2 * This file is part of the GROMACS molecular simulation package.
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.
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.
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.
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.
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.
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.
36 * Note: this file was generated by the GROMACS avx_256_double kernel generator.
44 #include "../nb_kernel.h"
45 #include "types/simple.h"
49 #include "gromacs/simd/math_x86_avx_256_double.h"
50 #include "kernelutil_x86_avx_256_double.h"
53 * Gromacs nonbonded kernel: nb_kernel_ElecNone_VdwLJEwSh_GeomP1P1_VF_avx_256_double
54 * Electrostatics interaction: None
55 * VdW interaction: LJEwald
56 * Geometry: Particle-Particle
57 * Calculate force/pot: PotentialAndForce
60 nb_kernel_ElecNone_VdwLJEwSh_GeomP1P1_VF_avx_256_double
61 (t_nblist * gmx_restrict nlist,
62 rvec * gmx_restrict xx,
63 rvec * gmx_restrict ff,
64 t_forcerec * gmx_restrict fr,
65 t_mdatoms * gmx_restrict mdatoms,
66 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
67 t_nrnb * gmx_restrict nrnb)
69 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
70 * just 0 for non-waters.
71 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
72 * jnr indices corresponding to data put in the four positions in the SIMD register.
74 int i_shift_offset,i_coord_offset,outeriter,inneriter;
75 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
76 int jnrA,jnrB,jnrC,jnrD;
77 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
78 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
79 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
80 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
82 real *shiftvec,*fshift,*x,*f;
83 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
85 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
86 real * vdwioffsetptr0;
87 real * vdwgridioffsetptr0;
88 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
89 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
90 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
91 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
93 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
96 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
97 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
100 __m256d ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
101 __m256d one_half = _mm256_set1_pd(0.5);
102 __m256d minus_one = _mm256_set1_pd(-1.0);
103 __m256d dummy_mask,cutoff_mask;
104 __m128 tmpmask0,tmpmask1;
105 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
106 __m256d one = _mm256_set1_pd(1.0);
107 __m256d two = _mm256_set1_pd(2.0);
113 jindex = nlist->jindex;
115 shiftidx = nlist->shift;
117 shiftvec = fr->shift_vec[0];
118 fshift = fr->fshift[0];
119 nvdwtype = fr->ntype;
121 vdwtype = mdatoms->typeA;
122 vdwgridparam = fr->ljpme_c6grid;
123 sh_lj_ewald = _mm256_set1_pd(fr->ic->sh_lj_ewald);
124 ewclj = _mm256_set1_pd(fr->ewaldcoeff_lj);
125 ewclj2 = _mm256_mul_pd(minus_one,_mm256_mul_pd(ewclj,ewclj));
127 rcutoff_scalar = fr->rvdw;
128 rcutoff = _mm256_set1_pd(rcutoff_scalar);
129 rcutoff2 = _mm256_mul_pd(rcutoff,rcutoff);
131 sh_vdw_invrcut6 = _mm256_set1_pd(fr->ic->sh_invrc6);
132 rvdw = _mm256_set1_pd(fr->rvdw);
134 /* Avoid stupid compiler warnings */
135 jnrA = jnrB = jnrC = jnrD = 0;
144 for(iidx=0;iidx<4*DIM;iidx++)
149 /* Start outer loop over neighborlists */
150 for(iidx=0; iidx<nri; iidx++)
152 /* Load shift vector for this list */
153 i_shift_offset = DIM*shiftidx[iidx];
155 /* Load limits for loop over neighbors */
156 j_index_start = jindex[iidx];
157 j_index_end = jindex[iidx+1];
159 /* Get outer coordinate index */
161 i_coord_offset = DIM*inr;
163 /* Load i particle coords and add shift vector */
164 gmx_mm256_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
166 fix0 = _mm256_setzero_pd();
167 fiy0 = _mm256_setzero_pd();
