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_single kernel generator.
44 #include "../nb_kernel.h"
45 #include "gromacs/gmxlib/nrnb.h"
47 #include "kernelutil_x86_avx_256_single.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJEw_GeomW4W4_VF_avx_256_single
51 * Electrostatics interaction: Ewald
52 * VdW interaction: LJEwald
53 * Geometry: Water4-Water4
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecEw_VdwLJEw_GeomW4W4_VF_avx_256_single
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,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight 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 jnrE,jnrF,jnrG,jnrH;
75 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
76 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
77 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
78 int j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
79 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
81 real *shiftvec,*fshift,*x,*f;
82 real *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
84 __m256 tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
85 real * vdwioffsetptr0;
86 real * vdwgridioffsetptr0;
87 __m256 ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
88 real * vdwioffsetptr1;
89 real * vdwgridioffsetptr1;
90 __m256 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
91 real * vdwioffsetptr2;
92 real * vdwgridioffsetptr2;
93 __m256 ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
94 real * vdwioffsetptr3;
95 real * vdwgridioffsetptr3;
96 __m256 ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
97 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
98 __m256 jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
99 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D,vdwjidx1E,vdwjidx1F,vdwjidx1G,vdwjidx1H;
100 __m256 jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
101 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D,vdwjidx2E,vdwjidx2F,vdwjidx2G,vdwjidx2H;
102 __m256 jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
103 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D,vdwjidx3E,vdwjidx3F,vdwjidx3G,vdwjidx3H;
104 __m256 jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
105 __m256 dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
106 __m256 dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
107 __m256 dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
108 __m256 dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
109 __m256 dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
110 __m256 dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
111 __m256 dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
112 __m256 dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
113 __m256 dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
114 __m256 dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
115 __m256 velec,felec,velecsum,facel,crf,krf,krf2;
118 __m256 rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
121 __m256 one_sixth = _mm256_set1_ps(1.0/6.0);
122 __m256 one_twelfth = _mm256_set1_ps(1.0/12.0);
134 __m256 ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
135 __m256 one_half = _mm256_set1_ps(0.5);
136 __m256 minus_one = _mm256_set1_ps(-1.0);
138 __m128i ewitab_lo,ewitab_hi;
139 __m256 ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
140 __m256 beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
142 __m256 dummy_mask,cutoff_mask;
143 __m256 signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
144 __m256 one = _mm256_set1_ps(1.0);
145 __m256 two = _mm256_set1_ps(2.0);
151 jindex = nlist->jindex;
153 shiftidx = nlist->shift;
155 shiftvec = fr->shift_vec[0];
156 fshift = fr->fshift[0];
157 facel = _mm256_set1_ps(fr->ic->epsfac);
158 charge = mdatoms->chargeA;
159 nvdwtype = fr->ntype;
161 vdwtype = mdatoms->typeA;
162 vdwgridparam = fr->ljpme_c6grid;
163 sh_lj_ewald = _mm256_set1_ps(fr->ic->sh_lj_ewald);
164 ewclj = _mm256_set1_ps(fr->ic->ewaldcoeff_lj);
165 ewclj2 = _mm256_mul_ps(minus_one,_mm256_mul_ps(ewclj,ewclj));
167 sh_ewald = _mm256_set1_ps(fr->ic->sh_ewald);
168 beta = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
169 beta2 = _mm256_mul_ps(beta,beta);
170 beta3 = _mm256_mul_ps(beta,beta2);
172 ewtab = fr->ic->tabq_coul_FDV0;
173 ewtabscale = _mm256_set1_ps(fr->ic->tabq_scale);
174 ewtabhalfspace = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
176 /* Setup water-specific parameters */
177 inr = nlist->iinr[0];
178 iq1 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
179 iq2 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
180 iq3 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+3]));
181 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
182 vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
184 jq1 = _mm256_set1_ps(charge[inr+1]);
185 jq2 = _mm256_set1_ps(charge[inr+2]);
186 jq3 = _mm256_set1_ps(charge[inr+3]);
187 vdwjidx0A = 2*vdwtype[inr+0];
188 c6_00 = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A]);
189 c12_00 = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A+1]);
190 c6grid_00 = _mm256_set1_ps(vdwgridioffsetptr0[vdwjidx0A]);
191 qq11 = _mm256_mul_ps(iq1,jq1);
192 qq12 = _mm256_mul_ps(iq1,jq2);
193 qq13 = _mm256_mul_ps(iq1,jq3);
194 qq21 = _mm256_mul_ps(iq2,jq1);
195 qq22 = _mm256_mul_ps(iq2,jq2);
196 qq23 = _mm256_mul_ps(iq2,jq3);
197 qq31 = _mm256_mul_ps(iq3,jq1);
198 qq32 = _mm256_mul_ps(iq3,jq2);
199 qq33 = _mm256_mul_ps(iq3,jq3);
201 /* Avoid stupid compiler warnings */
202 jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
215 for(iidx=0;iidx<4*DIM;iidx++)
220 /* Start outer loop over neighborlists */
221 for(iidx=0; iidx<nri; iidx++)
223 /* Load shift vector for this list */
224 i_shift_offset = DIM*shiftidx[iidx];
226 /* Load limits for loop over neighbors */
227 j_index_start = jindex[iidx];
228 j_index_end = jindex[iidx+1];
230 /* Get outer coordinate index */
232 i_coord_offset = DIM*inr;
234 /* Load i particle coords and add shift vector */
235 gmx_mm256_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
236 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
