2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2012,2013,2014,2015,2017, 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 "gromacs/gmxlib/nrnb.h"
47 #include "kernelutil_x86_avx_256_double.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecCoul_VdwLJ_GeomW3W3_VF_avx_256_double
51 * Electrostatics interaction: Coulomb
52 * VdW interaction: LennardJones
53 * Geometry: Water3-Water3
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecCoul_VdwLJ_GeomW3W3_VF_avx_256_double
58 (t_nblist * gmx_restrict nlist,
59 rvec * gmx_restrict xx,
60 rvec * gmx_restrict ff,
61 struct t_forcerec * gmx_restrict fr,
62 t_mdatoms * gmx_restrict mdatoms,
63 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
64 t_nrnb * gmx_restrict nrnb)
66 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
67 * just 0 for non-waters.
68 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
69 * jnr indices corresponding to data put in the four positions in the SIMD register.
71 int i_shift_offset,i_coord_offset,outeriter,inneriter;
72 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
73 int jnrA,jnrB,jnrC,jnrD;
74 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
75 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
76 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
77 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
79 real *shiftvec,*fshift,*x,*f;
80 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
82 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
83 real * vdwioffsetptr0;
84 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
85 real * vdwioffsetptr1;
86 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
87 real * vdwioffsetptr2;
88 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
89 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
90 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
91 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
92 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
93 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
94 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
95 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
96 __m256d dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
97 __m256d dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
98 __m256d dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
99 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
100 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
101 __m256d dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
102 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
103 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
104 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
107 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
110 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
111 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
112 __m256d dummy_mask,cutoff_mask;
113 __m128 tmpmask0,tmpmask1;
114 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
115 __m256d one = _mm256_set1_pd(1.0);
116 __m256d two = _mm256_set1_pd(2.0);
122 jindex = nlist->jindex;
124 shiftidx = nlist->shift;
126 shiftvec = fr->shift_vec[0];
127 fshift = fr->fshift[0];
128 facel = _mm256_set1_pd(fr->ic->epsfac);
129 charge = mdatoms->chargeA;
130 nvdwtype = fr->ntype;
132 vdwtype = mdatoms->typeA;
134 /* Setup water-specific parameters */
135 inr = nlist->iinr[0];
136 iq0 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
137 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
138 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
139 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
141 jq0 = _mm256_set1_pd(charge[inr+0]);
142 jq1 = _mm256_set1_pd(charge[inr+1]);
143 jq2 = _mm256_set1_pd(charge[inr+2]);
144 vdwjidx0A = 2*vdwtype[inr+0];
145 qq00 = _mm256_mul_pd(iq0,jq0);
146 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
147 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
148 qq01 = _mm256_mul_pd(iq0,jq1);
149 qq02 = _mm256_mul_pd(iq0,jq2);
150 qq10 = _mm256_mul_pd(iq1,jq0);
151 qq11 = _mm256_mul_pd(iq1,jq1);
152 qq12 = _mm256_mul_pd(iq1,jq2);
153 qq20 = _mm256_mul_pd(iq2,jq0);
154 qq21 = _mm256_mul_pd(iq2,jq1);
155 qq22 = _mm256_mul_pd(iq2,jq2);
157 /* Avoid stupid compiler warnings */
158 jnrA = jnrB = jnrC = jnrD = 0;
167 for(iidx=0;iidx<4*DIM;iidx++)
172 /* Start outer loop over neighborlists */
173 for(iidx=0; iidx<nri; iidx++)
175 /* Load shift vector for this list */
176 i_shift_offset = DIM*shiftidx[iidx];
178 /* Load limits for loop over neighbors */
179 j_index_start = jindex[iidx];
180 j_index_end = jindex[iidx+1];
182 /* Get outer coordinate index */
184 i_coord_offset = DIM*inr;
186 /* Load i particle coords and add shift vector */
187 gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
188 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
190 fix0 = _mm256_setzero_pd();
191 fiy0 = _mm256_setzero_pd();
192 fiz0 = _mm256_setzero_pd();
193 fix1 = _mm256_setzero_pd();
194 fiy1 = _mm256_setzero_pd();
195 fiz1 = _mm256_setzero_pd();
196 fix2 = _mm256_setzero_pd();
197 fiy2 = _mm256_setzero_pd();
198 fiz2 = _mm256_setzero_pd();
200 /* Reset potential sums */
201 velecsum = _mm256_setzero_pd();
202 vvdwsum = _mm256_setzero_pd();
204 /* Start inner kernel loop */
205 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
208 /* Get j neighbor index, and coordinate index */
213 j_coord_offsetA = DIM*jnrA;
214 j_coord_offsetB = DIM*jnrB;
215 j_coord_offsetC = DIM*jnrC;
216 j_coord_offsetD = DIM*jnrD;
218 /* load j atom coordinates */
219 gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
220 x+j_coord_offsetC,x+j_coord_offsetD,
221 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
223 /* Calculate displacement vector */
224 dx00 = _mm256_sub_pd(ix0,jx0);
225 dy00 = _mm256_sub_pd(iy0,jy0);
226 dz00 = _mm256_sub_pd(iz0,jz0);
227 dx01 = _mm256_sub_pd(ix0,jx1);
228 dy01 = _mm256_sub_pd(iy0,jy1);
229 dz01 = _mm256_sub_pd(iz0,jz1);
230 dx02 = _mm256_sub_pd(ix0,jx2);
231 dy02 = _mm256_sub_pd(iy0,jy2);
232 dz02 = _mm256_sub_pd(iz0,jz2);
233 dx10 = _mm256_sub_pd(ix1,jx0);
234 dy10 = _mm256_sub_pd(iy1,jy0);
235 dz10 = _mm256_sub_pd(iz1,jz0);
236 dx11 = _mm256_sub_pd(ix1,jx1);
237 dy11 = _mm256_sub_pd(iy1,jy1);
238 dz11 = _mm256_sub_pd(iz1,jz1);
239 dx12 = _mm256_sub_pd(ix1,jx2);
240 dy12 = _mm256_sub_pd(iy1,jy2);
241 dz12 = _mm256_sub_pd(iz1,jz2);
242 dx20 = _mm256_sub_pd(ix2,jx0);
243 dy20 = _mm256_sub_pd(iy2,jy0);
244 dz20 = _mm256_sub_pd(iz2,jz0);
245 dx21 = _mm256_sub_pd(ix2,jx1);
246 dy21 = _mm256_sub_pd(iy2,jy1);
247 dz21 = _mm256_sub_pd(iz2,jz1);
248 dx22 = _mm256_sub_pd(ix2,jx2);
249 dy22 = _mm256_sub_pd(iy2,jy2);
250 dz22 = _mm256_sub_pd(iz2,jz2);
252 /* Calculate squared distance and things based on it */
253 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
254 rsq01 = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
255 rsq02 = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
256 rsq10 = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
257 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
258 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
259 rsq20 = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
260 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
261 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
263 rinv00 = avx256_invsqrt_d(rsq00);
264 rinv01 = avx256_invsqrt_d(rsq01);
265 rinv02 = avx256_invsqrt_d(rsq02);
266 rinv10 = avx256_invsqrt_d(rsq10);
267 rinv11 = avx256_invsqrt_d(rsq11);
268 rinv12 = avx256_invsqrt_d(rsq12);
269 rinv20 = avx256_invsqrt_d(rsq20);
270 rinv21 = avx256_invsqrt_d(rsq21);
271 rinv22 = avx256_invsqrt_d(rsq22);
273 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
274 rinvsq01 = _mm256_mul_pd(rinv01,rinv01);
275 rinvsq02 = _mm256_mul_pd(rinv02,rinv02);
276 rinvsq10 = _mm256_mul_pd(rinv10,rinv10);
277 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
278 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
279 rinvsq20 = _mm256_mul_pd(rinv20,rinv20);