168 fiz0 = _mm256_setzero_pd();
170 /* Load parameters for i particles */
171 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
172 vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
174 /* Reset potential sums */
175 vvdwsum = _mm256_setzero_pd();
177 /* Start inner kernel loop */
178 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
181 /* Get j neighbor index, and coordinate index */
186 j_coord_offsetA = DIM*jnrA;
187 j_coord_offsetB = DIM*jnrB;
188 j_coord_offsetC = DIM*jnrC;
189 j_coord_offsetD = DIM*jnrD;
191 /* load j atom coordinates */
192 gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
193 x+j_coord_offsetC,x+j_coord_offsetD,
196 /* Calculate displacement vector */
197 dx00 = _mm256_sub_pd(ix0,jx0);
198 dy00 = _mm256_sub_pd(iy0,jy0);
199 dz00 = _mm256_sub_pd(iz0,jz0);
201 /* Calculate squared distance and things based on it */
202 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
204 rinv00 = gmx_mm256_invsqrt_pd(rsq00);
206 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
208 /* Load parameters for j particles */
209 vdwjidx0A = 2*vdwtype[jnrA+0];
210 vdwjidx0B = 2*vdwtype[jnrB+0];
211 vdwjidx0C = 2*vdwtype[jnrC+0];
212 vdwjidx0D = 2*vdwtype[jnrD+0];
214 /**************************
215 * CALCULATE INTERACTIONS *
216 **************************/
218 if (gmx_mm256_any_lt(rsq00,rcutoff2))
221 r00 = _mm256_mul_pd(rsq00,rinv00);
223 /* Compute parameters for interactions between i and j atoms */
224 gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
225 vdwioffsetptr0+vdwjidx0B,
226 vdwioffsetptr0+vdwjidx0C,
227 vdwioffsetptr0+vdwjidx0D,
230 c6grid_00 = gmx_mm256_load_4real_swizzle_pd(vdwgridioffsetptr0+vdwjidx0A,
231 vdwgridioffsetptr0+vdwjidx0B,
232 vdwgridioffsetptr0+vdwjidx0C,
233 vdwgridioffsetptr0+vdwjidx0D);
235 /* Analytical LJ-PME */
236 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
237 ewcljrsq = _mm256_mul_pd(ewclj2,rsq00);
238 ewclj6 = _mm256_mul_pd(ewclj2,_mm256_mul_pd(ewclj2,ewclj2));
239 exponent = gmx_simd_exp_d(ewcljrsq);
240 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
241 poly = _mm256_mul_pd(exponent,_mm256_add_pd(_mm256_sub_pd(one,ewcljrsq),_mm256_mul_pd(_mm256_mul_pd(ewcljrsq,ewcljrsq),one_half)));
242 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
243 vvdw6 = _mm256_mul_pd(_mm256_sub_pd(c6_00,_mm256_mul_pd(c6grid_00,_mm256_sub_pd(one,poly))),rinvsix);
244 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
245 vvdw = _mm256_sub_pd(_mm256_mul_pd( _mm256_sub_pd(vvdw12 , _mm256_mul_pd(c12_00,_mm256_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
246 _mm256_mul_pd( _mm256_sub_pd(vvdw6,_mm256_add_pd(_mm256_mul_pd(c6_00,sh_vdw_invrcut6),_mm256_mul_pd(c6grid_00,sh_lj_ewald))),one_sixth));
247 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
248 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,_mm256_sub_pd(vvdw6,_mm256_mul_pd(_mm256_mul_pd(c6grid_00,one_sixth),_mm256_mul_pd(exponent,ewclj6)))),rinvsq00);
250 cutoff_mask = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
252 /* Update potential sum for this i atom from the interaction with this j atom. */
253 vvdw = _mm256_and_pd(vvdw,cutoff_mask);
254 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
258 fscal = _mm256_and_pd(fscal,cutoff_mask);
260 /* Calculate temporary vectorial force */
261 tx = _mm256_mul_pd(fscal,dx00);
262 ty = _mm256_mul_pd(fscal,dy00);
263 tz = _mm256_mul_pd(fscal,dz00);
265 /* Update vectorial force */
266 fix0 = _mm256_add_pd(fix0,tx);
267 fiy0 = _mm256_add_pd(fiy0,ty);
268 fiz0 = _mm256_add_pd(fiz0,tz);
270 fjptrA = f+j_coord_offsetA;
271 fjptrB = f+j_coord_offsetB;
272 fjptrC = f+j_coord_offsetC;
273 fjptrD = f+j_coord_offsetD;
274 gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
278 /* Inner loop uses 62 flops */
284 /* Get j neighbor index, and coordinate index */
285 jnrlistA = jjnr[jidx];
286 jnrlistB = jjnr[jidx+1];
287 jnrlistC = jjnr[jidx+2];
288 jnrlistD = jjnr[jidx+3];
289 /* Sign of each element will be negative for non-real atoms.