238 fix0 = _mm256_setzero_ps();
239 fiy0 = _mm256_setzero_ps();
240 fiz0 = _mm256_setzero_ps();
241 fix1 = _mm256_setzero_ps();
242 fiy1 = _mm256_setzero_ps();
243 fiz1 = _mm256_setzero_ps();
244 fix2 = _mm256_setzero_ps();
245 fiy2 = _mm256_setzero_ps();
246 fiz2 = _mm256_setzero_ps();
247 fix3 = _mm256_setzero_ps();
248 fiy3 = _mm256_setzero_ps();
249 fiz3 = _mm256_setzero_ps();
251 /* Reset potential sums */
252 velecsum = _mm256_setzero_ps();
253 vvdwsum = _mm256_setzero_ps();
255 /* Start inner kernel loop */
256 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
259 /* Get j neighbor index, and coordinate index */
268 j_coord_offsetA = DIM*jnrA;
269 j_coord_offsetB = DIM*jnrB;
270 j_coord_offsetC = DIM*jnrC;
271 j_coord_offsetD = DIM*jnrD;
272 j_coord_offsetE = DIM*jnrE;
273 j_coord_offsetF = DIM*jnrF;
274 j_coord_offsetG = DIM*jnrG;
275 j_coord_offsetH = DIM*jnrH;
277 /* load j atom coordinates */
278 gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
279 x+j_coord_offsetC,x+j_coord_offsetD,
280 x+j_coord_offsetE,x+j_coord_offsetF,
281 x+j_coord_offsetG,x+j_coord_offsetH,
282 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
283 &jy2,&jz2,&jx3,&jy3,&jz3);
285 /* Calculate displacement vector */
286 dx00 = _mm256_sub_ps(ix0,jx0);
287 dy00 = _mm256_sub_ps(iy0,jy0);
288 dz00 = _mm256_sub_ps(iz0,jz0);
289 dx11 = _mm256_sub_ps(ix1,jx1);
290 dy11 = _mm256_sub_ps(iy1,jy1);
291 dz11 = _mm256_sub_ps(iz1,jz1);
292 dx12 = _mm256_sub_ps(ix1,jx2);
293 dy12 = _mm256_sub_ps(iy1,jy2);
294 dz12 = _mm256_sub_ps(iz1,jz2);
295 dx13 = _mm256_sub_ps(ix1,jx3);
296 dy13 = _mm256_sub_ps(iy1,jy3);
297 dz13 = _mm256_sub_ps(iz1,jz3);
298 dx21 = _mm256_sub_ps(ix2,jx1);
299 dy21 = _mm256_sub_ps(iy2,jy1);
300 dz21 = _mm256_sub_ps(iz2,jz1);
301 dx22 = _mm256_sub_ps(ix2,jx2);
302 dy22 = _mm256_sub_ps(iy2,jy2);
303 dz22 = _mm256_sub_ps(iz2,jz2);
304 dx23 = _mm256_sub_ps(ix2,jx3);
305 dy23 = _mm256_sub_ps(iy2,jy3);
306 dz23 = _mm256_sub_ps(iz2,jz3);
307 dx31 = _mm256_sub_ps(ix3,jx1);
308 dy31 = _mm256_sub_ps(iy3,jy1);
309 dz31 = _mm256_sub_ps(iz3,jz1);
310 dx32 = _mm256_sub_ps(ix3,jx2);
311 dy32 = _mm256_sub_ps(iy3,jy2);
312 dz32 = _mm256_sub_ps(iz3,jz2);
313 dx33 = _mm256_sub_ps(ix3,jx3);
314 dy33 = _mm256_sub_ps(iy3,jy3);
315 dz33 = _mm256_sub_ps(iz3,jz3);
317 /* Calculate squared distance and things based on it */
318 rsq00 = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
319 rsq11 = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
320 rsq12 = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
321 rsq13 = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
322 rsq21 = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
323 rsq22 = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
324 rsq23 = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
325 rsq31 = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
326 rsq32 = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
327 rsq33 = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
329 rinv00 = avx256_invsqrt_f(rsq00);
330 rinv11 = avx256_invsqrt_f(rsq11);
331 rinv12 = avx256_invsqrt_f(rsq12);
332 rinv13 = avx256_invsqrt_f(rsq13);
333 rinv21 = avx256_invsqrt_f(rsq21);
334 rinv22 = avx256_invsqrt_f(rsq22);
335 rinv23 = avx256_invsqrt_f(rsq23);
336 rinv31 = avx256_invsqrt_f(rsq31);
337 rinv32 = avx256_invsqrt_f(rsq32);
338 rinv33 = avx256_invsqrt_f(rsq33);
340 rinvsq00 = _mm256_mul_ps(rinv00,rinv00);
341 rinvsq11 = _mm256_mul_ps(rinv11,rinv11);
342 rinvsq12 = _mm256_mul_ps(rinv12,rinv12);
343 rinvsq13 = _mm256_mul_ps(rinv13,rinv13);
344 rinvsq21 = _mm256_mul_ps(rinv21,rinv21);
345 rinvsq22 = _mm256_mul_ps(rinv22,rinv22);
346 rinvsq23 = _mm256_mul_ps(rinv23,rinv23);
347 rinvsq31 = _mm256_mul_ps(rinv31,rinv31);
348 rinvsq32 = _mm256_mul_ps(rinv32,rinv32);
349 rinvsq33 = _mm256_mul_ps(rinv33,rinv33);
351 fjx0 = _mm256_setzero_ps();
352 fjy0 = _mm256_setzero_ps();
353 fjz0 = _mm256_setzero_ps();
354 fjx1 = _mm256_setzero_ps();
355 fjy1 = _mm256_setzero_ps();
356 fjz1 = _mm256_setzero_ps();
357 fjx2 = _mm256_setzero_ps();
358 fjy2 = _mm256_setzero_ps();
359 fjz2 = _mm256_setzero_ps();
360 fjx3 = _mm256_setzero_ps();
361 fjy3 = _mm256_setzero_ps();
362 fjz3 = _mm256_setzero_ps();
364 /**************************
365 * CALCULATE INTERACTIONS *
366 **************************/
368 r00 = _mm256_mul_ps(rsq00,rinv00);
370 /* Analytical LJ-PME */
371 rinvsix = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
372 ewcljrsq = _mm256_mul_ps(ewclj2,rsq00);
373 ewclj6 = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
374 exponent = avx256_exp_f(ewcljrsq);
375 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
376 poly = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
377 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
378 vvdw6 = _mm256_mul_ps(_mm256_sub_ps(c6_00,_mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly))),rinvsix);
379 vvdw12 = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
380 vvdw = _mm256_sub_ps(_mm256_mul_ps(vvdw12,one_twelfth),_mm256_mul_ps(vvdw6,one_sixth));
381 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
382 fvdw = _mm256_mul_ps(_mm256_sub_ps(vvdw12,_mm256_sub_ps(vvdw6,_mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6)))),rinvsq00);
384 /* Update potential sum for this i atom from the interaction with this j atom. */
385 vvdwsum = _mm256_add_ps(vvdwsum,vvdw);
389 /* Calculate temporary vectorial force */
390 tx = _mm256_mul_ps(fscal,dx00);
391 ty = _mm256_mul_ps(fscal,dy00);
392 tz = _mm256_mul_ps(fscal,dz00);
394 /* Update vectorial force */
395 fix0 = _mm256_add_ps(fix0,tx);
396 fiy0 = _mm256_add_ps(fiy0,ty);
397 fiz0 = _mm256_add_ps(fiz0,tz);
399 fjx0 = _mm256_add_ps(fjx0,tx);
400 fjy0 = _mm256_add_ps(fjy0,ty);
401 fjz0 = _mm256_add_ps(fjz0,tz);
403 /**************************
404 * CALCULATE INTERACTIONS *
405 **************************/
407 r11 = _mm256_mul_ps(rsq11,rinv11);
409 /* EWALD ELECTROSTATICS */
411 /* Analytical PME correction */
412 zeta2 = _mm256_mul_ps(beta2,rsq11);
413 rinv3 = _mm256_mul_ps(rinvsq11,rinv11);
414 pmecorrF = avx256_pmecorrF_f(zeta2);