280 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
281 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
283 fjx0 = _mm256_setzero_pd();
284 fjy0 = _mm256_setzero_pd();
285 fjz0 = _mm256_setzero_pd();
286 fjx1 = _mm256_setzero_pd();
287 fjy1 = _mm256_setzero_pd();
288 fjz1 = _mm256_setzero_pd();
289 fjx2 = _mm256_setzero_pd();
290 fjy2 = _mm256_setzero_pd();
291 fjz2 = _mm256_setzero_pd();
293 /**************************
294 * CALCULATE INTERACTIONS *
295 **************************/
297 /* COULOMB ELECTROSTATICS */
298 velec = _mm256_mul_pd(qq00,rinv00);
299 felec = _mm256_mul_pd(velec,rinvsq00);
301 /* LENNARD-JONES DISPERSION/REPULSION */
303 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
304 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
305 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
306 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
307 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
309 /* Update potential sum for this i atom from the interaction with this j atom. */
310 velecsum = _mm256_add_pd(velecsum,velec);
311 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
313 fscal = _mm256_add_pd(felec,fvdw);
315 /* Calculate temporary vectorial force */
316 tx = _mm256_mul_pd(fscal,dx00);
317 ty = _mm256_mul_pd(fscal,dy00);
318 tz = _mm256_mul_pd(fscal,dz00);
320 /* Update vectorial force */
321 fix0 = _mm256_add_pd(fix0,tx);
322 fiy0 = _mm256_add_pd(fiy0,ty);
323 fiz0 = _mm256_add_pd(fiz0,tz);
325 fjx0 = _mm256_add_pd(fjx0,tx);
326 fjy0 = _mm256_add_pd(fjy0,ty);
327 fjz0 = _mm256_add_pd(fjz0,tz);
329 /**************************
330 * CALCULATE INTERACTIONS *
331 **************************/
333 /* COULOMB ELECTROSTATICS */
334 velec = _mm256_mul_pd(qq01,rinv01);
335 felec = _mm256_mul_pd(velec,rinvsq01);
337 /* Update potential sum for this i atom from the interaction with this j atom. */
338 velecsum = _mm256_add_pd(velecsum,velec);
342 /* Calculate temporary vectorial force */
343 tx = _mm256_mul_pd(fscal,dx01);
344 ty = _mm256_mul_pd(fscal,dy01);
345 tz = _mm256_mul_pd(fscal,dz01);
347 /* Update vectorial force */
348 fix0 = _mm256_add_pd(fix0,tx);
349 fiy0 = _mm256_add_pd(fiy0,ty);
350 fiz0 = _mm256_add_pd(fiz0,tz);
352 fjx1 = _mm256_add_pd(fjx1,tx);
353 fjy1 = _mm256_add_pd(fjy1,ty);
354 fjz1 = _mm256_add_pd(fjz1,tz);
356 /**************************
357 * CALCULATE INTERACTIONS *
358 **************************/
360 /* COULOMB ELECTROSTATICS */
361 velec = _mm256_mul_pd(qq02,rinv02);
362 felec = _mm256_mul_pd(velec,rinvsq02);
364 /* Update potential sum for this i atom from the interaction with this j atom. */
365 velecsum = _mm256_add_pd(velecsum,velec);
369 /* Calculate temporary vectorial force */
370 tx = _mm256_mul_pd(fscal,dx02);
371 ty = _mm256_mul_pd(fscal,dy02);
372 tz = _mm256_mul_pd(fscal,dz02);
374 /* Update vectorial force */
375 fix0 = _mm256_add_pd(fix0,tx);
376 fiy0 = _mm256_add_pd(fiy0,ty);
377 fiz0 = _mm256_add_pd(fiz0,tz);
379 fjx2 = _mm256_add_pd(fjx2,tx);
380 fjy2 = _mm256_add_pd(fjy2,ty);
381 fjz2 = _mm256_add_pd(fjz2,tz);
383 /**************************
384 * CALCULATE INTERACTIONS *
385 **************************/
387 /* COULOMB ELECTROSTATICS */
388 velec = _mm256_mul_pd(qq10,rinv10);
389 felec = _mm256_mul_pd(velec,rinvsq10);
391 /* Update potential sum for this i atom from the interaction with this j atom. */
392 velecsum = _mm256_add_pd(velecsum,velec);
396 /* Calculate temporary vectorial force */
397 tx = _mm256_mul_pd(fscal,dx10);
398 ty = _mm256_mul_pd(fscal,dy10);
399 tz = _mm256_mul_pd(fscal,dz10);
401 /* Update vectorial force */
402 fix1 = _mm256_add_pd(fix1,tx);
403 fiy1 = _mm256_add_pd(fiy1,ty);
404 fiz1 = _mm256_add_pd(fiz1,tz);
406 fjx0 = _mm256_add_pd(fjx0,tx);
407 fjy0 = _mm256_add_pd(fjy0,ty);
408 fjz0 = _mm256_add_pd(fjz0,tz);
410 /**************************
411 * CALCULATE INTERACTIONS *
412 **************************/
414 /* COULOMB ELECTROSTATICS */
415 velec = _mm256_mul_pd(qq11,rinv11);
416 felec = _mm256_mul_pd(velec,rinvsq11);
418 /* Update potential sum for this i atom from the interaction with this j atom. */
419 velecsum = _mm256_add_pd(velecsum,velec);
423 /* Calculate temporary vectorial force */
424 tx = _mm256_mul_pd(fscal,dx11);
425 ty = _mm256_mul_pd(fscal,dy11);
426 tz = _mm256_mul_pd(fscal,dz11);
428 /* Update vectorial force */
429 fix1 = _mm256_add_pd(fix1,tx);
430 fiy1 = _mm256_add_pd(fiy1,ty);
431 fiz1 = _mm256_add_pd(fiz1,tz);
433 fjx1 = _mm256_add_pd(fjx1,tx);
434 fjy1 = _mm256_add_pd(fjy1,ty);
435 fjz1 = _mm256_add_pd(fjz1,tz);
437 /**************************
438 * CALCULATE INTERACTIONS *
439 **************************/
441 /* COULOMB ELECTROSTATICS */
442 velec = _mm256_mul_pd(qq12,rinv12);
443 felec = _mm256_mul_pd(velec,rinvsq12);
445 /* Update potential sum for this i atom from the interaction with this j atom. */
446 velecsum = _mm256_add_pd(velecsum,velec);
450 /* Calculate temporary vectorial force */
451 tx = _mm256_mul_pd(fscal,dx12);
452 ty = _mm256_mul_pd(fscal,dy12);
453 tz = _mm256_mul_pd(fscal,dz12);
455 /* Update vectorial force */
456 fix1 = _mm256_add_pd(fix1,tx);
457 fiy1 = _mm256_add_pd(fiy1,ty);
458 fiz1 = _mm256_add_pd(fiz1,tz);
460 fjx2 = _mm256_add_pd(fjx2,tx);
461 fjy2 = _mm256_add_pd(fjy2,ty);
462 fjz2 = _mm256_add_pd(fjz2,tz);
464 /**************************
465 * CALCULATE INTERACTIONS *
466 **************************/
468 /* COULOMB ELECTROSTATICS */
469 velec = _mm256_mul_pd(qq20,rinv20);
470 felec = _mm256_mul_pd(velec,rinvsq20);
472 /* Update potential sum for this i atom from the interaction with this j atom. */
473 velecsum = _mm256_add_pd(velecsum,velec);
477 /* Calculate temporary vectorial force */
478 tx = _mm256_mul_pd(fscal,dx20);
479 ty = _mm256_mul_pd(fscal,dy20);
480 tz = _mm256_mul_pd(fscal,dz20);
482 /* Update vectorial force */
483 fix2 = _mm256_add_pd(fix2,tx);
484 fiy2 = _mm256_add_pd(fiy2,ty);
485 fiz2 = _mm256_add_pd(fiz2,tz);
487 fjx0 = _mm256_add_pd(fjx0,tx);
488 fjy0 = _mm256_add_pd(fjy0,ty);
489 fjz0 = _mm256_add_pd(fjz0,tz);
491 /**************************
492 * CALCULATE INTERACTIONS *
493 **************************/
495 /* COULOMB ELECTROSTATICS */
496 velec = _mm256_mul_pd(qq21,rinv21);
497 felec = _mm256_mul_pd(velec,rinvsq21);
499 /* Update potential sum for this i atom from the interaction with this j atom. */
500 velecsum = _mm256_add_pd(velecsum,velec);
504 /* Calculate temporary vectorial force */
505 tx = _mm256_mul_pd(fscal,dx21);
506 ty = _mm256_mul_pd(fscal,dy21);
507 tz = _mm256_mul_pd(fscal,dz21);
509 /* Update vectorial force */
510 fix2 = _mm256_add_pd(fix2,tx);
511 fiy2 = _mm256_add_pd(fiy2,ty);
512 fiz2 = _mm256_add_pd(fiz2,tz);
514 fjx1 = _mm256_add_pd(fjx1,tx);
515 fjy1 = _mm256_add_pd(fjy1,ty);
516 fjz1 = _mm256_add_pd(fjz1,tz);
518 /**************************
519 * CALCULATE INTERACTIONS *
520 **************************/
522 /* COULOMB ELECTROSTATICS */
523 velec = _mm256_mul_pd(qq22,rinv22);
524 felec = _mm256_mul_pd(velec,rinvsq22);
526 /* Update potential sum for this i atom from the interaction with this j atom. */
527 velecsum = _mm256_add_pd(velecsum,velec);
531 /* Calculate temporary vectorial force */
532 tx = _mm256_mul_pd(fscal,dx22);
533 ty = _mm256_mul_pd(fscal,dy22);
534 tz = _mm256_mul_pd(fscal,dz22);
536 /* Update vectorial force */
537 fix2 = _mm256_add_pd(fix2,tx);
538 fiy2 = _mm256_add_pd(fiy2,ty);
539 fiz2 = _mm256_add_pd(fiz2,tz);
541 fjx2 = _mm256_add_pd(fjx2,tx);
542 fjy2 = _mm256_add_pd(fjy2,ty);
543 fjz2 = _mm256_add_pd(fjz2,tz);
545 fjptrA = f+j_coord_offsetA;
546 fjptrB = f+j_coord_offsetB;
547 fjptrC = f+j_coord_offsetC;
548 fjptrD = f+j_coord_offsetD;
550 gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
551 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
553 /* Inner loop uses 255 flops */
559 /* Get j neighbor index, and coordinate index */
560 jnrlistA = jjnr[jidx];
561 jnrlistB = jjnr[jidx+1];
562 jnrlistC = jjnr[jidx+2];
563 jnrlistD = jjnr[jidx+3];
564 /* Sign of each element will be negative for non-real atoms.