290 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
291 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
293 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
295 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
296 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
297 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
299 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
300 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
301 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
302 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
303 j_coord_offsetA = DIM*jnrA;
304 j_coord_offsetB = DIM*jnrB;
305 j_coord_offsetC = DIM*jnrC;
306 j_coord_offsetD = DIM*jnrD;
308 /* load j atom coordinates */
309 gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
310 x+j_coord_offsetC,x+j_coord_offsetD,
313 /* Calculate displacement vector */
314 dx00 = _mm256_sub_pd(ix0,jx0);
315 dy00 = _mm256_sub_pd(iy0,jy0);
316 dz00 = _mm256_sub_pd(iz0,jz0);
318 /* Calculate squared distance and things based on it */
319 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
321 rinv00 = gmx_mm256_invsqrt_pd(rsq00);
323 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
325 /* Load parameters for j particles */
326 vdwjidx0A = 2*vdwtype[jnrA+0];
327 vdwjidx0B = 2*vdwtype[jnrB+0];
328 vdwjidx0C = 2*vdwtype[jnrC+0];
329 vdwjidx0D = 2*vdwtype[jnrD+0];
331 /**************************
332 * CALCULATE INTERACTIONS *
333 **************************/
335 if (gmx_mm256_any_lt(rsq00,rcutoff2))
338 r00 = _mm256_mul_pd(rsq00,rinv00);
339 r00 = _mm256_andnot_pd(dummy_mask,r00);
341 /* Compute parameters for interactions between i and j atoms */
342 gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
343 vdwioffsetptr0+vdwjidx0B,
344 vdwioffsetptr0+vdwjidx0C,
345 vdwioffsetptr0+vdwjidx0D,
348 c6grid_00 = gmx_mm256_load_4real_swizzle_pd(vdwgridioffsetptr0+vdwjidx0A,
349 vdwgridioffsetptr0+vdwjidx0B,
350 vdwgridioffsetptr0+vdwjidx0C,
351 vdwgridioffsetptr0+vdwjidx0D);
353 /* Analytical LJ-PME */
354 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
355 ewcljrsq = _mm256_mul_pd(ewclj2,rsq00);
356 ewclj6 = _mm256_mul_pd(ewclj2,_mm256_mul_pd(ewclj2,ewclj2));
357 exponent = gmx_simd_exp_d(ewcljrsq);
358 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
359 poly = _mm256_mul_pd(exponent,_mm256_add_pd(_mm256_sub_pd(one,ewcljrsq),_mm256_mul_pd(_mm256_mul_pd(ewcljrsq,ewcljrsq),one_half)));
360 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
361 vvdw6 = _mm256_mul_pd(_mm256_sub_pd(c6_00,_mm256_mul_pd(c6grid_00,_mm256_sub_pd(one,poly))),rinvsix);
362 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
363 vvdw = _mm256_sub_pd(_mm256_mul_pd( _mm256_sub_pd(vvdw12 , _mm256_mul_pd(c12_00,_mm256_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6))), one_twelfth) ,
364 _mm256_mul_pd( _mm256_sub_pd(vvdw6,_mm256_add_pd(_mm256_mul_pd(c6_00,sh_vdw_invrcut6),_mm256_mul_pd(c6grid_00,sh_lj_ewald))),one_sixth));
365 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
366 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,_mm256_sub_pd(vvdw6,_mm256_mul_pd(_mm256_mul_pd(c6grid_00,one_sixth),_mm256_mul_pd(exponent,ewclj6)))),rinvsq00);
368 cutoff_mask = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
370 /* Update potential sum for this i atom from the interaction with this j atom. */
371 vvdw = _mm256_and_pd(vvdw,cutoff_mask);
372 vvdw = _mm256_andnot_pd(dummy_mask,vvdw);
373 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
377 fscal = _mm256_and_pd(fscal,cutoff_mask);
379 fscal = _mm256_andnot_pd(dummy_mask,fscal);
381 /* Calculate temporary vectorial force */
382 tx = _mm256_mul_pd(fscal,dx00);
383 ty = _mm256_mul_pd(fscal,dy00);
384 tz = _mm256_mul_pd(fscal,dz00);
386 /* Update vectorial force */
387 fix0 = _mm256_add_pd(fix0,tx);
388 fiy0 = _mm256_add_pd(fiy0,ty);