415 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
416 felec = _mm256_mul_ps(qq11,felec);
417 pmecorrV = avx256_pmecorrV_f(zeta2);
418 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
419 velec = _mm256_sub_ps(rinv11,pmecorrV);
420 velec = _mm256_mul_ps(qq11,velec);
422 /* Update potential sum for this i atom from the interaction with this j atom. */
423 velecsum = _mm256_add_ps(velecsum,velec);
427 /* Calculate temporary vectorial force */
428 tx = _mm256_mul_ps(fscal,dx11);
429 ty = _mm256_mul_ps(fscal,dy11);
430 tz = _mm256_mul_ps(fscal,dz11);
432 /* Update vectorial force */
433 fix1 = _mm256_add_ps(fix1,tx);
434 fiy1 = _mm256_add_ps(fiy1,ty);
435 fiz1 = _mm256_add_ps(fiz1,tz);
437 fjx1 = _mm256_add_ps(fjx1,tx);
438 fjy1 = _mm256_add_ps(fjy1,ty);
439 fjz1 = _mm256_add_ps(fjz1,tz);
441 /**************************
442 * CALCULATE INTERACTIONS *
443 **************************/
445 r12 = _mm256_mul_ps(rsq12,rinv12);
447 /* EWALD ELECTROSTATICS */
449 /* Analytical PME correction */
450 zeta2 = _mm256_mul_ps(beta2,rsq12);
451 rinv3 = _mm256_mul_ps(rinvsq12,rinv12);
452 pmecorrF = avx256_pmecorrF_f(zeta2);
453 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
454 felec = _mm256_mul_ps(qq12,felec);
455 pmecorrV = avx256_pmecorrV_f(zeta2);
456 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
457 velec = _mm256_sub_ps(rinv12,pmecorrV);
458 velec = _mm256_mul_ps(qq12,velec);
460 /* Update potential sum for this i atom from the interaction with this j atom. */
461 velecsum = _mm256_add_ps(velecsum,velec);
465 /* Calculate temporary vectorial force */
466 tx = _mm256_mul_ps(fscal,dx12);
467 ty = _mm256_mul_ps(fscal,dy12);
468 tz = _mm256_mul_ps(fscal,dz12);
470 /* Update vectorial force */
471 fix1 = _mm256_add_ps(fix1,tx);
472 fiy1 = _mm256_add_ps(fiy1,ty);
473 fiz1 = _mm256_add_ps(fiz1,tz);
475 fjx2 = _mm256_add_ps(fjx2,tx);
476 fjy2 = _mm256_add_ps(fjy2,ty);
477 fjz2 = _mm256_add_ps(fjz2,tz);
479 /**************************
480 * CALCULATE INTERACTIONS *
481 **************************/
483 r13 = _mm256_mul_ps(rsq13,rinv13);
485 /* EWALD ELECTROSTATICS */
487 /* Analytical PME correction */
488 zeta2 = _mm256_mul_ps(beta2,rsq13);
489 rinv3 = _mm256_mul_ps(rinvsq13,rinv13);
490 pmecorrF = avx256_pmecorrF_f(zeta2);
491 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
492 felec = _mm256_mul_ps(qq13,felec);
493 pmecorrV = avx256_pmecorrV_f(zeta2);
494 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
495 velec = _mm256_sub_ps(rinv13,pmecorrV);
496 velec = _mm256_mul_ps(qq13,velec);
498 /* Update potential sum for this i atom from the interaction with this j atom. */
499 velecsum = _mm256_add_ps(velecsum,velec);
503 /* Calculate temporary vectorial force */
504 tx = _mm256_mul_ps(fscal,dx13);
505 ty = _mm256_mul_ps(fscal,dy13);
506 tz = _mm256_mul_ps(fscal,dz13);
508 /* Update vectorial force */
509 fix1 = _mm256_add_ps(fix1,tx);
510 fiy1 = _mm256_add_ps(fiy1,ty);
511 fiz1 = _mm256_add_ps(fiz1,tz);
513 fjx3 = _mm256_add_ps(fjx3,tx);
514 fjy3 = _mm256_add_ps(fjy3,ty);
515 fjz3 = _mm256_add_ps(fjz3,tz);
517 /**************************
518 * CALCULATE INTERACTIONS *
519 **************************/
521 r21 = _mm256_mul_ps(rsq21,rinv21);
523 /* EWALD ELECTROSTATICS */
525 /* Analytical PME correction */
526 zeta2 = _mm256_mul_ps(beta2,rsq21);
527 rinv3 = _mm256_mul_ps(rinvsq21,rinv21);
528 pmecorrF = avx256_pmecorrF_f(zeta2);
529 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
530 felec = _mm256_mul_ps(qq21,felec);
531 pmecorrV = avx256_pmecorrV_f(zeta2);
532 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
533 velec = _mm256_sub_ps(rinv21,pmecorrV);
534 velec = _mm256_mul_ps(qq21,velec);
536 /* Update potential sum for this i atom from the interaction with this j atom. */
537 velecsum = _mm256_add_ps(velecsum,velec);
541 /* Calculate temporary vectorial force */
542 tx = _mm256_mul_ps(fscal,dx21);
543 ty = _mm256_mul_ps(fscal,dy21);
544 tz = _mm256_mul_ps(fscal,dz21);
546 /* Update vectorial force */
547 fix2 = _mm256_add_ps(fix2,tx);
548 fiy2 = _mm256_add_ps(fiy2,ty);
549 fiz2 = _mm256_add_ps(fiz2,tz);
551 fjx1 = _mm256_add_ps(fjx1,tx);
552 fjy1 = _mm256_add_ps(fjy1,ty);
553 fjz1 = _mm256_add_ps(fjz1,tz);
555 /**************************
556 * CALCULATE INTERACTIONS *
557 **************************/
559 r22 = _mm256_mul_ps(rsq22,rinv22);
561 /* EWALD ELECTROSTATICS */
563 /* Analytical PME correction */
564 zeta2 = _mm256_mul_ps(beta2,rsq22);
565 rinv3 = _mm256_mul_ps(rinvsq22,rinv22);
566 pmecorrF = avx256_pmecorrF_f(zeta2);
567 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
568 felec = _mm256_mul_ps(qq22,felec);
569 pmecorrV = avx256_pmecorrV_f(zeta2);
570 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
571 velec = _mm256_sub_ps(rinv22,pmecorrV);
572 velec = _mm256_mul_ps(qq22,velec);
574 /* Update potential sum for this i atom from the interaction with this j atom. */
575 velecsum = _mm256_add_ps(velecsum,velec);
579 /* Calculate temporary vectorial force */
580 tx = _mm256_mul_ps(fscal,dx22);
581 ty = _mm256_mul_ps(fscal,dy22);
582 tz = _mm256_mul_ps(fscal,dz22);
584 /* Update vectorial force */
585 fix2 = _mm256_add_ps(fix2,tx);
586 fiy2 = _mm256_add_ps(fiy2,ty);
587 fiz2 = _mm256_add_ps(fiz2,tz);
589 fjx2 = _mm256_add_ps(fjx2,tx);
590 fjy2 = _mm256_add_ps(fjy2,ty);
591 fjz2 = _mm256_add_ps(fjz2,tz);
593 /**************************
594 * CALCULATE INTERACTIONS *
595 **************************/
597 r23 = _mm256_mul_ps(rsq23,rinv23);
599 /* EWALD ELECTROSTATICS */
601 /* Analytical PME correction */
602 zeta2 = _mm256_mul_ps(beta2,rsq23);
603 rinv3 = _mm256_mul_ps(rinvsq23,rinv23);
604 pmecorrF = avx256_pmecorrF_f(zeta2);
605 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
606 felec = _mm256_mul_ps(qq23,felec);
607 pmecorrV = avx256_pmecorrV_f(zeta2);
608 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
609 velec = _mm256_sub_ps(rinv23,pmecorrV);
610 velec = _mm256_mul_ps(qq23,velec);
612 /* Update potential sum for this i atom from the interaction with this j atom. */
613 velecsum = _mm256_add_ps(velecsum,velec);
