(iy2,jy1);
1944             dz21             = _mm_sub_pd(iz2,jz1);
1945             dx22             = _mm_sub_pd(ix2,jx2);
1946             dy22             = _mm_sub_pd(iy2,jy2);
1947             dz22             = _mm_sub_pd(iz2,jz2);
1948             dx23             = _mm_sub_pd(ix2,jx3);
1949             dy23             = _mm_sub_pd(iy2,jy3);
1950             dz23             = _mm_sub_pd(iz2,jz3);
1951             dx31             = _mm_sub_pd(ix3,jx1);
1952             dy31             = _mm_sub_pd(iy3,jy1);
1953             dz31             = _mm_sub_pd(iz3,jz1);
1954             dx32             = _mm_sub_pd(ix3,jx2);
1955             dy32             = _mm_sub_pd(iy3,jy2);
1956             dz32             = _mm_sub_pd(iz3,jz2);
1957             dx33             = _mm_sub_pd(ix3,jx3);
1958             dy33             = _mm_sub_pd(iy3,jy3);
1959             dz33             = _mm_sub_pd(iz3,jz3);
1960
1961             /* Calculate squared distance and things based on it */
1962             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1963             rsq11            = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1964             rsq12            = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1965             rsq13            = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
1966             rsq21            = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1967             rsq22            = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1968             rsq23            = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
1969             rsq31            = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
1970             rsq32            = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
1971             rsq33            = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
1972
1973             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1974             rinv11           = gmx_mm_invsqrt_pd(rsq11);
1975             rinv12           = gmx_mm_invsqrt_pd(rsq12);
1976             rinv13           = gmx_mm_invsqrt_pd(rsq13);
1977             rinv21           = gmx_mm_invsqrt_pd(rsq21);
1978             rinv22           = gmx_mm_invsqrt_pd(rsq22);
1979             rinv23           = gmx_mm_invsqrt_pd(rsq23);
1980             rinv31           = gmx_mm_invsqrt_pd(rsq31);
1981             rinv32           = gmx_mm_invsqrt_pd(rsq32);
1982             rinv33           = gmx_mm_invsqrt_pd(rsq33);
1983
1984             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
1985             rinvsq11         = _mm_mul_pd(rinv11,rinv11);
1986             rinvsq12         = _mm_mul_pd(rinv12,rinv12);
1987             rinvsq13         = _mm_mul_pd(rinv13,rinv13);
1988             rinvsq21         = _mm_mul_pd(rinv21,rinv21);
1989             rinvsq22         = _mm_mul_pd(rinv22,rinv22);
1990             rinvsq23         = _mm_mul_pd(rinv23,rinv23);
1991             rinvsq31         = _mm_mul_pd(rinv31,rinv31);
1992             rinvsq32         = _mm_mul_pd(rinv32,rinv32);
1993             rinvsq33         = _mm_mul_pd(rinv33,rinv33);
1994
1995             fjx0             = _mm_setzero_pd();
1996             fjy0             = _mm_setzero_pd();
1997             fjz0             = _mm_setzero_pd();
1998             fjx1             = _mm_setzero_pd();
1999             fjy1             = _mm_setzero_pd();
2000             fjz1             = _mm_setzero_pd();
2001             fjx2             = _mm_setzero_pd();
2002             fjy2             = _mm_setzero_pd();
2003             fjz2             = _mm_setzero_pd();
2004             fjx3             = _mm_setzero_pd();
2005             fjy3             = _mm_setzero_pd();
2006             fjz3             = _mm_setzero_pd();
2007
2008             /**************************
2009              * CALCULATE INTERACTIONS *
2010              **************************/
2011
2012             r00              = _mm_mul_pd(rsq00,rinv00);
2013
2014             /* Analytical LJ-PME */
2015             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
2016             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
2017             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
2018             exponent         = gmx_simd_exp_d(ewcljrsq);
2019             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
2020             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
2021             /* f6A = 6 * C6grid * (1 - poly) */
2022             f6A              = _mm_mul_pd(c6grid_00,_mm_sub_pd(one,poly));
2023             /* f6B = C6grid * exponent * beta^6 */
2024             f6B              = _mm_mul_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6));
2025             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
2026             fvdw              = _mm_mul_pd(_mm_macc_pd(_mm_msub_pd(c12_00,rinvsix,_mm_sub_pd(c6_00,f6A)),rinvsix,f6B),rinvsq00);