565 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
566 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
568 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
570 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
571 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
572 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
574 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
575 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
576 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
577 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
578 j_coord_offsetA = DIM*jnrA;
579 j_coord_offsetB = DIM*jnrB;
580 j_coord_offsetC = DIM*jnrC;
581 j_coord_offsetD = DIM*jnrD;
583 /* load j atom coordinates */
584 gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
585 x+j_coord_offsetC,x+j_coord_offsetD,
586 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
588 /* Calculate displacement vector */
589 dx00 = _mm256_sub_pd(ix0,jx0);
590 dy00 = _mm256_sub_pd(iy0,jy0);
591 dz00 = _mm256_sub_pd(iz0,jz0);
592 dx01 = _mm256_sub_pd(ix0,jx1);
593 dy01 = _mm256_sub_pd(iy0,jy1);
594 dz01 = _mm256_sub_pd(iz0,jz1);
595 dx02 = _mm256_sub_pd(ix0,jx2);
596 dy02 = _mm256_sub_pd(iy0,jy2);
597 dz02 = _mm256_sub_pd(iz0,jz2);
598 dx10 = _mm256_sub_pd(ix1,jx0);
599 dy10 = _mm256_sub_pd(iy1,jy0);
600 dz10 = _mm256_sub_pd(iz1,jz0);
601 dx11 = _mm256_sub_pd(ix1,jx1);
602 dy11 = _mm256_sub_pd(iy1,jy1);
603 dz11 = _mm256_sub_pd(iz1,jz1);
604 dx12 = _mm256_sub_pd(ix1,jx2);
605 dy12 = _mm256_sub_pd(iy1,jy2);
606 dz12 = _mm256_sub_pd(iz1,jz2);
607 dx20 = _mm256_sub_pd(ix2,jx0);
608 dy20 = _mm256_sub_pd(iy2,jy0);
609 dz20 = _mm256_sub_pd(iz2,jz0);
610 dx21 = _mm256_sub_pd(ix2,jx1);
611 dy21 = _mm256_sub_pd(iy2,jy1);
612 dz21 = _mm256_sub_pd(iz2,jz1);
613 dx22 = _mm256_sub_pd(ix2,jx2);
614 dy22 = _mm256_sub_pd(iy2,jy2);
615 dz22 = _mm256_sub_pd(iz2,jz2);
617 /* Calculate squared distance and things based on it */
618 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
619 rsq01 = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
620 rsq02 = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
621 rsq10 = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
622 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
623 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
624 rsq20 = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
625 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
626 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
628 rinv00 = avx256_invsqrt_d(rsq00);
629 rinv01 = avx256_invsqrt_d(rsq01);
630 rinv02 = avx256_invsqrt_d(rsq02);
631 rinv10 = avx256_invsqrt_d(rsq10);
632 rinv11 = avx256_invsqrt_d(rsq11);
633 rinv12 = avx256_invsqrt_d(rsq12);
634 rinv20 = avx256_invsqrt_d(rsq20);
635 rinv21 = avx256_invsqrt_d(rsq21);
636 rinv22 = avx256_invsqrt_d(rsq22);
638 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
639 rinvsq01 = _mm256_mul_pd(rinv01,rinv01);
640 rinvsq02 = _mm256_mul_pd(rinv02,rinv02);
641 rinvsq10 = _mm256_mul_pd(rinv10,rinv10);
642 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
643 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
644 rinvsq20 = _mm256_mul_pd(rinv20,rinv20);
645 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
646 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
648 fjx0 = _mm256_setzero_pd();
649 fjy0 = _mm256_setzero_pd();
650 fjz0 = _mm256_setzero_pd();
651 fjx1 = _mm256_setzero_pd();
652 fjy1 = _mm256_setzero_pd();
653 fjz1 = _mm256_setzero_pd();
654 fjx2 = _mm256_setzero_pd();
655 fjy2 = _mm256_setzero_pd();
656 fjz2 = _mm256_setzero_pd();
658 /**************************
659 * CALCULATE INTERACTIONS *
660 **************************/
662 /* COULOMB ELECTROSTATICS */
663 velec = _mm256_mul_pd(qq00,rinv00);
664 felec = _mm256_mul_pd(velec,rinvsq00);
666 /* LENNARD-JONES DISPERSION/REPULSION */
668 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
669 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
670 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
671 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
672 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
674 /* Update potential sum for this i atom from the interaction with this j atom. */
675 velec = _mm256_andnot_pd(dummy_mask,velec);
676 velecsum = _mm256_add_pd(velecsum,velec);
677 vvdw = _mm256_andnot_pd(dummy_mask,vvdw);
678 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
680 fscal = _mm256_add_pd(felec,fvdw);
682 fscal = _mm256_andnot_pd(dummy_mask,fscal);
684 /* Calculate temporary vectorial force */
685 tx = _mm256_mul_pd(fscal,dx00);
686 ty = _mm256_mul_pd(fscal,dy00);
687 tz = _mm256_mul_pd(fscal,dz00);
689 /* Update vectorial force */
690 fix0 = _mm256_add_pd(fix0,tx);
691 fiy0 = _mm256_add_pd(fiy0,ty);
692 fiz0 = _mm256_add_pd(fiz0,tz);
694 fjx0 = _mm256_add_pd(fjx0,tx);
695 fjy0 = _mm256_add_pd(fjy0,ty);
696 fjz0 = _mm256_add_pd(fjz0,tz);
698 /**************************
699 * CALCULATE INTERACTIONS *
700 **************************/
702 /* COULOMB ELECTROSTATICS */
703 velec = _mm256_mul_pd(qq01,rinv01);
704 felec = _mm256_mul_pd(velec,rinvsq01);
706 /* Update potential sum for this i atom from the interaction with this j atom. */
707 velec = _mm256_andnot_pd(dummy_mask,velec);
708 velecsum = _mm256_add_pd(velecsum,velec);
712 fscal = _mm256_andnot_pd(dummy_mask,fscal);
714 /* Calculate temporary vectorial force */
715 tx = _mm256_mul_pd(fscal,dx01);
716 ty = _mm256_mul_pd(fscal,dy01);
717 tz = _mm256_mul_pd(fscal,dz01);
719 /* Update vectorial force */
720 fix0 = _mm256_add_pd(fix0,tx);
721 fiy0 = _mm256_add_pd(fiy0,ty);
722 fiz0 = _mm256_add_pd(fiz0,tz);
724 fjx1 = _mm256_add_pd(fjx1,tx);
725 fjy1 = _mm256_add_pd(fjy1,ty);
726 fjz1 = _mm256_add_pd(fjz1,tz);
728 /**************************
729 * CALCULATE INTERACTIONS *
730 **************************/
732 /* COULOMB ELECTROSTATICS */
733 velec = _mm256_mul_pd(qq02,rinv02);
734 felec = _mm256_mul_pd(velec,rinvsq02);
736 /* Update potential sum for this i atom from the interaction with this j atom. */
737 velec = _mm256_andnot_pd(dummy_mask,velec);
738 velecsum = _mm256_add_pd(velecsum,velec);
742 fscal = _mm256_andnot_pd(dummy_mask,fscal);
744 /* Calculate temporary vectorial force */
745 tx = _mm256_mul_pd(fscal,dx02);
746 ty = _mm256_mul_pd(fscal,dy02);
747 tz = _mm256_mul_pd(fscal,dz02);
749 /* Update vectorial force */
750 fix0 = _mm256_add_pd(fix0,tx);
751 fiy0 = _mm256_add_pd(fiy0,ty);
752 fiz0 = _mm256_add_pd(fiz0,tz);
754 fjx2 = _mm256_add_pd(fjx2,tx);
755 fjy2 = _mm256_add_pd(fjy2,ty);
756 fjz2 = _mm256_add_pd(fjz2,tz);
758 /**************************
