src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_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 "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  50 #include "kernelutil_x86_avx_128_fma_double.h"
  51
  52 /*
  53  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_VF_avx_128_fma_double
  54  * Electrostatics interaction: ReactionField
  55  * VdW interaction:            CubicSplineTable
  56  * Geometry:                   Particle-Particle
  57  * Calculate force/pot:        PotentialAndForce
  58  */
  59 void
  60 nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_VF_avx_128_fma_double
  61                     (t_nblist                    * gmx_restrict       nlist,
  62                      rvec                        * gmx_restrict          xx,
  63                      rvec                        * gmx_restrict          ff,
  64                      t_forcerec                  * gmx_restrict          fr,
  65                      t_mdatoms                   * gmx_restrict     mdatoms,
  66                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  67                      t_nrnb                      * gmx_restrict        nrnb)
  68 {
  69     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  70      * just 0 for non-waters.
  71      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
  72      * jnr indices corresponding to data put in the four positions in the SIMD register.
  73      */
  74     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  75     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  76     int              jnrA,jnrB;
  77     int              j_coord_offsetA,j_coord_offsetB;
  78     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  79     real             rcutoff_scalar;
  80     real             *shiftvec,*fshift,*x,*f;
  81     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  82     int              vdwioffset0;
  83     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  84     int              vdwjidx0A,vdwjidx0B;
  85     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  86     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  87     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
  88     real             *charge;
  89     int              nvdwtype;
  90     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  91     int              *vdwtype;
  92     real             *vdwparam;
  93     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
  94     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
  95     __m128i          vfitab;
  96     __m128i          ifour       = _mm_set1_epi32(4);
  97     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
  98     real             *vftab;
  99     __m128d          dummy_mask,cutoff_mask;
 100     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 101     __m128d          one     = _mm_set1_pd(1.0);
 102     __m128d          two     = _mm_set1_pd(2.0);
 103     x                = xx[0];
 104     f                = ff[0];
 105
 106     nri              = nlist->nri;
 107     iinr             = nlist->iinr;
 108     jindex           = nlist->jindex;
 109     jjnr             = nlist->jjnr;
 110     shiftidx         = nlist->shift;
 111     gid              = nlist->gid;
 112     shiftvec         = fr->shift_vec[0];
 113     fshift           = fr->fshift[0];
 114     facel            = _mm_set1_pd(fr->epsfac);
 115     charge           = mdatoms->chargeA;
 116     krf              = _mm_set1_pd(fr->ic->k_rf);
 117     krf2             = _mm_set1_pd(fr->ic->k_rf*2.0);
 118     crf              = _mm_set1_pd(fr->ic->c_rf);
 119     nvdwtype         = fr->ntype;
 120     vdwparam         = fr->nbfp;
 121     vdwtype          = mdatoms->typeA;
 122
 123     vftab            = kernel_data->table_vdw->data;
 124     vftabscale       = _mm_set1_pd(kernel_data->table_vdw->scale);
 125
 126     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 127     rcutoff_scalar   = fr->rcoulomb;
 128     rcutoff          = _mm_set1_pd(rcutoff_scalar);
 129     rcutoff2         = _mm_mul_pd(rcutoff,rcutoff);
 130
 131     /* Avoid stupid compiler warnings */
 132     jnrA = jnrB = 0;
 133     j_coord_offsetA = 0;
 134     j_coord_offsetB = 0;
 135
 136     outeriter        = 0;
 137     inneriter        = 0;
 138
 139     /* Start outer loop over neighborlists */
 140     for(iidx=0; iidx<nri; iidx++)
 141     {
 142         /* Load shift vector for this list */
 143         i_shift_offset   = DIM*shiftidx[iidx];
 144
 145         /* Load limits for loop over neighbors */
 146         j_index_start    = jindex[iidx];
 147         j_index_end      = jindex[iidx+1];
 148
 149         /* Get outer coordinate index */
 150         inr              = iinr[iidx];
 151         i_coord_offset   = DIM*inr;
 152
 153         /* Load i particle coords and add shift vector */
 154         gmx_mm_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 155
 156         fix0             = _mm_setzero_pd();
 157         fiy0             = _mm_setzero_pd();
 158         fiz0             = _mm_setzero_pd();
 159
 160         /* Load parameters for i particles */
 161         iq0              = _mm_mul_pd(facel,_mm_load1_pd(charge+inr+0));
 162         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 163
 164         /* Reset potential sums */
 165         velecsum         = _mm_setzero_pd();
 166         vvdwsum          = _mm_setzero_pd();