2027
2028             fscal            = fvdw;
2029
2030             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2031
2032             /* Update vectorial force */
2033             fix0             = _mm_macc_pd(dx00,fscal,fix0);
2034             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
2035             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
2036
2037             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
2038             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
2039             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
2040
2041             /**************************
2042              * CALCULATE INTERACTIONS *
2043              **************************/
2044
2045             r11              = _mm_mul_pd(rsq11,rinv11);
2046
2047             /* EWALD ELECTROSTATICS */
2048
2049             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2050             ewrt             = _mm_mul_pd(r11,ewtabscale);
2051             ewitab           = _mm_cvttpd_epi32(ewrt);
2052 #ifdef __XOP__
2053             eweps            = _mm_frcz_pd(ewrt);
2054 #else
2055             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2056 #endif
2057             twoeweps         = _mm_add_pd(eweps,eweps);
2058             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2059             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2060             felec            = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
2061
2062             fscal            = felec;
2063
2064             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2065
2066             /* Update vectorial force */
2067             fix1             = _mm_macc_pd(dx11,fscal,fix1);
2068             fiy1             = _mm_macc_pd(dy11,fscal,fiy1);
2069             fiz1             = _mm_macc_pd(dz11,fscal,fiz1);
2070
2071             fjx1             = _mm_macc_pd(dx11,fscal,fjx1);
2072             fjy1             = _mm_macc_pd(dy11,fscal,fjy1);
2073             fjz1             = _mm_macc_pd(dz11,fscal,fjz1);
2074
2075             /**************************
2076              * CALCULATE INTERACTIONS *
2077              **************************/
2078
2079             r12              = _mm_mul_pd(rsq12,rinv12);
2080
2081             /* EWALD ELECTROSTATICS */
2082
2083             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2084             ewrt             = _mm_mul_pd(r12,ewtabscale);
2085             ewitab           = _mm_cvttpd_epi32(ewrt);
2086 #ifdef __XOP__
2087             eweps            = _mm_frcz_pd(ewrt);
2088 #else
2089             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2090 #endif
2091             twoeweps         = _mm_add_pd(eweps,eweps);
2092             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2093             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2094             felec            = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
2095
2096             fscal            = felec;
2097
2098             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2099
2100             /* Update vectorial force */
2101             fix1             = _mm_macc_pd(dx12,fscal,fix1);
2102             fiy1             = _mm_macc_pd(dy12,fscal,fiy1);
2103             fiz1             = _mm_macc_pd(dz12,fscal,fiz1);
2104
2105             fjx2             = _mm_macc_pd(dx12,fscal,fjx2);
2106             fjy2             = _mm_macc_pd(dy12,fscal,fjy2);
2107             fjz2             = _mm_macc_pd(dz12,fscal,fjz2);
2108
2109             /**************************
2110              * CALCULATE INTERACTIONS *
2111              **************************/
2112
2113             r13              = _mm_mul_pd(rsq13,rinv13);
2114
2115             /* EWALD ELECTROSTATICS */
2116
2117             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2118             ewrt             = _mm_mul_pd(r13,ewtabscale);
2119             ewitab           = _mm_cvttpd_epi32(ewrt);
2120 #ifdef __XOP__
2121             eweps            = _mm_frcz_pd(ewrt);
2122 #else
2123             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2124 #endif
2125             twoeweps         = _mm_add_pd(eweps,eweps);
2126             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2127             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2128             felec            = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
2129
2130             fscal            = felec;
2131
2132             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2133
2134             /* Update vectorial force */
2135             fix1             = _mm_macc_pd(dx13,fscal,fix1);
2136             fiy1             = _mm_macc_pd(dy13,fscal,fiy1);
2137             fiz1             = _mm_macc_pd(dz13,fscal,fiz1);
2138
2139             fjx3             = _mm_macc_pd(dx13,fscal,fjx3);