759 * CALCULATE INTERACTIONS *
760 **************************/
762 /* COULOMB ELECTROSTATICS */
763 velec = _mm256_mul_pd(qq10,rinv10);
764 felec = _mm256_mul_pd(velec,rinvsq10);
766 /* Update potential sum for this i atom from the interaction with this j atom. */
767 velec = _mm256_andnot_pd(dummy_mask,velec);
768 velecsum = _mm256_add_pd(velecsum,velec);
772 fscal = _mm256_andnot_pd(dummy_mask,fscal);
774 /* Calculate temporary vectorial force */
775 tx = _mm256_mul_pd(fscal,dx10);
776 ty = _mm256_mul_pd(fscal,dy10);
777 tz = _mm256_mul_pd(fscal,dz10);
779 /* Update vectorial force */
780 fix1 = _mm256_add_pd(fix1,tx);
781 fiy1 = _mm256_add_pd(fiy1,ty);
782 fiz1 = _mm256_add_pd(fiz1,tz);
784 fjx0 = _mm256_add_pd(fjx0,tx);
785 fjy0 = _mm256_add_pd(fjy0,ty);
786 fjz0 = _mm256_add_pd(fjz0,tz);
788 /**************************
789 * CALCULATE INTERACTIONS *
790 **************************/
792 /* COULOMB ELECTROSTATICS */
793 velec = _mm256_mul_pd(qq11,rinv11);
794 felec = _mm256_mul_pd(velec,rinvsq11);
796 /* Update potential sum for this i atom from the interaction with this j atom. */
797 velec = _mm256_andnot_pd(dummy_mask,velec);
798 velecsum = _mm256_add_pd(velecsum,velec);
802 fscal = _mm256_andnot_pd(dummy_mask,fscal);
804 /* Calculate temporary vectorial force */
805 tx = _mm256_mul_pd(fscal,dx11);
806 ty = _mm256_mul_pd(fscal,dy11);
807 tz = _mm256_mul_pd(fscal,dz11);
809 /* Update vectorial force */
810 fix1 = _mm256_add_pd(fix1,tx);
811 fiy1 = _mm256_add_pd(fiy1,ty);
812 fiz1 = _mm256_add_pd(fiz1,tz);
814 fjx1 = _mm256_add_pd(fjx1,tx);
815 fjy1 = _mm256_add_pd(fjy1,ty);
816 fjz1 = _mm256_add_pd(fjz1,tz);
818 /**************************
819 * CALCULATE INTERACTIONS *
820 **************************/
822 /* COULOMB ELECTROSTATICS */
823 velec = _mm256_mul_pd(qq12,rinv12);
824 felec = _mm256_mul_pd(velec,rinvsq12);
826 /* Update potential sum for this i atom from the interaction with this j atom. */
827 velec = _mm256_andnot_pd(dummy_mask,velec);
828 velecsum = _mm256_add_pd(velecsum,velec);
832 fscal = _mm256_andnot_pd(dummy_mask,fscal);
834 /* Calculate temporary vectorial force */
835 tx = _mm256_mul_pd(fscal,dx12);
836 ty = _mm256_mul_pd(fscal,dy12);
837 tz = _mm256_mul_pd(fscal,dz12);
839 /* Update vectorial force */
840 fix1 = _mm256_add_pd(fix1,tx);
841 fiy1 = _mm256_add_pd(fiy1,ty);
842 fiz1 = _mm256_add_pd(fiz1,tz);
844 fjx2 = _mm256_add_pd(fjx2,tx);
845 fjy2 = _mm256_add_pd(fjy2,ty);
846 fjz2 = _mm256_add_pd(fjz2,tz);
848 /**************************
849 * CALCULATE INTERACTIONS *
850 **************************/
852 /* COULOMB ELECTROSTATICS */
853 velec = _mm256_mul_pd(qq20,rinv20);
854 felec = _mm256_mul_pd(velec,rinvsq20);
856 /* Update potential sum for this i atom from the interaction with this j atom. */
857 velec = _mm256_andnot_pd(dummy_mask,velec);
858 velecsum = _mm256_add_pd(velecsum,velec);
862 fscal = _mm256_andnot_pd(dummy_mask,fscal);
864 /* Calculate temporary vectorial force */
865 tx = _mm256_mul_pd(fscal,dx20);
866 ty = _mm256_mul_pd(fscal,dy20);
867 tz = _mm256_mul_pd(fscal,dz20);
869 /* Update vectorial force */
870 fix2 = _mm256_add_pd(fix2,tx);
871 fiy2 = _mm256_add_pd(fiy2,ty);
872 fiz2 = _mm256_add_pd(fiz2,tz);
874 fjx0 = _mm256_add_pd(fjx0,tx);
875 fjy0 = _mm256_add_pd(fjy0,ty);
876 fjz0 = _mm256_add_pd(fjz0,tz);
878 /**************************
879 * CALCULATE INTERACTIONS *
880 **************************/
882 /* COULOMB ELECTROSTATICS */
883 velec = _mm256_mul_pd(qq21,rinv21);
884 felec = _mm256_mul_pd(velec,rinvsq21);
886 /* Update potential sum for this i atom from the interaction with this j atom. */
887 velec = _mm256_andnot_pd(dummy_mask,velec);
888 velecsum = _mm256_add_pd(velecsum,velec);
892 fscal = _mm256_andnot_pd(dummy_mask,fscal);
894 /* Calculate temporary vectorial force */
895 tx = _mm256_mul_pd(fscal,dx21);
896 ty = _mm256_mul_pd(fscal,dy21);
897 tz = _mm256_mul_pd(fscal,dz21);
899 /* Update vectorial force */
900 fix2 = _mm256_add_pd(fix2,tx);
901 fiy2 = _mm256_add_pd(fiy2,ty);
902 fiz2 = _mm256_add_pd(fiz2,tz);
904 fjx1 = _mm256_add_pd(fjx1,tx);
905 fjy1 = _mm256_add_pd(fjy1,ty);
906 fjz1 = _mm256_add_pd(fjz1,tz);
908 /**************************
909 * CALCULATE INTERACTIONS *
910 **************************/
912 /* COULOMB ELECTROSTATICS */
913 velec = _mm256_mul_pd(qq22,rinv22);
914 felec = _mm256_mul_pd(velec,rinvsq22);
916 /* Update potential sum for this i atom from the interaction with this j atom. */
917 velec = _mm256_andnot_pd(dummy_mask,velec);
918 velecsum = _mm256_add_pd(velecsum,velec);
922 fscal = _mm256_andnot_pd(dummy_mask,fscal);
924 /* Calculate temporary vectorial force */
925 tx = _mm256_mul_pd(fscal,dx22);
926 ty = _mm256_mul_pd(fscal,dy22);
927 tz = _mm256_mul_pd(fscal,dz22);
929 /* Update vectorial force */
930 fix2 = _mm256_add_pd(fix2,tx);
931 fiy2 = _mm256_add_pd(fiy2,ty);
932 fiz2 = _mm256_add_pd(fiz2,tz);
934 fjx2 = _mm256_add_pd(fjx2,tx);
935 fjy2 = _mm256_add_pd(fjy2,ty);
936 fjz2 = _mm256_add_pd(fjz2,tz);
938 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
939 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
940 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
941 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
943 gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
944 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
946 /* Inner loop uses 255 flops */
949 /* End of innermost loop */
951 gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
952 f+i_coord_offset,fshift+i_shift_offset);
955 /* Update potential energies */
956 gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
957 gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
959 /* Increment number of inner iterations */
960 inneriter += j_index_end - j_index_start;
962 /* Outer loop uses 20 flops */
965 /* Increment number of outer iterations */
968 /* Update outer/inner flops */
970 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3W3_VF,outeriter*20 + inneriter*255);
973 * Gromacs nonbonded kernel: nb_kernel_ElecCoul_VdwLJ_GeomW3W3_F_avx_256_double
974 * Electrostatics interaction: Coulomb
975 * VdW interaction: LennardJones
976 * Geometry: Water3-Water3
977 * Calculate force/pot: Force
980 nb_kernel_ElecCoul_VdwLJ_GeomW3W3_F_avx_256_double
981 (t_nblist * gmx_restrict nlist,
982 rvec * gmx_restrict xx,
983 rvec * gmx_restrict ff,
984 struct t_forcerec * gmx_restrict fr,
985 t_mdatoms * gmx_restrict mdatoms,
986 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
987 t_nrnb * gmx_restrict nrnb)
989 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
990 * just 0 for non-waters.