617 /* Calculate temporary vectorial force */
618 tx = _mm256_mul_ps(fscal,dx23);
619 ty = _mm256_mul_ps(fscal,dy23);
620 tz = _mm256_mul_ps(fscal,dz23);
622 /* Update vectorial force */
623 fix2 = _mm256_add_ps(fix2,tx);
624 fiy2 = _mm256_add_ps(fiy2,ty);
625 fiz2 = _mm256_add_ps(fiz2,tz);
627 fjx3 = _mm256_add_ps(fjx3,tx);
628 fjy3 = _mm256_add_ps(fjy3,ty);
629 fjz3 = _mm256_add_ps(fjz3,tz);
631 /**************************
632 * CALCULATE INTERACTIONS *
633 **************************/
635 r31 = _mm256_mul_ps(rsq31,rinv31);
637 /* EWALD ELECTROSTATICS */
639 /* Analytical PME correction */
640 zeta2 = _mm256_mul_ps(beta2,rsq31);
641 rinv3 = _mm256_mul_ps(rinvsq31,rinv31);
642 pmecorrF = avx256_pmecorrF_f(zeta2);
643 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
644 felec = _mm256_mul_ps(qq31,felec);
645 pmecorrV = avx256_pmecorrV_f(zeta2);
646 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
647 velec = _mm256_sub_ps(rinv31,pmecorrV);
648 velec = _mm256_mul_ps(qq31,velec);
650 /* Update potential sum for this i atom from the interaction with this j atom. */
651 velecsum = _mm256_add_ps(velecsum,velec);
655 /* Calculate temporary vectorial force */
656 tx = _mm256_mul_ps(fscal,dx31);
657 ty = _mm256_mul_ps(fscal,dy31);
658 tz = _mm256_mul_ps(fscal,dz31);
660 /* Update vectorial force */
661 fix3 = _mm256_add_ps(fix3,tx);
662 fiy3 = _mm256_add_ps(fiy3,ty);
663 fiz3 = _mm256_add_ps(fiz3,tz);
665 fjx1 = _mm256_add_ps(fjx1,tx);
666 fjy1 = _mm256_add_ps(fjy1,ty);
667 fjz1 = _mm256_add_ps(fjz1,tz);
669 /**************************
670 * CALCULATE INTERACTIONS *
671 **************************/
673 r32 = _mm256_mul_ps(rsq32,rinv32);
675 /* EWALD ELECTROSTATICS */
677 /* Analytical PME correction */
678 zeta2 = _mm256_mul_ps(beta2,rsq32);
679 rinv3 = _mm256_mul_ps(rinvsq32,rinv32);
680 pmecorrF = avx256_pmecorrF_f(zeta2);
681 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
682 felec = _mm256_mul_ps(qq32,felec);
683 pmecorrV = avx256_pmecorrV_f(zeta2);
684 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
685 velec = _mm256_sub_ps(rinv32,pmecorrV);
686 velec = _mm256_mul_ps(qq32,velec);
688 /* Update potential sum for this i atom from the interaction with this j atom. */
689 velecsum = _mm256_add_ps(velecsum,velec);
693 /* Calculate temporary vectorial force */
694 tx = _mm256_mul_ps(fscal,dx32);
695 ty = _mm256_mul_ps(fscal,dy32);
696 tz = _mm256_mul_ps(fscal,dz32);
698 /* Update vectorial force */
699 fix3 = _mm256_add_ps(fix3,tx);
700 fiy3 = _mm256_add_ps(fiy3,ty);
701 fiz3 = _mm256_add_ps(fiz3,tz);
703 fjx2 = _mm256_add_ps(fjx2,tx);
704 fjy2 = _mm256_add_ps(fjy2,ty);
705 fjz2 = _mm256_add_ps(fjz2,tz);
707 /**************************
708 * CALCULATE INTERACTIONS *
709 **************************/
711 r33 = _mm256_mul_ps(rsq33,rinv33);
713 /* EWALD ELECTROSTATICS */
715 /* Analytical PME correction */
716 zeta2 = _mm256_mul_ps(beta2,rsq33);
717 rinv3 = _mm256_mul_ps(rinvsq33,rinv33);
718 pmecorrF = avx256_pmecorrF_f(zeta2);
719 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
720 felec = _mm256_mul_ps(qq33,felec);
721 pmecorrV = avx256_pmecorrV_f(zeta2);
722 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
723 velec = _mm256_sub_ps(rinv33,pmecorrV);
724 velec = _mm256_mul_ps(qq33,velec);
726 /* Update potential sum for this i atom from the interaction with this j atom. */
727 velecsum = _mm256_add_ps(velecsum,velec);
731 /* Calculate temporary vectorial force */
732 tx = _mm256_mul_ps(fscal,dx33);
733 ty = _mm256_mul_ps(fscal,dy33);
734 tz = _mm256_mul_ps(fscal,dz33);
736 /* Update vectorial force */
737 fix3 = _mm256_add_ps(fix3,tx);
738 fiy3 = _mm256_add_ps(fiy3,ty);
739 fiz3 = _mm256_add_ps(fiz3,tz);
741 fjx3 = _mm256_add_ps(fjx3,tx);
742 fjy3 = _mm256_add_ps(fjy3,ty);
743 fjz3 = _mm256_add_ps(fjz3,tz);
745 fjptrA = f+j_coord_offsetA;
746 fjptrB = f+j_coord_offsetB;
747 fjptrC = f+j_coord_offsetC;
748 fjptrD = f+j_coord_offsetD;
749 fjptrE = f+j_coord_offsetE;
750 fjptrF = f+j_coord_offsetF;
751 fjptrG = f+j_coord_offsetG;
752 fjptrH = f+j_coord_offsetH;
754 gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
755 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
756 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
758 /* Inner loop uses 810 flops */
764 /* Get j neighbor index, and coordinate index */
765 jnrlistA = jjnr[jidx];
766 jnrlistB = jjnr[jidx+1];
767 jnrlistC = jjnr[jidx+2];
768 jnrlistD = jjnr[jidx+3];
769 jnrlistE = jjnr[jidx+4];
770 jnrlistF = jjnr[jidx+5];
771 jnrlistG = jjnr[jidx+6];
772 jnrlistH = jjnr[jidx+7];
773 /* Sign of each element will be negative for non-real atoms.
774 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
775 * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
777 dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
778 gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
780 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
781 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
782 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
783 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
784 jnrE = (jnrlistE>=0) ? jnrlistE : 0;
785 jnrF = (jnrlistF>=0) ? jnrlistF : 0;
786 jnrG = (jnrlistG>=0) ? jnrlistG : 0;
787 jnrH = (jnrlistH>=0) ? jnrlistH : 0;
788 j_coord_offsetA = DIM*jnrA;
789 j_coord_offsetB = DIM*jnrB;
790 j_coord_offsetC = DIM*jnrC;
791 j_coord_offsetD = DIM*jnrD;
792 j_coord_offsetE = DIM*jnrE;
793 j_coord_offsetF = DIM*jnrF;
794 j_coord_offsetG = DIM*jnrG;
795 j_coord_offsetH = DIM*jnrH;
797 /* load j atom coordinates */
798 gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
799 x+j_coord_offsetC,x+j_coord_offsetD,
800 x+j_coord_offsetE,x+j_coord_offsetF,
801 x+j_coord_offsetG,x+j_coord_offsetH,
802 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
803 &jy2,&jz2,&jx3,&jy3,&jz3);
805 /* Calculate displacement vector */
806 dx00 = _mm256_sub_ps(ix0,jx0);
807 dy00 = _mm256_sub_ps(iy0,jy0);
808 dz00 = _mm256_sub_ps(iz0,jz0);
809 dx11 = _mm256_sub_ps(ix1,jx1);
810 dy11 = _mm256_sub_ps(iy1,jy1);
811 dz11 = _mm256_sub_ps(iz1,jz1);
812 dx12 = _mm256_sub_ps(ix1,jx2);