 167
 168         /* Start inner kernel loop */
 169         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 170         {
 171
 172             /* Get j neighbor index, and coordinate index */
 173             jnrA             = jjnr[jidx];
 174             jnrB             = jjnr[jidx+1];
 175             j_coord_offsetA  = DIM*jnrA;
 176             j_coord_offsetB  = DIM*jnrB;
 177
 178             /* load j atom coordinates */
 179             gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 180                                               &jx0,&jy0,&jz0);
 181
 182             /* Calculate displacement vector */
 183             dx00             = _mm_sub_pd(ix0,jx0);
 184             dy00             = _mm_sub_pd(iy0,jy0);
 185             dz00             = _mm_sub_pd(iz0,jz0);
 186
 187             /* Calculate squared distance and things based on it */
 188             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 189
 190             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 191
 192             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 193
 194             /* Load parameters for j particles */
 195             jq0              = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
 196             vdwjidx0A        = 2*vdwtype[jnrA+0];
 197             vdwjidx0B        = 2*vdwtype[jnrB+0];
 198
 199             /**************************
 200              * CALCULATE INTERACTIONS *
 201              **************************/
 202
 203             if (gmx_mm_any_lt(rsq00,rcutoff2))
 204             {
 205
 206             r00              = _mm_mul_pd(rsq00,rinv00);
 207
 208             /* Compute parameters for interactions between i and j atoms */
 209             qq00             = _mm_mul_pd(iq0,jq0);
 210             gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
 211                                          vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
 212
 213             /* Calculate table index by multiplying r with table scale and truncate to integer */
 214             rt               = _mm_mul_pd(r00,vftabscale);
 215             vfitab           = _mm_cvttpd_epi32(rt);
 216 #ifdef __XOP__
 217             vfeps            = _mm_frcz_pd(rt);
 218 #else
 219             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 220 #endif
 221             twovfeps         = _mm_add_pd(vfeps,vfeps);
 222             vfitab           = _mm_slli_epi32(vfitab,3);
 223
 224             /* REACTION-FIELD ELECTROSTATICS */
 225             velec            = _mm_mul_pd(qq00,_mm_sub_pd(_mm_macc_pd(krf,rsq00,rinv00),crf));
 226             felec            = _mm_mul_pd(qq00,_mm_msub_pd(rinv00,rinvsq00,krf2));
 227
 228             /* CUBIC SPLINE TABLE DISPERSION */
 229             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 230             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 231             GMX_MM_TRANSPOSE2_PD(Y,F);
 232             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 233             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 234             GMX_MM_TRANSPOSE2_PD(G,H);
 235             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 236             VV               = _mm_macc_pd(vfeps,Fp,Y);
 237             vvdw6            = _mm_mul_pd(c6_00,VV);
 238             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 239             fvdw6            = _mm_mul_pd(c6_00,FF);
 240
 241             /* CUBIC SPLINE TABLE REPULSION */
 242             vfitab           = _mm_add_epi32(vfitab,ifour);
 243             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 244             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 245             GMX_MM_TRANSPOSE2_PD(Y,F);
 246             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 247             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 248             GMX_MM_TRANSPOSE2_PD(G,H);
 249             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 250             VV               = _mm_macc_pd(vfeps,Fp,Y);
 251             vvdw12           = _mm_mul_pd(c12_00,VV);
 252             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 253             fvdw12           = _mm_mul_pd(c12_00,FF);
 254             vvdw             = _mm_add_pd(vvdw12,vvdw6);
 255             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 256
 257             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 258
 259             /* Update potential sum for this i atom from the interaction with this j atom. */
 260             velec            = _mm_and_pd(velec,cutoff_mask);
 261             velecsum         = _mm_add_pd(velecsum,velec);
 262             vvdw             = _mm_and_pd(vvdw,cutoff_mask);
 263             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 264
 265             fscal            = _mm_add_pd(felec,fvdw);
 266
 267             fscal            = _mm_and_pd(fscal,cutoff_mask);
 268
 269             /* Update vectorial force */
 270             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 271             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 272             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 273
 274             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,
 275                                                    _mm_mul_pd(dx00,fscal),