991 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
992 * jnr indices corresponding to data put in the four positions in the SIMD register.
994 int i_shift_offset,i_coord_offset,outeriter,inneriter;
995 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
996 int jnrA,jnrB,jnrC,jnrD;
997 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
998 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
999 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1000 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
1001 real rcutoff_scalar;
1002 real *shiftvec,*fshift,*x,*f;
1003 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1004 real scratch[4*DIM];
1005 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1006 real * vdwioffsetptr0;
1007 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1008 real * vdwioffsetptr1;
1009 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1010 real * vdwioffsetptr2;
1011 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1012 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1013 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1014 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1015 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1016 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1017 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1018 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1019 __m256d dx01,dy01,dz01,rsq01,rinv01,rinvsq01,r01,qq01,c6_01,c12_01;
1020 __m256d dx02,dy02,dz02,rsq02,rinv02,rinvsq02,r02,qq02,c6_02,c12_02;
1021 __m256d dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
1022 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1023 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1024 __m256d dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
1025 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1026 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1027 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
1030 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1033 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
1034 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
1035 __m256d dummy_mask,cutoff_mask;
1036 __m128 tmpmask0,tmpmask1;
1037 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
1038 __m256d one = _mm256_set1_pd(1.0);
1039 __m256d two = _mm256_set1_pd(2.0);
1045 jindex = nlist->jindex;
1047 shiftidx = nlist->shift;
1049 shiftvec = fr->shift_vec[0];
1050 fshift = fr->fshift[0];
1051 facel = _mm256_set1_pd(fr->ic->epsfac);
1052 charge = mdatoms->chargeA;
1053 nvdwtype = fr->ntype;
1054 vdwparam = fr->nbfp;
1055 vdwtype = mdatoms->typeA;
1057 /* Setup water-specific parameters */
1058 inr = nlist->iinr[0];
1059 iq0 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+0]));
1060 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
1061 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
1062 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
1064 jq0 = _mm256_set1_pd(charge[inr+0]);
1065 jq1 = _mm256_set1_pd(charge[inr+1]);
1066 jq2 = _mm256_set1_pd(charge[inr+2]);
1067 vdwjidx0A = 2*vdwtype[inr+0];
1068 qq00 = _mm256_mul_pd(iq0,jq0);
1069 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
1070 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
1071 qq01 = _mm256_mul_pd(iq0,jq1);
1072 qq02 = _mm256_mul_pd(iq0,jq2);
1073 qq10 = _mm256_mul_pd(iq1,jq0);
1074 qq11 = _mm256_mul_pd(iq1,jq1);
1075 qq12 = _mm256_mul_pd(iq1,jq2);
1076 qq20 = _mm256_mul_pd(iq2,jq0);
1077 qq21 = _mm256_mul_pd(iq2,jq1);
1078 qq22 = _mm256_mul_pd(iq2,jq2);
1080 /* Avoid stupid compiler warnings */
1081 jnrA = jnrB = jnrC = jnrD = 0;
1082 j_coord_offsetA = 0;
1083 j_coord_offsetB = 0;
1084 j_coord_offsetC = 0;
1085 j_coord_offsetD = 0;
1090 for(iidx=0;iidx<4*DIM;iidx++)
1092 scratch[iidx] = 0.0;
1095 /* Start outer loop over neighborlists */
1096 for(iidx=0; iidx<nri; iidx++)
1098 /* Load shift vector for this list */
1099 i_shift_offset = DIM*shiftidx[iidx];
1101 /* Load limits for loop over neighbors */
1102 j_index_start = jindex[iidx];
1103 j_index_end = jindex[iidx+1];
1105 /* Get outer coordinate index */
1107 i_coord_offset = DIM*inr;
1109 /* Load i particle coords and add shift vector */
1110 gmx_mm256_load_shift_and_3rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1111 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2);
1113 fix0 = _mm256_setzero_pd();
1114 fiy0 = _mm256_setzero_pd();
1115 fiz0 = _mm256_setzero_pd();
1116 fix1 = _mm256_setzero_pd();
1117 fiy1 = _mm256_setzero_pd();
1118 fiz1 = _mm256_setzero_pd();
1119 fix2 = _mm256_setzero_pd();
1120 fiy2 = _mm256_setzero_pd();
1121 fiz2 = _mm256_setzero_pd();
1123 /* Start inner kernel loop */
1124 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1127 /* Get j neighbor index, and coordinate index */
1129 jnrB = jjnr[jidx+1];
1130 jnrC = jjnr[jidx+2];
1131 jnrD = jjnr[jidx+3];
1132 j_coord_offsetA = DIM*jnrA;
1133 j_coord_offsetB = DIM*jnrB;
1134 j_coord_offsetC = DIM*jnrC;
1135 j_coord_offsetD = DIM*jnrD;
1137 /* load j atom coordinates */
1138 gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1139 x+j_coord_offsetC,x+j_coord_offsetD,
1140 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1142 /* Calculate displacement vector */
1143 dx00 = _mm256_sub_pd(ix0,jx0);
1144 dy00 = _mm256_sub_pd(iy0,jy0);
1145 dz00 = _mm256_sub_pd(iz0,jz0);
1146 dx01 = _mm256_sub_pd(ix0,jx1);
1147 dy01 = _mm256_sub_pd(iy0,jy1);
1148 dz01 = _mm256_sub_pd(iz0,jz1);
1149 dx02 = _mm256_sub_pd(ix0,jx2);
1150 dy02 = _mm256_sub_pd(iy0,jy2);
1151 dz02 = _mm256_sub_pd(iz0,jz2);
1152 dx10 = _mm256_sub_pd(ix1,jx0);
1153 dy10 = _mm256_sub_pd(iy1,jy0);
1154 dz10 = _mm256_sub_pd(iz1,jz0);
1155 dx11 = _mm256_sub_pd(ix1,jx1);
1156 dy11 = _mm256_sub_pd(iy1,jy1);
1157 dz11 = _mm256_sub_pd(iz1,jz1);
1158 dx12 = _mm256_sub_pd(ix1,jx2);
1159 dy12 = _mm256_sub_pd(iy1,jy2);
1160 dz12 = _mm256_sub_pd(iz1,jz2);
1161 dx20 = _mm256_sub_pd(ix2,jx0);
1162 dy20 = _mm256_sub_pd(iy2,jy0);
1163 dz20 = _mm256_sub_pd(iz2,jz0);
1164 dx21 = _mm256_sub_pd(ix2,jx1);
1165 dy21 = _mm256_sub_pd(iy2,jy1);
1166 dz21 = _mm256_sub_pd(iz2,jz1);
1167 dx22 = _mm256_sub_pd(ix2,jx2);
1168 dy22 = _mm256_sub_pd(iy2,jy2);
1169 dz22 = _mm256_sub_pd(iz2,jz2);
1171 /* Calculate squared distance and things based on it */