389 fiz0 = _mm256_add_pd(fiz0,tz);
391 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
392 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
393 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
394 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
395 gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
399 /* Inner loop uses 63 flops */
402 /* End of innermost loop */
404 gmx_mm256_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
405 f+i_coord_offset,fshift+i_shift_offset);
408 /* Update potential energies */
409 gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
411 /* Increment number of inner iterations */
412 inneriter += j_index_end - j_index_start;
414 /* Outer loop uses 7 flops */
417 /* Increment number of outer iterations */
420 /* Update outer/inner flops */
422 inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*7 + inneriter*63);
425 * Gromacs nonbonded kernel: nb_kernel_ElecNone_VdwLJEwSh_GeomP1P1_F_avx_256_double
426 * Electrostatics interaction: None
427 * VdW interaction: LJEwald
428 * Geometry: Particle-Particle
429 * Calculate force/pot: Force
432 nb_kernel_ElecNone_VdwLJEwSh_GeomP1P1_F_avx_256_double
433 (t_nblist * gmx_restrict nlist,
434 rvec * gmx_restrict xx,
435 rvec * gmx_restrict ff,
436 t_forcerec * gmx_restrict fr,
437 t_mdatoms * gmx_restrict mdatoms,
438 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
439 t_nrnb * gmx_restrict nrnb)
441 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
442 * just 0 for non-waters.
443 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
444 * jnr indices corresponding to data put in the four positions in the SIMD register.
446 int i_shift_offset,i_coord_offset,outeriter,inneriter;
447 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
448 int jnrA,jnrB,jnrC,jnrD;
449 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
450 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
451 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
452 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
454 real *shiftvec,*fshift,*x,*f;
455 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
457 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
458 real * vdwioffsetptr0;
459 real * vdwgridioffsetptr0;
460 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
461 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
462 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
463 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
465 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
468 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
469 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
472 __m256d ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
473 __m256d one_half = _mm256_set1_pd(0.5);
474 __m256d minus_one = _mm256_set1_pd(-1.0);
475 __m256d dummy_mask,cutoff_mask;
476 __m128 tmpmask0,tmpmask1;
477 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
478 __m256d one = _mm256_set1_pd(1.0);
479 __m256d two = _mm256_set1_pd(2.0);
485 jindex = nlist->jindex;
487 shiftidx = nlist->shift;
489 shiftvec = fr->shift_vec[0];
490 fshift = fr->fshift[0];
491 nvdwtype = fr->ntype;
493 vdwtype = mdatoms->typeA;
494 vdwgridparam = fr->ljpme_c6grid;
495 sh_lj_ewald = _mm256_set1_pd(fr->ic->sh_lj_ewald);
496 ewclj = _mm256_set1_pd(fr->ewaldcoeff_lj);
497 ewclj2 = _mm256_mul_pd(minus_one,_mm256_mul_pd(ewclj,ewclj));
499 rcutoff_scalar = fr->rvdw;
500 rcutoff = _mm256_set1_pd(rcutoff_scalar);
501 rcutoff2 = _mm256_mul_pd(rcutoff,rcutoff);
503 sh_vdw_invrcut6 = _mm256_set1_pd(fr->ic->sh_invrc6);
504 rvdw = _mm256_set1_pd(fr->rvdw);
506 /* Avoid stupid compiler warnings */
507 jnrA = jnrB = jnrC = jnrD = 0;
516 for(iidx=0;iidx<4*DIM;iidx++)
521 /* Start outer loop over neighborlists */
522 for(iidx=0; iidx<nri; iidx++)
524 /* Load shift vector for this list */
525 i_shift_offset = DIM*shiftidx[iidx];