 276                                                    _mm_mul_pd(dy00,fscal),
 277                                                    _mm_mul_pd(dz00,fscal));
 278
 279             }
 280
 281             /* Inner loop uses 75 flops */
 282         }
 283
 284         if(jidx<j_index_end)
 285         {
 286
 287             jnrA             = jjnr[jidx];
 288             j_coord_offsetA  = DIM*jnrA;
 289
 290             /* load j atom coordinates */
 291             gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 292                                               &jx0,&jy0,&jz0);
 293
 294             /* Calculate displacement vector */
 295             dx00             = _mm_sub_pd(ix0,jx0);
 296             dy00             = _mm_sub_pd(iy0,jy0);
 297             dz00             = _mm_sub_pd(iz0,jz0);
 298
 299             /* Calculate squared distance and things based on it */
 300             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 301
 302             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 303
 304             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 305
 306             /* Load parameters for j particles */
 307             jq0              = _mm_load_sd(charge+jnrA+0);
 308             vdwjidx0A        = 2*vdwtype[jnrA+0];
 309
 310             /**************************
 311              * CALCULATE INTERACTIONS *
 312              **************************/
 313
 314             if (gmx_mm_any_lt(rsq00,rcutoff2))
 315             {
 316
 317             r00              = _mm_mul_pd(rsq00,rinv00);
 318
 319             /* Compute parameters for interactions between i and j atoms */
 320             qq00             = _mm_mul_pd(iq0,jq0);
 321             gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
 322
 323             /* Calculate table index by multiplying r with table scale and truncate to integer */
 324             rt               = _mm_mul_pd(r00,vftabscale);
 325             vfitab           = _mm_cvttpd_epi32(rt);
 326 #ifdef __XOP__
 327             vfeps            = _mm_frcz_pd(rt);
 328 #else
 329             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 330 #endif
 331             twovfeps         = _mm_add_pd(vfeps,vfeps);
 332             vfitab           = _mm_slli_epi32(vfitab,3);
 333
 334             /* REACTION-FIELD ELECTROSTATICS */
 335             velec            = _mm_mul_pd(qq00,_mm_sub_pd(_mm_macc_pd(krf,rsq00,rinv00),crf));
 336             felec            = _mm_mul_pd(qq00,_mm_msub_pd(rinv00,rinvsq00,krf2));
 337
 338             /* CUBIC SPLINE TABLE DISPERSION */
 339             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 340             F                = _mm_setzero_pd();
 341             GMX_MM_TRANSPOSE2_PD(Y,F);
 342             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 343             H                = _mm_setzero_pd();
 344             GMX_MM_TRANSPOSE2_PD(G,H);
 345             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 346             VV               = _mm_macc_pd(vfeps,Fp,Y);
 347             vvdw6            = _mm_mul_pd(c6_00,VV);
 348             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 349             fvdw6            = _mm_mul_pd(c6_00,FF);
 350
 351             /* CUBIC SPLINE TABLE REPULSION */
 352             vfitab           = _mm_add_epi32(vfitab,ifour);
 353             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 354             F                = _mm_setzero_pd();
 355             GMX_MM_TRANSPOSE2_PD(Y,F);
 356             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 357             H                = _mm_setzero_pd();
 358             GMX_MM_TRANSPOSE2_PD(G,H);
 359             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 360             VV               = _mm_macc_pd(vfeps,Fp,Y);
 361             vvdw12           = _mm_mul_pd(c12_00,VV);
 362             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 363             fvdw12           = _mm_mul_pd(c12_00,FF);
 364             vvdw             = _mm_add_pd(vvdw12,vvdw6);
 365             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 366
 367             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 368
 369             /* Update potential sum for this i atom from the interaction with this j atom. */
 370             velec            = _mm_and_pd(velec,cutoff_mask);
 371             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 372             velecsum         = _mm_add_pd(velecsum,velec);
 373             vvdw             = _mm_and_pd(vvdw,cutoff_mask);
 374             vvdw             = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
 375             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 376
 377             fscal            = _mm_add_pd(felec,fvdw);
 378
 379             fscal            = _mm_and_pd(fscal,cutoff_mask);
 380
 381             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 382
 383             /* Update vectorial force */
 384             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 385             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 386             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 387
 388             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,
 389                                                    _mm_mul_pd(dx00,fscal),
 390                                                    _mm_mul_pd(dy00,fscal),