1172 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1173 rsq01 = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
1174 rsq02 = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
1175 rsq10 = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1176 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1177 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1178 rsq20 = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1179 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1180 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1182 rinv00 = avx256_invsqrt_d(rsq00);
1183 rinv01 = avx256_invsqrt_d(rsq01);
1184 rinv02 = avx256_invsqrt_d(rsq02);
1185 rinv10 = avx256_invsqrt_d(rsq10);
1186 rinv11 = avx256_invsqrt_d(rsq11);
1187 rinv12 = avx256_invsqrt_d(rsq12);
1188 rinv20 = avx256_invsqrt_d(rsq20);
1189 rinv21 = avx256_invsqrt_d(rsq21);
1190 rinv22 = avx256_invsqrt_d(rsq22);
1192 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
1193 rinvsq01 = _mm256_mul_pd(rinv01,rinv01);
1194 rinvsq02 = _mm256_mul_pd(rinv02,rinv02);
1195 rinvsq10 = _mm256_mul_pd(rinv10,rinv10);
1196 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
1197 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
1198 rinvsq20 = _mm256_mul_pd(rinv20,rinv20);
1199 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
1200 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
1202 fjx0 = _mm256_setzero_pd();
1203 fjy0 = _mm256_setzero_pd();
1204 fjz0 = _mm256_setzero_pd();
1205 fjx1 = _mm256_setzero_pd();
1206 fjy1 = _mm256_setzero_pd();
1207 fjz1 = _mm256_setzero_pd();
1208 fjx2 = _mm256_setzero_pd();
1209 fjy2 = _mm256_setzero_pd();
1210 fjz2 = _mm256_setzero_pd();
1212 /**************************
1213 * CALCULATE INTERACTIONS *
1214 **************************/
1216 /* COULOMB ELECTROSTATICS */
1217 velec = _mm256_mul_pd(qq00,rinv00);
1218 felec = _mm256_mul_pd(velec,rinvsq00);
1220 /* LENNARD-JONES DISPERSION/REPULSION */
1222 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1223 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1225 fscal = _mm256_add_pd(felec,fvdw);
1227 /* Calculate temporary vectorial force */
1228 tx = _mm256_mul_pd(fscal,dx00);
1229 ty = _mm256_mul_pd(fscal,dy00);
1230 tz = _mm256_mul_pd(fscal,dz00);
1232 /* Update vectorial force */
1233 fix0 = _mm256_add_pd(fix0,tx);
1234 fiy0 = _mm256_add_pd(fiy0,ty);
1235 fiz0 = _mm256_add_pd(fiz0,tz);
1237 fjx0 = _mm256_add_pd(fjx0,tx);
1238 fjy0 = _mm256_add_pd(fjy0,ty);
1239 fjz0 = _mm256_add_pd(fjz0,tz);
1241 /**************************
1242 * CALCULATE INTERACTIONS *
1243 **************************/
1245 /* COULOMB ELECTROSTATICS */
1246 velec = _mm256_mul_pd(qq01,rinv01);
1247 felec = _mm256_mul_pd(velec,rinvsq01);
1251 /* Calculate temporary vectorial force */
1252 tx = _mm256_mul_pd(fscal,dx01);
1253 ty = _mm256_mul_pd(fscal,dy01);
1254 tz = _mm256_mul_pd(fscal,dz01);
1256 /* Update vectorial force */
1257 fix0 = _mm256_add_pd(fix0,tx);
1258 fiy0 = _mm256_add_pd(fiy0,ty);
1259 fiz0 = _mm256_add_pd(fiz0,tz);
1261 fjx1 = _mm256_add_pd(fjx1,tx);
1262 fjy1 = _mm256_add_pd(fjy1,ty);
1263 fjz1 = _mm256_add_pd(fjz1,tz);
1265 /**************************
1266 * CALCULATE INTERACTIONS *
1267 **************************/
1269 /* COULOMB ELECTROSTATICS */
1270 velec = _mm256_mul_pd(qq02,rinv02);
1271 felec = _mm256_mul_pd(velec,rinvsq02);
1275 /* Calculate temporary vectorial force */
1276 tx = _mm256_mul_pd(fscal,dx02);
1277 ty = _mm256_mul_pd(fscal,dy02);
1278 tz = _mm256_mul_pd(fscal,dz02);
1280 /* Update vectorial force */
1281 fix0 = _mm256_add_pd(fix0,tx);
1282 fiy0 = _mm256_add_pd(fiy0,ty);
1283 fiz0 = _mm256_add_pd(fiz0,tz);
1285 fjx2 = _mm256_add_pd(fjx2,tx);
1286 fjy2 = _mm256_add_pd(fjy2,ty);
1287 fjz2 = _mm256_add_pd(fjz2,tz);
1289 /**************************
1290 * CALCULATE INTERACTIONS *
1291 **************************/
1293 /* COULOMB ELECTROSTATICS */
1294 velec = _mm256_mul_pd(qq10,rinv10);
1295 felec = _mm256_mul_pd(velec,rinvsq10);
1299 /* Calculate temporary vectorial force */
1300 tx = _mm256_mul_pd(fscal,dx10);
1301 ty = _mm256_mul_pd(fscal,dy10);
1302 tz = _mm256_mul_pd(fscal,dz10);
1304 /* Update vectorial force */
1305 fix1 = _mm256_add_pd(fix1,tx);
1306 fiy1 = _mm256_add_pd(fiy1,ty);
1307 fiz1 = _mm256_add_pd(fiz1,tz);
1309 fjx0 = _mm256_add_pd(fjx0,tx);
1310 fjy0 = _mm256_add_pd(fjy0,ty);
1311 fjz0 = _mm256_add_pd(fjz0,tz);
1313 /**************************
1314 * CALCULATE INTERACTIONS *
1315 **************************/
1317 /* COULOMB ELECTROSTATICS */
1318 velec = _mm256_mul_pd(qq11,rinv11);
1319 felec = _mm256_mul_pd(velec,rinvsq11);
1323 /* Calculate temporary vectorial force */
1324 tx = _mm256_mul_pd(fscal,dx11);
1325 ty = _mm256_mul_pd(fscal,dy11);
1326 tz = _mm256_mul_pd(fscal,dz11);
1328 /* Update vectorial force */
1329 fix1 = _mm256_add_pd(fix1,tx);
1330 fiy1 = _mm256_add_pd(fiy1,ty);
1331 fiz1 = _mm256_add_pd(fiz1,tz);
1333 fjx1 = _mm256_add_pd(fjx1,tx);
1334 fjy1 = _mm256_add_pd(fjy1,ty);
1335 fjz1 = _mm256_add_pd(fjz1,tz);
1337 /**************************
1338 * CALCULATE INTERACTIONS *
1339 **************************/
1341 /* COULOMB ELECTROSTATICS */
1342 velec = _mm256_mul_pd(qq12,rinv12);
1343 felec = _mm256_mul_pd(velec,rinvsq12);
1347 /* Calculate temporary vectorial force */
1348 tx = _mm256_mul_pd(fscal,dx12);
1349 ty = _mm256_mul_pd(fscal,dy12);
1350 tz = _mm256_mul_pd(fscal,dz12);
1352 /* Update vectorial force */
1353 fix1 = _mm256_add_pd(fix1,tx);
1354 fiy1 = _mm256_add_pd(fiy1,ty);
1355 fiz1 = _mm256_add_pd(fiz1,tz);
1357 fjx2 = _mm256_add_pd(fjx2,tx);
1358 fjy2 = _mm256_add_pd(fjy2,ty);
1359 fjz2 = _mm256_add_pd(fjz2,tz);
1361 /**************************
1362 * CALCULATE INTERACTIONS *
1363 **************************/
1365 /* COULOMB ELECTROSTATICS */
1366 velec = _mm256_mul_pd(qq20,rinv20);
1367 felec = _mm256_mul_pd(velec,rinvsq20);
1371 /* Calculate temporary vectorial force */
1372 tx = _mm256_mul_pd(fscal,dx20);
1373 ty = _mm256_mul_pd(fscal,dy20);
1374 tz = _mm256_mul_pd(fscal,dz20);
1376 /* Update vectorial force */
1377 fix2 = _mm256_add_pd(fix2,tx);
1378 fiy2 = _mm256_add_pd(fiy2,ty);
1379 fiz2 = _mm256_add_pd(fiz2,tz);
1381 fjx0 = _mm256_add_pd(fjx0,tx);
1382 fjy0 = _mm256_add_pd(fjy0,ty);
1383 fjz0 = _mm256_add_pd(fjz0,tz);
1385 /**************************
1386 * CALCULATE INTERACTIONS *