527 /* Load limits for loop over neighbors */
528 j_index_start = jindex[iidx];
529 j_index_end = jindex[iidx+1];
531 /* Get outer coordinate index */
533 i_coord_offset = DIM*inr;
535 /* Load i particle coords and add shift vector */
536 gmx_mm256_load_shift_and_1rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,&ix0,&iy0,&iz0);
538 fix0 = _mm256_setzero_pd();
539 fiy0 = _mm256_setzero_pd();
540 fiz0 = _mm256_setzero_pd();
542 /* Load parameters for i particles */
543 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
544 vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
546 /* Start inner kernel loop */
547 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
550 /* Get j neighbor index, and coordinate index */
555 j_coord_offsetA = DIM*jnrA;
556 j_coord_offsetB = DIM*jnrB;
557 j_coord_offsetC = DIM*jnrC;
558 j_coord_offsetD = DIM*jnrD;
560 /* load j atom coordinates */
561 gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
562 x+j_coord_offsetC,x+j_coord_offsetD,
565 /* Calculate displacement vector */
566 dx00 = _mm256_sub_pd(ix0,jx0);
567 dy00 = _mm256_sub_pd(iy0,jy0);
568 dz00 = _mm256_sub_pd(iz0,jz0);
570 /* Calculate squared distance and things based on it */
571 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
573 rinv00 = gmx_mm256_invsqrt_pd(rsq00);
575 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
577 /* Load parameters for j particles */
578 vdwjidx0A = 2*vdwtype[jnrA+0];
579 vdwjidx0B = 2*vdwtype[jnrB+0];
580 vdwjidx0C = 2*vdwtype[jnrC+0];
581 vdwjidx0D = 2*vdwtype[jnrD+0];
583 /**************************
584 * CALCULATE INTERACTIONS *
585 **************************/
587 if (gmx_mm256_any_lt(rsq00,rcutoff2))
590 r00 = _mm256_mul_pd(rsq00,rinv00);
592 /* Compute parameters for interactions between i and j atoms */
593 gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
594 vdwioffsetptr0+vdwjidx0B,
595 vdwioffsetptr0+vdwjidx0C,
596 vdwioffsetptr0+vdwjidx0D,
599 c6grid_00 = gmx_mm256_load_4real_swizzle_pd(vdwgridioffsetptr0+vdwjidx0A,
600 vdwgridioffsetptr0+vdwjidx0B,
601 vdwgridioffsetptr0+vdwjidx0C,
602 vdwgridioffsetptr0+vdwjidx0D);
604 /* Analytical LJ-PME */
605 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
606 ewcljrsq = _mm256_mul_pd(ewclj2,rsq00);
607 ewclj6 = _mm256_mul_pd(ewclj2,_mm256_mul_pd(ewclj2,ewclj2));
608 exponent = gmx_simd_exp_d(ewcljrsq);
609 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
610 poly = _mm256_mul_pd(exponent,_mm256_add_pd(_mm256_sub_pd(one,ewcljrsq),_mm256_mul_pd(_mm256_mul_pd(ewcljrsq,ewcljrsq),one_half)));
611 /* f6A = 6 * C6grid * (1 - poly) */
612 f6A = _mm256_mul_pd(c6grid_00,_mm256_sub_pd(one,poly));
613 /* f6B = C6grid * exponent * beta^6 */
614 f6B = _mm256_mul_pd(_mm256_mul_pd(c6grid_00,one_sixth),_mm256_mul_pd(exponent,ewclj6));
615 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
616 fvdw = _mm256_mul_pd(_mm256_add_pd(_mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),_mm256_sub_pd(c6_00,f6A)),rinvsix),f6B),rinvsq00);
618 cutoff_mask = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
622 fscal = _mm256_and_pd(fscal,cutoff_mask);
624 /* Calculate temporary vectorial force */
625 tx = _mm256_mul_pd(fscal,dx00);
626 ty = _mm256_mul_pd(fscal,dy00);
627 tz = _mm256_mul_pd(fscal,dz00);
629 /* Update vectorial force */
630 fix0 = _mm256_add_pd(fix0,tx);
631 fiy0 = _mm256_add_pd(fiy0,ty);
632 fiz0 = _mm256_add_pd(fiz0,tz);
634 fjptrA = f+j_coord_offsetA;
635 fjptrB = f+j_coord_offsetB;
636 fjptrC = f+j_coord_offsetC;
637 fjptrD = f+j_coord_offsetD;
638 gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
642 /* Inner loop uses 49 flops */
648 /* Get j neighbor index, and coordinate index */
649 jnrlistA = jjnr[jidx];
650 jnrlistB = jjnr[jidx+1];
651 jnrlistC = jjnr[jidx+2];
652 jnrlistD = jjnr[jidx+3];
653 /* Sign of each element will be negative for non-real atoms.