813 dy12 = _mm256_sub_ps(iy1,jy2);
814 dz12 = _mm256_sub_ps(iz1,jz2);
815 dx13 = _mm256_sub_ps(ix1,jx3);
816 dy13 = _mm256_sub_ps(iy1,jy3);
817 dz13 = _mm256_sub_ps(iz1,jz3);
818 dx21 = _mm256_sub_ps(ix2,jx1);
819 dy21 = _mm256_sub_ps(iy2,jy1);
820 dz21 = _mm256_sub_ps(iz2,jz1);
821 dx22 = _mm256_sub_ps(ix2,jx2);
822 dy22 = _mm256_sub_ps(iy2,jy2);
823 dz22 = _mm256_sub_ps(iz2,jz2);
824 dx23 = _mm256_sub_ps(ix2,jx3);
825 dy23 = _mm256_sub_ps(iy2,jy3);
826 dz23 = _mm256_sub_ps(iz2,jz3);
827 dx31 = _mm256_sub_ps(ix3,jx1);
828 dy31 = _mm256_sub_ps(iy3,jy1);
829 dz31 = _mm256_sub_ps(iz3,jz1);
830 dx32 = _mm256_sub_ps(ix3,jx2);
831 dy32 = _mm256_sub_ps(iy3,jy2);
832 dz32 = _mm256_sub_ps(iz3,jz2);
833 dx33 = _mm256_sub_ps(ix3,jx3);
834 dy33 = _mm256_sub_ps(iy3,jy3);
835 dz33 = _mm256_sub_ps(iz3,jz3);
837 /* Calculate squared distance and things based on it */
838 rsq00 = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
839 rsq11 = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
840 rsq12 = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
841 rsq13 = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
842 rsq21 = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
843 rsq22 = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
844 rsq23 = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
845 rsq31 = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
846 rsq32 = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
847 rsq33 = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
849 rinv00 = avx256_invsqrt_f(rsq00);
850 rinv11 = avx256_invsqrt_f(rsq11);
851 rinv12 = avx256_invsqrt_f(rsq12);
852 rinv13 = avx256_invsqrt_f(rsq13);
853 rinv21 = avx256_invsqrt_f(rsq21);
854 rinv22 = avx256_invsqrt_f(rsq22);
855 rinv23 = avx256_invsqrt_f(rsq23);
856 rinv31 = avx256_invsqrt_f(rsq31);
857 rinv32 = avx256_invsqrt_f(rsq32);
858 rinv33 = avx256_invsqrt_f(rsq33);
860 rinvsq00 = _mm256_mul_ps(rinv00,rinv00);
861 rinvsq11 = _mm256_mul_ps(rinv11,rinv11);
862 rinvsq12 = _mm256_mul_ps(rinv12,rinv12);
863 rinvsq13 = _mm256_mul_ps(rinv13,rinv13);
864 rinvsq21 = _mm256_mul_ps(rinv21,rinv21);
865 rinvsq22 = _mm256_mul_ps(rinv22,rinv22);
866 rinvsq23 = _mm256_mul_ps(rinv23,rinv23);
867 rinvsq31 = _mm256_mul_ps(rinv31,rinv31);
868 rinvsq32 = _mm256_mul_ps(rinv32,rinv32);
869 rinvsq33 = _mm256_mul_ps(rinv33,rinv33);
871 fjx0 = _mm256_setzero_ps();
872 fjy0 = _mm256_setzero_ps();
873 fjz0 = _mm256_setzero_ps();
874 fjx1 = _mm256_setzero_ps();
875 fjy1 = _mm256_setzero_ps();
876 fjz1 = _mm256_setzero_ps();
877 fjx2 = _mm256_setzero_ps();
878 fjy2 = _mm256_setzero_ps();
879 fjz2 = _mm256_setzero_ps();
880 fjx3 = _mm256_setzero_ps();
881 fjy3 = _mm256_setzero_ps();
882 fjz3 = _mm256_setzero_ps();
884 /**************************
885 * CALCULATE INTERACTIONS *
886 **************************/
888 r00 = _mm256_mul_ps(rsq00,rinv00);
889 r00 = _mm256_andnot_ps(dummy_mask,r00);
891 /* Analytical LJ-PME */
892 rinvsix = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
893 ewcljrsq = _mm256_mul_ps(ewclj2,rsq00);
894 ewclj6 = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
895 exponent = avx256_exp_f(ewcljrsq);
896 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
897 poly = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
898 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
899 vvdw6 = _mm256_mul_ps(_mm256_sub_ps(c6_00,_mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly))),rinvsix);
900 vvdw12 = _mm256_mul_ps(c12_00,_mm256_mul_ps(rinvsix,rinvsix));
901 vvdw = _mm256_sub_ps(_mm256_mul_ps(vvdw12,one_twelfth),_mm256_mul_ps(vvdw6,one_sixth));
902 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
903 fvdw = _mm256_mul_ps(_mm256_sub_ps(vvdw12,_mm256_sub_ps(vvdw6,_mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6)))),rinvsq00);
905 /* Update potential sum for this i atom from the interaction with this j atom. */
906 vvdw = _mm256_andnot_ps(dummy_mask,vvdw);
907 vvdwsum = _mm256_add_ps(vvdwsum,vvdw);
911 fscal = _mm256_andnot_ps(dummy_mask,fscal);
913 /* Calculate temporary vectorial force */
914 tx = _mm256_mul_ps(fscal,dx00);
915 ty = _mm256_mul_ps(fscal,dy00);
916 tz = _mm256_mul_ps(fscal,dz00);
918 /* Update vectorial force */
919 fix0 = _mm256_add_ps(fix0,tx);
920 fiy0 = _mm256_add_ps(fiy0,ty);
921 fiz0 = _mm256_add_ps(fiz0,tz);
923 fjx0 = _mm256_add_ps(fjx0,tx);
924 fjy0 = _mm256_add_ps(fjy0,ty);
925 fjz0 = _mm256_add_ps(fjz0,tz);
927 /**************************
928 * CALCULATE INTERACTIONS *
929 **************************/
931 r11 = _mm256_mul_ps(rsq11,rinv11);
932 r11 = _mm256_andnot_ps(dummy_mask,r11);
934 /* EWALD ELECTROSTATICS */
936 /* Analytical PME correction */
937 zeta2 = _mm256_mul_ps(beta2,rsq11);
938 rinv3 = _mm256_mul_ps(rinvsq11,rinv11);
939 pmecorrF = avx256_pmecorrF_f(zeta2);
940 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
941 felec = _mm256_mul_ps(qq11,felec);
942 pmecorrV = avx256_pmecorrV_f(zeta2);
943 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
944 velec = _mm256_sub_ps(rinv11,pmecorrV);
945 velec = _mm256_mul_ps(qq11,velec);
947 /* Update potential sum for this i atom from the interaction with this j atom. */
948 velec = _mm256_andnot_ps(dummy_mask,velec);
949 velecsum = _mm256_add_ps(velecsum,velec);
953 fscal = _mm256_andnot_ps(dummy_mask,fscal);
955 /* Calculate temporary vectorial force */
956 tx = _mm256_mul_ps(fscal,dx11);
957 ty = _mm256_mul_ps(fscal,dy11);
958 tz = _mm256_mul_ps(fscal,dz11);
960 /* Update vectorial force */
961 fix1 = _mm256_add_ps(fix1,tx);
962 fiy1 = _mm256_add_ps(fiy1,ty);
963 fiz1 = _mm256_add_ps(fiz1,tz);
965 fjx1 = _mm256_add_ps(fjx1,tx);
966 fjy1 = _mm256_add_ps(fjy1,ty);
967 fjz1 = _mm256_add_ps(fjz1,tz);
969 /**************************
970 * CALCULATE INTERACTIONS *
971 **************************/
973 r12 = _mm256_mul_ps(rsq12,rinv12);
974 r12 = _mm256_andnot_ps(dummy_mask,r12);
976 /* EWALD ELECTROSTATICS */
978 /* Analytical PME correction */
979 zeta2 = _mm256_mul_ps(beta2,rsq12);
980 rinv3 = _mm256_mul_ps(rinvsq12,rinv12);
981 pmecorrF = avx256_pmecorrF_f(zeta2);
982 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