2140             fjy3             = _mm_macc_pd(dy13,fscal,fjy3);
2141             fjz3             = _mm_macc_pd(dz13,fscal,fjz3);
2142
2143             /**************************
2144              * CALCULATE INTERACTIONS *
2145              **************************/
2146
2147             r21              = _mm_mul_pd(rsq21,rinv21);
2148
2149             /* EWALD ELECTROSTATICS */
2150
2151             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2152             ewrt             = _mm_mul_pd(r21,ewtabscale);
2153             ewitab           = _mm_cvttpd_epi32(ewrt);
2154 #ifdef __XOP__
2155             eweps            = _mm_frcz_pd(ewrt);
2156 #else
2157             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2158 #endif
2159             twoeweps         = _mm_add_pd(eweps,eweps);
2160             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2161             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2162             felec            = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
2163
2164             fscal            = felec;
2165
2166             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2167
2168             /* Update vectorial force */
2169             fix2             = _mm_macc_pd(dx21,fscal,fix2);
2170             fiy2             = _mm_macc_pd(dy21,fscal,fiy2);
2171             fiz2             = _mm_macc_pd(dz21,fscal,fiz2);
2172
2173             fjx1             = _mm_macc_pd(dx21,fscal,fjx1);
2174             fjy1             = _mm_macc_pd(dy21,fscal,fjy1);
2175             fjz1             = _mm_macc_pd(dz21,fscal,fjz1);
2176
2177             /**************************
2178              * CALCULATE INTERACTIONS *
2179              **************************/
2180
2181             r22              = _mm_mul_pd(rsq22,rinv22);
2182
2183             /* EWALD ELECTROSTATICS */
2184
2185             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2186             ewrt             = _mm_mul_pd(r22,ewtabscale);
2187             ewitab           = _mm_cvttpd_epi32(ewrt);
2188 #ifdef __XOP__
2189             eweps            = _mm_frcz_pd(ewrt);
2190 #else
2191             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2192 #endif
2193             twoeweps         = _mm_add_pd(eweps,eweps);
2194             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2195             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2196             felec            = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
2197
2198             fscal            = felec;
2199
2200             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2201
2202             /* Update vectorial force */
2203             fix2             = _mm_macc_pd(dx22,fscal,fix2);
2204             fiy2             = _mm_macc_pd(dy22,fscal,fiy2);
2205             fiz2             = _mm_macc_pd(dz22,fscal,fiz2);
2206
2207             fjx2             = _mm_macc_pd(dx22,fscal,fjx2);
2208             fjy2             = _mm_macc_pd(dy22,fscal,fjy2);
2209             fjz2             = _mm_macc_pd(dz22,fscal,fjz2);
2210
2211             /**************************
2212              * CALCULATE INTERACTIONS *
2213              **************************/
2214
2215             r23              = _mm_mul_pd(rsq23,rinv23);
2216
2217             /* EWALD ELECTROSTATICS */
2218
2219             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2220             ewrt             = _mm_mul_pd(r23,ewtabscale);
2221             ewitab           = _mm_cvttpd_epi32(ewrt);
2222 #ifdef __XOP__
2223             eweps            = _mm_frcz_pd(ewrt);
2224 #else
2225             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2226 #endif
2227             twoeweps         = _mm_add_pd(eweps,eweps);
2228             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2229             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2230             felec            = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
2231
2232             fscal            = felec;
2233
2234             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2235
2236             /* Update vectorial force */
2237             fix2             = _mm_macc_pd(dx23,fscal,fix2);
2238             fiy2             = _mm_macc_pd(dy23,fscal,fiy2);
2239             fiz2             = _mm_macc_pd(dz23,fscal,fiz2);
2240
2241             fjx3             = _mm_macc_pd(dx23,fscal,fjx3);
2242             fjy3             = _mm_macc_pd(dy23,fscal,fjy3);
2243             fjz3             = _mm_macc_pd(dz23,fscal,fjz3);
2244
2245             /**************************
2246              * CALCULATE INTERACTIONS *
2247              **************************/
2248
2249             r31              = _mm_mul_pd(rsq31,rinv31);
2250
2251             /* EWALD ELECTROSTATICS */
2252