654 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
655 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
657 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
659 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
660 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
661 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
663 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
664 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
665 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
666 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
667 j_coord_offsetA = DIM*jnrA;
668 j_coord_offsetB = DIM*jnrB;
669 j_coord_offsetC = DIM*jnrC;
670 j_coord_offsetD = DIM*jnrD;
672 /* load j atom coordinates */
673 gmx_mm256_load_1rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
674 x+j_coord_offsetC,x+j_coord_offsetD,
677 /* Calculate displacement vector */
678 dx00 = _mm256_sub_pd(ix0,jx0);
679 dy00 = _mm256_sub_pd(iy0,jy0);
680 dz00 = _mm256_sub_pd(iz0,jz0);
682 /* Calculate squared distance and things based on it */
683 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
685 rinv00 = gmx_mm256_invsqrt_pd(rsq00);
687 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
689 /* Load parameters for j particles */
690 vdwjidx0A = 2*vdwtype[jnrA+0];
691 vdwjidx0B = 2*vdwtype[jnrB+0];
692 vdwjidx0C = 2*vdwtype[jnrC+0];
693 vdwjidx0D = 2*vdwtype[jnrD+0];
695 /**************************
696 * CALCULATE INTERACTIONS *
697 **************************/
699 if (gmx_mm256_any_lt(rsq00,rcutoff2))
702 r00 = _mm256_mul_pd(rsq00,rinv00);
703 r00 = _mm256_andnot_pd(dummy_mask,r00);
705 /* Compute parameters for interactions between i and j atoms */
706 gmx_mm256_load_4pair_swizzle_pd(vdwioffsetptr0+vdwjidx0A,
707 vdwioffsetptr0+vdwjidx0B,
708 vdwioffsetptr0+vdwjidx0C,
709 vdwioffsetptr0+vdwjidx0D,
712 c6grid_00 = gmx_mm256_load_4real_swizzle_pd(vdwgridioffsetptr0+vdwjidx0A,
713 vdwgridioffsetptr0+vdwjidx0B,
714 vdwgridioffsetptr0+vdwjidx0C,
715 vdwgridioffsetptr0+vdwjidx0D);
717 /* Analytical LJ-PME */
718 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
719 ewcljrsq = _mm256_mul_pd(ewclj2,rsq00);
720 ewclj6 = _mm256_mul_pd(ewclj2,_mm256_mul_pd(ewclj2,ewclj2));
721 exponent = gmx_simd_exp_d(ewcljrsq);
722 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
723 poly = _mm256_mul_pd(exponent,_mm256_add_pd(_mm256_sub_pd(one,ewcljrsq),_mm256_mul_pd(_mm256_mul_pd(ewcljrsq,ewcljrsq),one_half)));
724 /* f6A = 6 * C6grid * (1 - poly) */
725 f6A = _mm256_mul_pd(c6grid_00,_mm256_sub_pd(one,poly));
726 /* f6B = C6grid * exponent * beta^6 */
727 f6B = _mm256_mul_pd(_mm256_mul_pd(c6grid_00,one_sixth),_mm256_mul_pd(exponent,ewclj6));
728 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
729 fvdw = _mm256_mul_pd(_mm256_add_pd(_mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),_mm256_sub_pd(c6_00,f6A)),rinvsix),f6B),rinvsq00);
731 cutoff_mask = _mm256_cmp_pd(rsq00,rcutoff2,_CMP_LT_OQ);
735 fscal = _mm256_and_pd(fscal,cutoff_mask);
737 fscal = _mm256_andnot_pd(dummy_mask,fscal);
739 /* Calculate temporary vectorial force */
740 tx = _mm256_mul_pd(fscal,dx00);
741 ty = _mm256_mul_pd(fscal,dy00);
742 tz = _mm256_mul_pd(fscal,dz00);
744 /* Update vectorial force */
745 fix0 = _mm256_add_pd(fix0,tx);
746 fiy0 = _mm256_add_pd(fiy0,ty);
747 fiz0 = _mm256_add_pd(fiz0,tz);
749 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
750 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
751 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
752 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
753 gmx_mm256_decrement_1rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,tx,ty,tz);
757 /* Inner loop uses 50 flops */
760 /* End of innermost loop */
762 gmx_mm256_update_iforce_1atom_swizzle_pd(fix0,fiy0,fiz0,
763 f+i_coord_offset,fshift+i_shift_offset);
765 /* Increment number of inner iterations */
766 inneriter += j_index_end - j_index_start;
768 /* Outer loop uses 6 flops */
771 /* Increment number of outer iterations */
774 /* Update outer/inner flops */
776 inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*6 + inneriter*50);