983 felec = _mm256_mul_ps(qq12,felec);
984 pmecorrV = avx256_pmecorrV_f(zeta2);
985 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
986 velec = _mm256_sub_ps(rinv12,pmecorrV);
987 velec = _mm256_mul_ps(qq12,velec);
989 /* Update potential sum for this i atom from the interaction with this j atom. */
990 velec = _mm256_andnot_ps(dummy_mask,velec);
991 velecsum = _mm256_add_ps(velecsum,velec);
995 fscal = _mm256_andnot_ps(dummy_mask,fscal);
997 /* Calculate temporary vectorial force */
998 tx = _mm256_mul_ps(fscal,dx12);
999 ty = _mm256_mul_ps(fscal,dy12);
1000 tz = _mm256_mul_ps(fscal,dz12);
1002 /* Update vectorial force */
1003 fix1 = _mm256_add_ps(fix1,tx);
1004 fiy1 = _mm256_add_ps(fiy1,ty);
1005 fiz1 = _mm256_add_ps(fiz1,tz);
1007 fjx2 = _mm256_add_ps(fjx2,tx);
1008 fjy2 = _mm256_add_ps(fjy2,ty);
1009 fjz2 = _mm256_add_ps(fjz2,tz);
1011 /**************************
1012 * CALCULATE INTERACTIONS *
1013 **************************/
1015 r13 = _mm256_mul_ps(rsq13,rinv13);
1016 r13 = _mm256_andnot_ps(dummy_mask,r13);
1018 /* EWALD ELECTROSTATICS */
1020 /* Analytical PME correction */
1021 zeta2 = _mm256_mul_ps(beta2,rsq13);
1022 rinv3 = _mm256_mul_ps(rinvsq13,rinv13);
1023 pmecorrF = avx256_pmecorrF_f(zeta2);
1024 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1025 felec = _mm256_mul_ps(qq13,felec);
1026 pmecorrV = avx256_pmecorrV_f(zeta2);
1027 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1028 velec = _mm256_sub_ps(rinv13,pmecorrV);
1029 velec = _mm256_mul_ps(qq13,velec);
1031 /* Update potential sum for this i atom from the interaction with this j atom. */
1032 velec = _mm256_andnot_ps(dummy_mask,velec);
1033 velecsum = _mm256_add_ps(velecsum,velec);
1037 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1039 /* Calculate temporary vectorial force */
1040 tx = _mm256_mul_ps(fscal,dx13);
1041 ty = _mm256_mul_ps(fscal,dy13);
1042 tz = _mm256_mul_ps(fscal,dz13);
1044 /* Update vectorial force */
1045 fix1 = _mm256_add_ps(fix1,tx);
1046 fiy1 = _mm256_add_ps(fiy1,ty);
1047 fiz1 = _mm256_add_ps(fiz1,tz);
1049 fjx3 = _mm256_add_ps(fjx3,tx);
1050 fjy3 = _mm256_add_ps(fjy3,ty);
1051 fjz3 = _mm256_add_ps(fjz3,tz);
1053 /**************************
1054 * CALCULATE INTERACTIONS *
1055 **************************/
1057 r21 = _mm256_mul_ps(rsq21,rinv21);
1058 r21 = _mm256_andnot_ps(dummy_mask,r21);
1060 /* EWALD ELECTROSTATICS */
1062 /* Analytical PME correction */
1063 zeta2 = _mm256_mul_ps(beta2,rsq21);
1064 rinv3 = _mm256_mul_ps(rinvsq21,rinv21);
1065 pmecorrF = avx256_pmecorrF_f(zeta2);
1066 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1067 felec = _mm256_mul_ps(qq21,felec);
1068 pmecorrV = avx256_pmecorrV_f(zeta2);
1069 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1070 velec = _mm256_sub_ps(rinv21,pmecorrV);
1071 velec = _mm256_mul_ps(qq21,velec);
1073 /* Update potential sum for this i atom from the interaction with this j atom. */
1074 velec = _mm256_andnot_ps(dummy_mask,velec);
1075 velecsum = _mm256_add_ps(velecsum,velec);
1079 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1081 /* Calculate temporary vectorial force */
1082 tx = _mm256_mul_ps(fscal,dx21);
1083 ty = _mm256_mul_ps(fscal,dy21);
1084 tz = _mm256_mul_ps(fscal,dz21);
1086 /* Update vectorial force */
1087 fix2 = _mm256_add_ps(fix2,tx);
1088 fiy2 = _mm256_add_ps(fiy2,ty);
1089 fiz2 = _mm256_add_ps(fiz2,tz);
1091 fjx1 = _mm256_add_ps(fjx1,tx);
1092 fjy1 = _mm256_add_ps(fjy1,ty);
1093 fjz1 = _mm256_add_ps(fjz1,tz);
1095 /**************************
1096 * CALCULATE INTERACTIONS *
1097 **************************/
1099 r22 = _mm256_mul_ps(rsq22,rinv22);
1100 r22 = _mm256_andnot_ps(dummy_mask,r22);
1102 /* EWALD ELECTROSTATICS */
1104 /* Analytical PME correction */
1105 zeta2 = _mm256_mul_ps(beta2,rsq22);
1106 rinv3 = _mm256_mul_ps(rinvsq22,rinv22);
1107 pmecorrF = avx256_pmecorrF_f(zeta2);
1108 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1109 felec = _mm256_mul_ps(qq22,felec);
1110 pmecorrV = avx256_pmecorrV_f(zeta2);
1111 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1112 velec = _mm256_sub_ps(rinv22,pmecorrV);
1113 velec = _mm256_mul_ps(qq22,velec);
1115 /* Update potential sum for this i atom from the interaction with this j atom. */
1116 velec = _mm256_andnot_ps(dummy_mask,velec);
1117 velecsum = _mm256_add_ps(velecsum,velec);
1121 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1123 /* Calculate temporary vectorial force */
1124 tx = _mm256_mul_ps(fscal,dx22);
1125 ty = _mm256_mul_ps(fscal,dy22);
1126 tz = _mm256_mul_ps(fscal,dz22);
1128 /* Update vectorial force */
1129 fix2 = _mm256_add_ps(fix2,tx);
1130 fiy2 = _mm256_add_ps(fiy2,ty);
1131 fiz2 = _mm256_add_ps(fiz2,tz);
1133 fjx2 = _mm256_add_ps(fjx2,tx);
1134 fjy2 = _mm256_add_ps(fjy2,ty);
1135 fjz2 = _mm256_add_ps(fjz2,tz);
1137 /**************************
1138 * CALCULATE INTERACTIONS *
1139 **************************/
1141 r23 = _mm256_mul_ps(rsq23,rinv23);
1142 r23 = _mm256_andnot_ps(dummy_mask,r23);
1144 /* EWALD ELECTROSTATICS */
1146 /* Analytical PME correction */
1147 zeta2 = _mm256_mul_ps(beta2,rsq23);
1148 rinv3 = _mm256_mul_ps(rinvsq23,rinv23);
1149 pmecorrF = avx256_pmecorrF_f(zeta2);
1150 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1151 felec = _mm256_mul_ps(qq23,felec);
1152 pmecorrV = avx256_pmecorrV_f(zeta2);
1153 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1154 velec = _mm256_sub_ps(rinv23,pmecorrV);
1155 velec = _mm256_mul_ps(qq23,velec);
1157 /* Update potential sum for this i atom from the interaction with this j atom. */
1158 velec = _mm256_andnot_ps(dummy_mask,velec);
1159 velecsum = _mm256_add_ps(velecsum,velec);
1163 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1165 /* Calculate temporary vectorial force */
1166 tx = _mm256_mul_ps(fscal,dx23);
1167 ty = _mm256_mul_ps(fscal,dy23);
1168 tz = _mm256_mul_ps(fscal,dz23);
1170 /* Update vectorial force */
1171 fix2 = _mm256_add_ps(fix2,tx);
1172 fiy2 = _mm256_add_ps(fiy2,ty);
1173 fiz2 = _mm256_add_ps(fiz2,tz);
1175 fjx3 = _mm256_add_ps(fjx3,tx);
1176 fjy3 = _mm256_add_ps(fjy3,ty);