 391                                                    _mm_mul_pd(dz00,fscal));
 392
 393             }
 394
 395             /* Inner loop uses 75 flops */
 396         }
 397
 398         /* End of innermost loop */
 399
 400         gmx_mm_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
 401                                               f+i_coord_offset,fshift+i_shift_offset);
 402
 403         ggid                        = gid[iidx];
 404         /* Update potential energies */
 405         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 406         gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
 407
 408         /* Increment number of inner iterations */
 409         inneriter                  += j_index_end - j_index_start;
 410
 411         /* Outer loop uses 9 flops */
 412     }
 413
 414     /* Increment number of outer iterations */
 415     outeriter        += nri;
 416
 417     /* Update outer/inner flops */
 418
 419     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*9 + inneriter*75);
 420 }
 421 /*
 422  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_F_avx_128_fma_double
 423  * Electrostatics interaction: ReactionField
 424  * VdW interaction:            CubicSplineTable
 425  * Geometry:                   Particle-Particle
 426  * Calculate force/pot:        Force
 427  */
 428 void
 429 nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_F_avx_128_fma_double
 430                     (t_nblist                    * gmx_restrict       nlist,
 431                      rvec                        * gmx_restrict          xx,
 432                      rvec                        * gmx_restrict          ff,
 433                      t_forcerec                  * gmx_restrict          fr,
 434                      t_mdatoms                   * gmx_restrict     mdatoms,
 435                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 436                      t_nrnb                      * gmx_restrict        nrnb)
 437 {
 438     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 439      * just 0 for non-waters.
 440      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
 441      * jnr indices corresponding to data put in the four positions in the SIMD register.
 442      */
 443     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 444     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 445     int              jnrA,jnrB;
 446     int              j_coord_offsetA,j_coord_offsetB;
 447     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 448     real             rcutoff_scalar;
 449     real             *shiftvec,*fshift,*x,*f;
 450     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 451     int              vdwioffset0;
 452     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 453     int              vdwjidx0A,vdwjidx0B;
 454     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 455     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 456     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 457     real             *charge;
 458     int              nvdwtype;
 459     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 460     int              *vdwtype;
 461     real             *vdwparam;
 462     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 463     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 464     __m128i          vfitab;
 465     __m128i          ifour       = _mm_set1_epi32(4);
 466     __m128d          rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF,twovfeps;
 467     real             *vftab;
 468     __m128d          dummy_mask,cutoff_mask;
 469     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 470     __m128d          one     = _mm_set1_pd(1.0);
 471     __m128d          two     = _mm_set1_pd(2.0);
 472     x                = xx[0];
 473     f                = ff[0];
 474
 475     nri              = nlist->nri;
 476     iinr             = nlist->iinr;
 477     jindex           = nlist->jindex;
 478     jjnr             = nlist->jjnr;
 479     shiftidx         = nlist->shift;
 480     gid              = nlist->gid;
 481     shiftvec         = fr->shift_vec[0];
 482     fshift           = fr->fshift[0];
 483     facel            = _mm_set1_pd(fr->epsfac);
 484     charge           = mdatoms->chargeA;
 485     krf              = _mm_set1_pd(fr->ic->k_rf);
 486     krf2             = _mm_set1_pd(fr->ic->k_rf*2.0);
 487     crf              = _mm_set1_pd(fr->ic->c_rf);
 488     nvdwtype         = fr->ntype;
 489     vdwparam         = fr->nbfp;
 490     vdwtype          = mdatoms->typeA;
 491
 492     vftab            = kernel_data->table_vdw->data;
 493     vftabscale       = _mm_set1_pd(kernel_data->table_vdw->scale);
 494
 495     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 496     rcutoff_scalar   = fr->rcoulomb;
 497     rcutoff          = _mm_set1_pd(rcutoff_scalar);
 498     rcutoff2         = _mm_mul_pd(rcutoff,rcutoff);
 499
 500     /* Avoid stupid compiler warnings */
 501     jnrA = jnrB = 0;
 502     j_coord_offsetA = 0;
 503     j_coord_offsetB = 0;
 504
 505     outeriter        = 0;
 506     inneriter        = 0;
 507
 508     /* Start outer loop over neighborlists */
 509     for(iidx=0; iidx<nri; iidx++)
 510     {
 511         /* Load shift vector for this list */
 512         i_shift_offset   = DIM*shiftidx[iidx];
 513
 514         /* Load limits for loop over neighbors */