2253             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2254             ewrt             = _mm_mul_pd(r31,ewtabscale);
2255             ewitab           = _mm_cvttpd_epi32(ewrt);
2256 #ifdef __XOP__
2257             eweps            = _mm_frcz_pd(ewrt);
2258 #else
2259             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2260 #endif
2261             twoeweps         = _mm_add_pd(eweps,eweps);
2262             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2263             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2264             felec            = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
2265
2266             fscal            = felec;
2267
2268             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2269
2270             /* Update vectorial force */
2271             fix3             = _mm_macc_pd(dx31,fscal,fix3);
2272             fiy3             = _mm_macc_pd(dy31,fscal,fiy3);
2273             fiz3             = _mm_macc_pd(dz31,fscal,fiz3);
2274
2275             fjx1             = _mm_macc_pd(dx31,fscal,fjx1);
2276             fjy1             = _mm_macc_pd(dy31,fscal,fjy1);
2277             fjz1             = _mm_macc_pd(dz31,fscal,fjz1);
2278
2279             /**************************
2280              * CALCULATE INTERACTIONS *
2281              **************************/
2282
2283             r32              = _mm_mul_pd(rsq32,rinv32);
2284
2285             /* EWALD ELECTROSTATICS */
2286
2287             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2288             ewrt             = _mm_mul_pd(r32,ewtabscale);
2289             ewitab           = _mm_cvttpd_epi32(ewrt);
2290 #ifdef __XOP__
2291             eweps            = _mm_frcz_pd(ewrt);
2292 #else
2293             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2294 #endif
2295             twoeweps         = _mm_add_pd(eweps,eweps);
2296             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2297             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2298             felec            = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
2299
2300             fscal            = felec;
2301
2302             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2303
2304             /* Update vectorial force */
2305             fix3             = _mm_macc_pd(dx32,fscal,fix3);
2306             fiy3             = _mm_macc_pd(dy32,fscal,fiy3);
2307             fiz3             = _mm_macc_pd(dz32,fscal,fiz3);
2308
2309             fjx2             = _mm_macc_pd(dx32,fscal,fjx2);
2310             fjy2             = _mm_macc_pd(dy32,fscal,fjy2);
2311             fjz2             = _mm_macc_pd(dz32,fscal,fjz2);
2312
2313             /**************************
2314              * CALCULATE INTERACTIONS *
2315              **************************/
2316
2317             r33              = _mm_mul_pd(rsq33,rinv33);
2318
2319             /* EWALD ELECTROSTATICS */
2320
2321             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2322             ewrt             = _mm_mul_pd(r33,ewtabscale);
2323             ewitab           = _mm_cvttpd_epi32(ewrt);
2324 #ifdef __XOP__
2325             eweps            = _mm_frcz_pd(ewrt);
2326 #else
2327             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
2328 #endif
2329             twoeweps         = _mm_add_pd(eweps,eweps);
2330             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
2331             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
2332             felec            = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
2333
2334             fscal            = felec;
2335
2336             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2337
2338             /* Update vectorial force */
2339             fix3             = _mm_macc_pd(dx33,fscal,fix3);
2340             fiy3             = _mm_macc_pd(dy33,fscal,fiy3);
2341             fiz3             = _mm_macc_pd(dz33,fscal,fiz3);
2342
2343             fjx3             = _mm_macc_pd(dx33,fscal,fjx3);
2344             fjy3             = _mm_macc_pd(dy33,fscal,fjy3);
2345             fjz3             = _mm_macc_pd(dz33,fscal,fjz3);
2346
2347             gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2348
2349             /* Inner loop uses 401 flops */
2350         }
2351
2352         /* End of innermost loop */
2353
2354         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
2355                                               f+i_coord_offset,fshift+i_shift_offset);
2356
2357         /* Increment number of inner iterations */
2358         inneriter                  += j_index_end - j_index_start;
2359
2360         /* Outer loop uses 24 flops */
2361     }
2362
2363     /* Increment number of outer iterations */
2364     outeriter        += nri;
2365
2366     /* Update outer/inner flops */
2367
2368     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*401);
2369 }