1177 fjz3 = _mm256_add_ps(fjz3,tz);
1179 /**************************
1180 * CALCULATE INTERACTIONS *
1181 **************************/
1183 r31 = _mm256_mul_ps(rsq31,rinv31);
1184 r31 = _mm256_andnot_ps(dummy_mask,r31);
1186 /* EWALD ELECTROSTATICS */
1188 /* Analytical PME correction */
1189 zeta2 = _mm256_mul_ps(beta2,rsq31);
1190 rinv3 = _mm256_mul_ps(rinvsq31,rinv31);
1191 pmecorrF = avx256_pmecorrF_f(zeta2);
1192 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1193 felec = _mm256_mul_ps(qq31,felec);
1194 pmecorrV = avx256_pmecorrV_f(zeta2);
1195 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1196 velec = _mm256_sub_ps(rinv31,pmecorrV);
1197 velec = _mm256_mul_ps(qq31,velec);
1199 /* Update potential sum for this i atom from the interaction with this j atom. */
1200 velec = _mm256_andnot_ps(dummy_mask,velec);
1201 velecsum = _mm256_add_ps(velecsum,velec);
1205 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1207 /* Calculate temporary vectorial force */
1208 tx = _mm256_mul_ps(fscal,dx31);
1209 ty = _mm256_mul_ps(fscal,dy31);
1210 tz = _mm256_mul_ps(fscal,dz31);
1212 /* Update vectorial force */
1213 fix3 = _mm256_add_ps(fix3,tx);
1214 fiy3 = _mm256_add_ps(fiy3,ty);
1215 fiz3 = _mm256_add_ps(fiz3,tz);
1217 fjx1 = _mm256_add_ps(fjx1,tx);
1218 fjy1 = _mm256_add_ps(fjy1,ty);
1219 fjz1 = _mm256_add_ps(fjz1,tz);
1221 /**************************
1222 * CALCULATE INTERACTIONS *
1223 **************************/
1225 r32 = _mm256_mul_ps(rsq32,rinv32);
1226 r32 = _mm256_andnot_ps(dummy_mask,r32);
1228 /* EWALD ELECTROSTATICS */
1230 /* Analytical PME correction */
1231 zeta2 = _mm256_mul_ps(beta2,rsq32);
1232 rinv3 = _mm256_mul_ps(rinvsq32,rinv32);
1233 pmecorrF = avx256_pmecorrF_f(zeta2);
1234 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1235 felec = _mm256_mul_ps(qq32,felec);
1236 pmecorrV = avx256_pmecorrV_f(zeta2);
1237 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1238 velec = _mm256_sub_ps(rinv32,pmecorrV);
1239 velec = _mm256_mul_ps(qq32,velec);
1241 /* Update potential sum for this i atom from the interaction with this j atom. */
1242 velec = _mm256_andnot_ps(dummy_mask,velec);
1243 velecsum = _mm256_add_ps(velecsum,velec);
1247 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1249 /* Calculate temporary vectorial force */
1250 tx = _mm256_mul_ps(fscal,dx32);
1251 ty = _mm256_mul_ps(fscal,dy32);
1252 tz = _mm256_mul_ps(fscal,dz32);
1254 /* Update vectorial force */
1255 fix3 = _mm256_add_ps(fix3,tx);
1256 fiy3 = _mm256_add_ps(fiy3,ty);
1257 fiz3 = _mm256_add_ps(fiz3,tz);
1259 fjx2 = _mm256_add_ps(fjx2,tx);
1260 fjy2 = _mm256_add_ps(fjy2,ty);
1261 fjz2 = _mm256_add_ps(fjz2,tz);
1263 /**************************
1264 * CALCULATE INTERACTIONS *
1265 **************************/
1267 r33 = _mm256_mul_ps(rsq33,rinv33);
1268 r33 = _mm256_andnot_ps(dummy_mask,r33);
1270 /* EWALD ELECTROSTATICS */
1272 /* Analytical PME correction */
1273 zeta2 = _mm256_mul_ps(beta2,rsq33);
1274 rinv3 = _mm256_mul_ps(rinvsq33,rinv33);
1275 pmecorrF = avx256_pmecorrF_f(zeta2);
1276 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1277 felec = _mm256_mul_ps(qq33,felec);
1278 pmecorrV = avx256_pmecorrV_f(zeta2);
1279 pmecorrV = _mm256_mul_ps(pmecorrV,beta);
1280 velec = _mm256_sub_ps(rinv33,pmecorrV);
1281 velec = _mm256_mul_ps(qq33,velec);
1283 /* Update potential sum for this i atom from the interaction with this j atom. */
1284 velec = _mm256_andnot_ps(dummy_mask,velec);
1285 velecsum = _mm256_add_ps(velecsum,velec);
1289 fscal = _mm256_andnot_ps(dummy_mask,fscal);
1291 /* Calculate temporary vectorial force */
1292 tx = _mm256_mul_ps(fscal,dx33);
1293 ty = _mm256_mul_ps(fscal,dy33);
1294 tz = _mm256_mul_ps(fscal,dz33);
1296 /* Update vectorial force */
1297 fix3 = _mm256_add_ps(fix3,tx);
1298 fiy3 = _mm256_add_ps(fiy3,ty);
1299 fiz3 = _mm256_add_ps(fiz3,tz);
1301 fjx3 = _mm256_add_ps(fjx3,tx);
1302 fjy3 = _mm256_add_ps(fjy3,ty);
1303 fjz3 = _mm256_add_ps(fjz3,tz);
1305 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1306 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1307 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1308 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1309 fjptrE = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
1310 fjptrF = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
1311 fjptrG = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
1312 fjptrH = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
1314 gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
1315 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1316 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1318 /* Inner loop uses 820 flops */
1321 /* End of innermost loop */
1323 gmx_mm256_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1324 f+i_coord_offset,fshift+i_shift_offset);
1327 /* Update potential energies */
1328 gmx_mm256_update_1pot_ps(velecsum,kernel_data->energygrp_elec+ggid);
1329 gmx_mm256_update_1pot_ps(vvdwsum,kernel_data->energygrp_vdw+ggid);
1331 /* Increment number of inner iterations */
1332 inneriter += j_index_end - j_index_start;
1334 /* Outer loop uses 26 flops */
1337 /* Increment number of outer iterations */
1340 /* Update outer/inner flops */
1342 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*820);
1345 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJEw_GeomW4W4_F_avx_256_single
1346 * Electrostatics interaction: Ewald
1347 * VdW interaction: LJEwald
1348 * Geometry: Water4-Water4
1349 * Calculate force/pot: Force
1352 nb_kernel_ElecEw_VdwLJEw_GeomW4W4_F_avx_256_single
1353 (t_nblist * gmx_restrict nlist,
1354 rvec * gmx_restrict xx,
1355 rvec * gmx_restrict ff,
1356 struct t_forcerec * gmx_restrict fr,
1357 t_mdatoms * gmx_restrict mdatoms,
1358 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1359 t_nrnb * gmx_restrict nrnb)
1361 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1362 * just 0 for non-waters.
1363 * Suffixes A,B,C,D,E,F,G,H refer to j loop unrolling done with AVX, e.g. for the eight different
1364 * jnr indices corresponding to data put in the four positions in the SIMD register.