 515         j_index_start    = jindex[iidx];
 516         j_index_end      = jindex[iidx+1];
 517
 518         /* Get outer coordinate index */
 519         inr              = iinr[iidx];
 520         i_coord_offset   = DIM*inr;
 521
 522         /* Load i particle coords and add shift vector */
 523         gmx_mm_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
 524
 525         fix0             = _mm_setzero_pd();
 526         fiy0             = _mm_setzero_pd();
 527         fiz0             = _mm_setzero_pd();
 528
 529         /* Load parameters for i particles */
 530         iq0              = _mm_mul_pd(facel,_mm_load1_pd(charge+inr+0));
 531         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 532
 533         /* Start inner kernel loop */
 534         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 535         {
 536
 537             /* Get j neighbor index, and coordinate index */
 538             jnrA             = jjnr[jidx];
 539             jnrB             = jjnr[jidx+1];
 540             j_coord_offsetA  = DIM*jnrA;
 541             j_coord_offsetB  = DIM*jnrB;
 542
 543             /* load j atom coordinates */
 544             gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 545                                               &jx0,&jy0,&jz0);
 546
 547             /* Calculate displacement vector */
 548             dx00             = _mm_sub_pd(ix0,jx0);
 549             dy00             = _mm_sub_pd(iy0,jy0);
 550             dz00             = _mm_sub_pd(iz0,jz0);
 551
 552             /* Calculate squared distance and things based on it */
 553             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 554
 555             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 556
 557             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 558
 559             /* Load parameters for j particles */
 560             jq0              = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
 561             vdwjidx0A        = 2*vdwtype[jnrA+0];
 562             vdwjidx0B        = 2*vdwtype[jnrB+0];
 563
 564             /**************************
 565              * CALCULATE INTERACTIONS *
 566              **************************/
 567
 568             if (gmx_mm_any_lt(rsq00,rcutoff2))
 569             {
 570
 571             r00              = _mm_mul_pd(rsq00,rinv00);
 572
 573             /* Compute parameters for interactions between i and j atoms */
 574             qq00             = _mm_mul_pd(iq0,jq0);
 575             gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
 576                                          vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
 577
 578             /* Calculate table index by multiplying r with table scale and truncate to integer */
 579             rt               = _mm_mul_pd(r00,vftabscale);
 580             vfitab           = _mm_cvttpd_epi32(rt);
 581 #ifdef __XOP__
 582             vfeps            = _mm_frcz_pd(rt);
 583 #else
 584             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 585 #endif
 586             twovfeps         = _mm_add_pd(vfeps,vfeps);
 587             vfitab           = _mm_slli_epi32(vfitab,3);
 588
 589             /* REACTION-FIELD ELECTROSTATICS */
 590             felec            = _mm_mul_pd(qq00,_mm_msub_pd(rinv00,rinvsq00,krf2));
 591
 592             /* CUBIC SPLINE TABLE DISPERSION */
 593             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 594             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 595             GMX_MM_TRANSPOSE2_PD(Y,F);
 596             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 597             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 598             GMX_MM_TRANSPOSE2_PD(G,H);
 599             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 600             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 601             fvdw6            = _mm_mul_pd(c6_00,FF);
 602
 603             /* CUBIC SPLINE TABLE REPULSION */
 604             vfitab           = _mm_add_epi32(vfitab,ifour);
 605             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 606             F                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) );
 607             GMX_MM_TRANSPOSE2_PD(Y,F);
 608             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 609             H                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,1) +2);
 610             GMX_MM_TRANSPOSE2_PD(G,H);
 611             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 612             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 613             fvdw12           = _mm_mul_pd(c12_00,FF);
 614             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 615
 616             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 617
 618             fscal            = _mm_add_pd(felec,fvdw);
 619
 620             fscal            = _mm_and_pd(fscal,cutoff_mask);
 621
 622             /* Update vectorial force */
 623             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 624             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 625             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 626
 627             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,
 628                                                    _mm_mul_pd(dx00,fscal),