1387 **************************/
1389 /* COULOMB ELECTROSTATICS */
1390 velec = _mm256_mul_pd(qq21,rinv21);
1391 felec = _mm256_mul_pd(velec,rinvsq21);
1395 /* Calculate temporary vectorial force */
1396 tx = _mm256_mul_pd(fscal,dx21);
1397 ty = _mm256_mul_pd(fscal,dy21);
1398 tz = _mm256_mul_pd(fscal,dz21);
1400 /* Update vectorial force */
1401 fix2 = _mm256_add_pd(fix2,tx);
1402 fiy2 = _mm256_add_pd(fiy2,ty);
1403 fiz2 = _mm256_add_pd(fiz2,tz);
1405 fjx1 = _mm256_add_pd(fjx1,tx);
1406 fjy1 = _mm256_add_pd(fjy1,ty);
1407 fjz1 = _mm256_add_pd(fjz1,tz);
1409 /**************************
1410 * CALCULATE INTERACTIONS *
1411 **************************/
1413 /* COULOMB ELECTROSTATICS */
1414 velec = _mm256_mul_pd(qq22,rinv22);
1415 felec = _mm256_mul_pd(velec,rinvsq22);
1419 /* Calculate temporary vectorial force */
1420 tx = _mm256_mul_pd(fscal,dx22);
1421 ty = _mm256_mul_pd(fscal,dy22);
1422 tz = _mm256_mul_pd(fscal,dz22);
1424 /* Update vectorial force */
1425 fix2 = _mm256_add_pd(fix2,tx);
1426 fiy2 = _mm256_add_pd(fiy2,ty);
1427 fiz2 = _mm256_add_pd(fiz2,tz);
1429 fjx2 = _mm256_add_pd(fjx2,tx);
1430 fjy2 = _mm256_add_pd(fjy2,ty);
1431 fjz2 = _mm256_add_pd(fjz2,tz);
1433 fjptrA = f+j_coord_offsetA;
1434 fjptrB = f+j_coord_offsetB;
1435 fjptrC = f+j_coord_offsetC;
1436 fjptrD = f+j_coord_offsetD;
1438 gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1439 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1441 /* Inner loop uses 241 flops */
1444 if(jidx<j_index_end)
1447 /* Get j neighbor index, and coordinate index */
1448 jnrlistA = jjnr[jidx];
1449 jnrlistB = jjnr[jidx+1];
1450 jnrlistC = jjnr[jidx+2];
1451 jnrlistD = jjnr[jidx+3];
1452 /* Sign of each element will be negative for non-real atoms.
1453 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1454 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1456 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1458 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1459 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1460 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1462 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
1463 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
1464 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
1465 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
1466 j_coord_offsetA = DIM*jnrA;
1467 j_coord_offsetB = DIM*jnrB;
1468 j_coord_offsetC = DIM*jnrC;
1469 j_coord_offsetD = DIM*jnrD;
1471 /* load j atom coordinates */
1472 gmx_mm256_load_3rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1473 x+j_coord_offsetC,x+j_coord_offsetD,
1474 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,&jy2,&jz2);
1476 /* Calculate displacement vector */
1477 dx00 = _mm256_sub_pd(ix0,jx0);
1478 dy00 = _mm256_sub_pd(iy0,jy0);
1479 dz00 = _mm256_sub_pd(iz0,jz0);
1480 dx01 = _mm256_sub_pd(ix0,jx1);
1481 dy01 = _mm256_sub_pd(iy0,jy1);
1482 dz01 = _mm256_sub_pd(iz0,jz1);
1483 dx02 = _mm256_sub_pd(ix0,jx2);
1484 dy02 = _mm256_sub_pd(iy0,jy2);
1485 dz02 = _mm256_sub_pd(iz0,jz2);
1486 dx10 = _mm256_sub_pd(ix1,jx0);
1487 dy10 = _mm256_sub_pd(iy1,jy0);
1488 dz10 = _mm256_sub_pd(iz1,jz0);
1489 dx11 = _mm256_sub_pd(ix1,jx1);
1490 dy11 = _mm256_sub_pd(iy1,jy1);
1491 dz11 = _mm256_sub_pd(iz1,jz1);
1492 dx12 = _mm256_sub_pd(ix1,jx2);
1493 dy12 = _mm256_sub_pd(iy1,jy2);
1494 dz12 = _mm256_sub_pd(iz1,jz2);
1495 dx20 = _mm256_sub_pd(ix2,jx0);
1496 dy20 = _mm256_sub_pd(iy2,jy0);
1497 dz20 = _mm256_sub_pd(iz2,jz0);
1498 dx21 = _mm256_sub_pd(ix2,jx1);
1499 dy21 = _mm256_sub_pd(iy2,jy1);
1500 dz21 = _mm256_sub_pd(iz2,jz1);
1501 dx22 = _mm256_sub_pd(ix2,jx2);
1502 dy22 = _mm256_sub_pd(iy2,jy2);
1503 dz22 = _mm256_sub_pd(iz2,jz2);
1505 /* Calculate squared distance and things based on it */
1506 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1507 rsq01 = gmx_mm256_calc_rsq_pd(dx01,dy01,dz01);
1508 rsq02 = gmx_mm256_calc_rsq_pd(dx02,dy02,dz02);
1509 rsq10 = gmx_mm256_calc_rsq_pd(dx10,dy10,dz10);
1510 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1511 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1512 rsq20 = gmx_mm256_calc_rsq_pd(dx20,dy20,dz20);
1513 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1514 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1516 rinv00 = avx256_invsqrt_d(rsq00);
1517 rinv01 = avx256_invsqrt_d(rsq01);
1518 rinv02 = avx256_invsqrt_d(rsq02);
1519 rinv10 = avx256_invsqrt_d(rsq10);
1520 rinv11 = avx256_invsqrt_d(rsq11);
1521 rinv12 = avx256_invsqrt_d(rsq12);
1522 rinv20 = avx256_invsqrt_d(rsq20);
1523 rinv21 = avx256_invsqrt_d(rsq21);
1524 rinv22 = avx256_invsqrt_d(rsq22);
1526 rinvsq00 = _mm256_mul_pd(rinv00,rinv00);
1527 rinvsq01 = _mm256_mul_pd(rinv01,rinv01);
1528 rinvsq02 = _mm256_mul_pd(rinv02,rinv02);
1529 rinvsq10 = _mm256_mul_pd(rinv10,rinv10);
1530 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
1531 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
1532 rinvsq20 = _mm256_mul_pd(rinv20,rinv20);
1533 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
1534 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
1536 fjx0 = _mm256_setzero_pd();
1537 fjy0 = _mm256_setzero_pd();
1538 fjz0 = _mm256_setzero_pd();
1539 fjx1 = _mm256_setzero_pd();
1540 fjy1 = _mm256_setzero_pd();
1541 fjz1 = _mm256_setzero_pd();
1542 fjx2 = _mm256_setzero_pd();
1543 fjy2 = _mm256_setzero_pd();
1544 fjz2 = _mm256_setzero_pd();
1546 /**************************
1547 * CALCULATE INTERACTIONS *
1548 **************************/
1550 /* COULOMB ELECTROSTATICS */
1551 velec = _mm256_mul_pd(qq00,rinv00);
1552 felec = _mm256_mul_pd(velec,rinvsq00);
1554 /* LENNARD-JONES DISPERSION/REPULSION */
1556 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1557 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1559 fscal = _mm256_add_pd(felec,fvdw);
1561 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1563 /* Calculate temporary vectorial force */
1564 tx = _mm256_mul_pd(fscal,dx00);
1565 ty = _mm256_mul_pd(fscal,dy00);
1566 tz = _mm256_mul_pd(fscal,dz00);
1568 /* Update vectorial force */
1569 fix0 = _mm256_add_pd(fix0,tx);
1570 fiy0 = _mm256_add_pd(fiy0,ty);
1571 fiz0 = _mm256_add_pd(fiz0,tz);
1573 fjx0 = _mm256_add_pd(fjx0,tx);
1574 fjy0 = _mm256_add_pd(fjy0,ty);
1575 fjz0 = _mm256_add_pd(fjz0,tz);
1577 /**************************
1578 * CALCULATE INTERACTIONS *