 629                                                    _mm_mul_pd(dy00,fscal),
 630                                                    _mm_mul_pd(dz00,fscal));
 631
 632             }
 633
 634             /* Inner loop uses 60 flops */
 635         }
 636
 637         if(jidx<j_index_end)
 638         {
 639
 640             jnrA             = jjnr[jidx];
 641             j_coord_offsetA  = DIM*jnrA;
 642
 643             /* load j atom coordinates */
 644             gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 645                                               &jx0,&jy0,&jz0);
 646
 647             /* Calculate displacement vector */
 648             dx00             = _mm_sub_pd(ix0,jx0);
 649             dy00             = _mm_sub_pd(iy0,jy0);
 650             dz00             = _mm_sub_pd(iz0,jz0);
 651
 652             /* Calculate squared distance and things based on it */
 653             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 654
 655             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 656
 657             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 658
 659             /* Load parameters for j particles */
 660             jq0              = _mm_load_sd(charge+jnrA+0);
 661             vdwjidx0A        = 2*vdwtype[jnrA+0];
 662
 663             /**************************
 664              * CALCULATE INTERACTIONS *
 665              **************************/
 666
 667             if (gmx_mm_any_lt(rsq00,rcutoff2))
 668             {
 669
 670             r00              = _mm_mul_pd(rsq00,rinv00);
 671
 672             /* Compute parameters for interactions between i and j atoms */
 673             qq00             = _mm_mul_pd(iq0,jq0);
 674             gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
 675
 676             /* Calculate table index by multiplying r with table scale and truncate to integer */
 677             rt               = _mm_mul_pd(r00,vftabscale);
 678             vfitab           = _mm_cvttpd_epi32(rt);
 679 #ifdef __XOP__
 680             vfeps            = _mm_frcz_pd(rt);
 681 #else
 682             vfeps            = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
 683 #endif
 684             twovfeps         = _mm_add_pd(vfeps,vfeps);
 685             vfitab           = _mm_slli_epi32(vfitab,3);
 686
 687             /* REACTION-FIELD ELECTROSTATICS */
 688             felec            = _mm_mul_pd(qq00,_mm_msub_pd(rinv00,rinvsq00,krf2));
 689
 690             /* CUBIC SPLINE TABLE DISPERSION */
 691             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 692             F                = _mm_setzero_pd();
 693             GMX_MM_TRANSPOSE2_PD(Y,F);
 694             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 695             H                = _mm_setzero_pd();
 696             GMX_MM_TRANSPOSE2_PD(G,H);
 697             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 698             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 699             fvdw6            = _mm_mul_pd(c6_00,FF);
 700
 701             /* CUBIC SPLINE TABLE REPULSION */
 702             vfitab           = _mm_add_epi32(vfitab,ifour);
 703             Y                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) );
 704             F                = _mm_setzero_pd();
 705             GMX_MM_TRANSPOSE2_PD(Y,F);
 706             G                = _mm_load_pd( vftab + _mm_extract_epi32(vfitab,0) +2);
 707             H                = _mm_setzero_pd();
 708             GMX_MM_TRANSPOSE2_PD(G,H);
 709             Fp               = _mm_macc_pd(vfeps,_mm_macc_pd(H,vfeps,G),F);
 710             FF               = _mm_macc_pd(vfeps,_mm_macc_pd(twovfeps,H,G),Fp);
 711             fvdw12           = _mm_mul_pd(c12_00,FF);
 712             fvdw             = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
 713
 714             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 715
 716             fscal            = _mm_add_pd(felec,fvdw);
 717
 718             fscal            = _mm_and_pd(fscal,cutoff_mask);
 719
 720             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 721
 722             /* Update vectorial force */
 723             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 724             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 725             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 726
 727             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,
 728                                                    _mm_mul_pd(dx00,fscal),
 729                                                    _mm_mul_pd(dy00,fscal),
 730                                                    _mm_mul_pd(dz00,fscal));
 731
 732             }
 733
 734             /* Inner loop uses 60 flops */
 735         }
 736
 737         /* End of innermost loop */
 738
 739         gmx_mm_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
 740                                               f+i_coord_offset,fshift+i_shift_offset);
 741
 742         /* Increment number of inner iterations */
 743         inneriter                  += j_index_end - j_index_start;
 744
 745         /* Outer loop uses 7 flops */
 746     }
 747
 748     /* Increment number of outer iterations */
 749     outeriter        += nri;
 750
 751     /* Update outer/inner flops */
 752
 753     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*7 + inneriter*60);
 754 }