1579 **************************/
1581 /* COULOMB ELECTROSTATICS */
1582 velec = _mm256_mul_pd(qq01,rinv01);
1583 felec = _mm256_mul_pd(velec,rinvsq01);
1587 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1589 /* Calculate temporary vectorial force */
1590 tx = _mm256_mul_pd(fscal,dx01);
1591 ty = _mm256_mul_pd(fscal,dy01);
1592 tz = _mm256_mul_pd(fscal,dz01);
1594 /* Update vectorial force */
1595 fix0 = _mm256_add_pd(fix0,tx);
1596 fiy0 = _mm256_add_pd(fiy0,ty);
1597 fiz0 = _mm256_add_pd(fiz0,tz);
1599 fjx1 = _mm256_add_pd(fjx1,tx);
1600 fjy1 = _mm256_add_pd(fjy1,ty);
1601 fjz1 = _mm256_add_pd(fjz1,tz);
1603 /**************************
1604 * CALCULATE INTERACTIONS *
1605 **************************/
1607 /* COULOMB ELECTROSTATICS */
1608 velec = _mm256_mul_pd(qq02,rinv02);
1609 felec = _mm256_mul_pd(velec,rinvsq02);
1613 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1615 /* Calculate temporary vectorial force */
1616 tx = _mm256_mul_pd(fscal,dx02);
1617 ty = _mm256_mul_pd(fscal,dy02);
1618 tz = _mm256_mul_pd(fscal,dz02);
1620 /* Update vectorial force */
1621 fix0 = _mm256_add_pd(fix0,tx);
1622 fiy0 = _mm256_add_pd(fiy0,ty);
1623 fiz0 = _mm256_add_pd(fiz0,tz);
1625 fjx2 = _mm256_add_pd(fjx2,tx);
1626 fjy2 = _mm256_add_pd(fjy2,ty);
1627 fjz2 = _mm256_add_pd(fjz2,tz);
1629 /**************************
1630 * CALCULATE INTERACTIONS *
1631 **************************/
1633 /* COULOMB ELECTROSTATICS */
1634 velec = _mm256_mul_pd(qq10,rinv10);
1635 felec = _mm256_mul_pd(velec,rinvsq10);
1639 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1641 /* Calculate temporary vectorial force */
1642 tx = _mm256_mul_pd(fscal,dx10);
1643 ty = _mm256_mul_pd(fscal,dy10);
1644 tz = _mm256_mul_pd(fscal,dz10);
1646 /* Update vectorial force */
1647 fix1 = _mm256_add_pd(fix1,tx);
1648 fiy1 = _mm256_add_pd(fiy1,ty);
1649 fiz1 = _mm256_add_pd(fiz1,tz);
1651 fjx0 = _mm256_add_pd(fjx0,tx);
1652 fjy0 = _mm256_add_pd(fjy0,ty);
1653 fjz0 = _mm256_add_pd(fjz0,tz);
1655 /**************************
1656 * CALCULATE INTERACTIONS *
1657 **************************/
1659 /* COULOMB ELECTROSTATICS */
1660 velec = _mm256_mul_pd(qq11,rinv11);
1661 felec = _mm256_mul_pd(velec,rinvsq11);
1665 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1667 /* Calculate temporary vectorial force */
1668 tx = _mm256_mul_pd(fscal,dx11);
1669 ty = _mm256_mul_pd(fscal,dy11);
1670 tz = _mm256_mul_pd(fscal,dz11);
1672 /* Update vectorial force */
1673 fix1 = _mm256_add_pd(fix1,tx);
1674 fiy1 = _mm256_add_pd(fiy1,ty);
1675 fiz1 = _mm256_add_pd(fiz1,tz);
1677 fjx1 = _mm256_add_pd(fjx1,tx);
1678 fjy1 = _mm256_add_pd(fjy1,ty);
1679 fjz1 = _mm256_add_pd(fjz1,tz);
1681 /**************************
1682 * CALCULATE INTERACTIONS *
1683 **************************/
1685 /* COULOMB ELECTROSTATICS */
1686 velec = _mm256_mul_pd(qq12,rinv12);
1687 felec = _mm256_mul_pd(velec,rinvsq12);
1691 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1693 /* Calculate temporary vectorial force */
1694 tx = _mm256_mul_pd(fscal,dx12);
1695 ty = _mm256_mul_pd(fscal,dy12);
1696 tz = _mm256_mul_pd(fscal,dz12);
1698 /* Update vectorial force */
1699 fix1 = _mm256_add_pd(fix1,tx);
1700 fiy1 = _mm256_add_pd(fiy1,ty);
1701 fiz1 = _mm256_add_pd(fiz1,tz);
1703 fjx2 = _mm256_add_pd(fjx2,tx);
1704 fjy2 = _mm256_add_pd(fjy2,ty);
1705 fjz2 = _mm256_add_pd(fjz2,tz);
1707 /**************************
1708 * CALCULATE INTERACTIONS *
1709 **************************/
1711 /* COULOMB ELECTROSTATICS */
1712 velec = _mm256_mul_pd(qq20,rinv20);
1713 felec = _mm256_mul_pd(velec,rinvsq20);
1717 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1719 /* Calculate temporary vectorial force */
1720 tx = _mm256_mul_pd(fscal,dx20);
1721 ty = _mm256_mul_pd(fscal,dy20);
1722 tz = _mm256_mul_pd(fscal,dz20);
1724 /* Update vectorial force */
1725 fix2 = _mm256_add_pd(fix2,tx);
1726 fiy2 = _mm256_add_pd(fiy2,ty);
1727 fiz2 = _mm256_add_pd(fiz2,tz);
1729 fjx0 = _mm256_add_pd(fjx0,tx);
1730 fjy0 = _mm256_add_pd(fjy0,ty);
1731 fjz0 = _mm256_add_pd(fjz0,tz);
1733 /**************************
1734 * CALCULATE INTERACTIONS *
1735 **************************/
1737 /* COULOMB ELECTROSTATICS */
1738 velec = _mm256_mul_pd(qq21,rinv21);
1739 felec = _mm256_mul_pd(velec,rinvsq21);
1743 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1745 /* Calculate temporary vectorial force */
1746 tx = _mm256_mul_pd(fscal,dx21);
1747 ty = _mm256_mul_pd(fscal,dy21);
1748 tz = _mm256_mul_pd(fscal,dz21);
1750 /* Update vectorial force */
1751 fix2 = _mm256_add_pd(fix2,tx);
1752 fiy2 = _mm256_add_pd(fiy2,ty);
1753 fiz2 = _mm256_add_pd(fiz2,tz);
1755 fjx1 = _mm256_add_pd(fjx1,tx);
1756 fjy1 = _mm256_add_pd(fjy1,ty);
1757 fjz1 = _mm256_add_pd(fjz1,tz);
1759 /**************************
1760 * CALCULATE INTERACTIONS *
1761 **************************/
1763 /* COULOMB ELECTROSTATICS */
1764 velec = _mm256_mul_pd(qq22,rinv22);
1765 felec = _mm256_mul_pd(velec,rinvsq22);
1769 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1771 /* Calculate temporary vectorial force */
1772 tx = _mm256_mul_pd(fscal,dx22);
1773 ty = _mm256_mul_pd(fscal,dy22);
1774 tz = _mm256_mul_pd(fscal,dz22);
1776 /* Update vectorial force */
1777 fix2 = _mm256_add_pd(fix2,tx);
1778 fiy2 = _mm256_add_pd(fiy2,ty);
1779 fiz2 = _mm256_add_pd(fiz2,tz);
1781 fjx2 = _mm256_add_pd(fjx2,tx);
1782 fjy2 = _mm256_add_pd(fjy2,ty);
1783 fjz2 = _mm256_add_pd(fjz2,tz);
1785 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1786 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1787 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1788 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1790 gmx_mm256_decrement_3rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1791 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2);
1793 /* Inner loop uses 241 flops */
1796 /* End of innermost loop */
1798 gmx_mm256_update_iforce_3atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,
1799 f+i_coord_offset,fshift+i_shift_offset);
1801 /* Increment number of inner iterations */
1802 inneriter += j_index_end - j_index_start;
1804 /* Outer loop uses 18 flops */
1807 /* Increment number of outer iterations */
1810 /* Update outer/inner flops */
1812 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3W3_F,outeriter*18 + inneriter*241);