1366 int i_shift_offset,i_coord_offset,outeriter,inneriter;
1367 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1368 int jnrA,jnrB,jnrC,jnrD;
1369 int jnrE,jnrF,jnrG,jnrH;
1370 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1371 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1372 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1373 int j_coord_offsetE,j_coord_offsetF,j_coord_offsetG,j_coord_offsetH;
1374 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
1375 real rcutoff_scalar;
1376 real *shiftvec,*fshift,*x,*f;
1377 real *fjptrA,*fjptrB,*fjptrC,*fjptrD,*fjptrE,*fjptrF,*fjptrG,*fjptrH;
1378 real scratch[4*DIM];
1379 __m256 tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1380 real * vdwioffsetptr0;
1381 real * vdwgridioffsetptr0;
1382 __m256 ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1383 real * vdwioffsetptr1;
1384 real * vdwgridioffsetptr1;
1385 __m256 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1386 real * vdwioffsetptr2;
1387 real * vdwgridioffsetptr2;
1388 __m256 ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1389 real * vdwioffsetptr3;
1390 real * vdwgridioffsetptr3;
1391 __m256 ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1392 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D,vdwjidx0E,vdwjidx0F,vdwjidx0G,vdwjidx0H;
1393 __m256 jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1394 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D,vdwjidx1E,vdwjidx1F,vdwjidx1G,vdwjidx1H;
1395 __m256 jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1396 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D,vdwjidx2E,vdwjidx2F,vdwjidx2G,vdwjidx2H;
1397 __m256 jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1398 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D,vdwjidx3E,vdwjidx3F,vdwjidx3G,vdwjidx3H;
1399 __m256 jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1400 __m256 dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1401 __m256 dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1402 __m256 dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1403 __m256 dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1404 __m256 dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1405 __m256 dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1406 __m256 dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1407 __m256 dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1408 __m256 dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1409 __m256 dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1410 __m256 velec,felec,velecsum,facel,crf,krf,krf2;
1413 __m256 rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1416 __m256 one_sixth = _mm256_set1_ps(1.0/6.0);
1417 __m256 one_twelfth = _mm256_set1_ps(1.0/12.0);
1429 __m256 ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
1430 __m256 one_half = _mm256_set1_ps(0.5);
1431 __m256 minus_one = _mm256_set1_ps(-1.0);
1433 __m128i ewitab_lo,ewitab_hi;
1434 __m256 ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1435 __m256 beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
1437 __m256 dummy_mask,cutoff_mask;
1438 __m256 signbit = _mm256_castsi256_ps( _mm256_set1_epi32(0x80000000) );
1439 __m256 one = _mm256_set1_ps(1.0);
1440 __m256 two = _mm256_set1_ps(2.0);
1446 jindex = nlist->jindex;
1448 shiftidx = nlist->shift;
1450 shiftvec = fr->shift_vec[0];
1451 fshift = fr->fshift[0];
1452 facel = _mm256_set1_ps(fr->ic->epsfac);
1453 charge = mdatoms->chargeA;
1454 nvdwtype = fr->ntype;
1455 vdwparam = fr->nbfp;
1456 vdwtype = mdatoms->typeA;
1457 vdwgridparam = fr->ljpme_c6grid;
1458 sh_lj_ewald = _mm256_set1_ps(fr->ic->sh_lj_ewald);
1459 ewclj = _mm256_set1_ps(fr->ic->ewaldcoeff_lj);
1460 ewclj2 = _mm256_mul_ps(minus_one,_mm256_mul_ps(ewclj,ewclj));
1462 sh_ewald = _mm256_set1_ps(fr->ic->sh_ewald);
1463 beta = _mm256_set1_ps(fr->ic->ewaldcoeff_q);
1464 beta2 = _mm256_mul_ps(beta,beta);
1465 beta3 = _mm256_mul_ps(beta,beta2);
1467 ewtab = fr->ic->tabq_coul_F;
1468 ewtabscale = _mm256_set1_ps(fr->ic->tabq_scale);
1469 ewtabhalfspace = _mm256_set1_ps(0.5/fr->ic->tabq_scale);
1471 /* Setup water-specific parameters */
1472 inr = nlist->iinr[0];
1473 iq1 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+1]));
1474 iq2 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+2]));
1475 iq3 = _mm256_mul_ps(facel,_mm256_set1_ps(charge[inr+3]));
1476 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
1477 vdwgridioffsetptr0 = vdwgridparam+2*nvdwtype*vdwtype[inr+0];
1479 jq1 = _mm256_set1_ps(charge[inr+1]);
1480 jq2 = _mm256_set1_ps(charge[inr+2]);
1481 jq3 = _mm256_set1_ps(charge[inr+3]);
1482 vdwjidx0A = 2*vdwtype[inr+0];
1483 c6_00 = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A]);
1484 c12_00 = _mm256_set1_ps(vdwioffsetptr0[vdwjidx0A+1]);
1485 c6grid_00 = _mm256_set1_ps(vdwgridioffsetptr0[vdwjidx0A]);
1486 qq11 = _mm256_mul_ps(iq1,jq1);
1487 qq12 = _mm256_mul_ps(iq1,jq2);
1488 qq13 = _mm256_mul_ps(iq1,jq3);
1489 qq21 = _mm256_mul_ps(iq2,jq1);
1490 qq22 = _mm256_mul_ps(iq2,jq2);
1491 qq23 = _mm256_mul_ps(iq2,jq3);
1492 qq31 = _mm256_mul_ps(iq3,jq1);
1493 qq32 = _mm256_mul_ps(iq3,jq2);
1494 qq33 = _mm256_mul_ps(iq3,jq3);
1496 /* Avoid stupid compiler warnings */
1497 jnrA = jnrB = jnrC = jnrD = jnrE = jnrF = jnrG = jnrH = 0;
1498 j_coord_offsetA = 0;
1499 j_coord_offsetB = 0;
1500 j_coord_offsetC = 0;
1501 j_coord_offsetD = 0;
1502 j_coord_offsetE = 0;
1503 j_coord_offsetF = 0;
1504 j_coord_offsetG = 0;
1505 j_coord_offsetH = 0;
1510 for(iidx=0;iidx<4*DIM;iidx++)
1512 scratch[iidx] = 0.0;
1515 /* Start outer loop over neighborlists */
1516 for(iidx=0; iidx<nri; iidx++)
1518 /* Load shift vector for this list */
1519 i_shift_offset = DIM*shiftidx[iidx];
1521 /* Load limits for loop over neighbors */
1522 j_index_start = jindex[iidx];
1523 j_index_end = jindex[iidx+1];
1525 /* Get outer coordinate index */
1527 i_coord_offset = DIM*inr;
1529 /* Load i particle coords and add shift vector */
1530 gmx_mm256_load_shift_and_4rvec_broadcast_ps(shiftvec+i_shift_offset,x+i_coord_offset,
1531 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1533 fix0 = _mm256_setzero_ps();
1534 fiy0 = _mm256_setzero_ps();
1535 fiz0 = _mm256_setzero_ps();
1536 fix1 = _mm256_setzero_ps();
1537 fiy1 = _mm256_setzero_ps();
1538 fiz1 = _mm256_setzero_ps();
1539 fix2 = _mm256_setzero_ps();
1540 fiy2 = _mm256_setzero_ps();
1541 fiz2 = _mm256_setzero_ps();
1542 fix3 = _mm256_setzero_ps();
1543 fiy3 = _mm256_setzero_ps();
1544 fiz3 = _mm256_setzero_ps();
1546 /* Start inner kernel loop */
1547 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+7]>=0; jidx+=8)
1550 /* Get j neighbor index, and coordinate index */
1552 jnrB = jjnr[jidx+1];
1553 jnrC = jjnr[jidx+2];
1554 jnrD = jjnr[jidx+3];
1555 jnrE = jjnr[jidx+4];
1556 jnrF = jjnr[jidx+5];
1557 jnrG = jjnr[jidx+6];
1558 jnrH = jjnr[jidx+7];
1559 j_coord_offsetA = DIM*jnrA;
1560 j_coord_offsetB = DIM*jnrB;
1561 j_coord_offsetC = DIM*jnrC;
1562 j_coord_offsetD = DIM*jnrD;
1563 j_coord_offsetE = DIM*jnrE;
1564 j_coord_offsetF = DIM*jnrF;
1565 j_coord_offsetG = DIM*jnrG;
1566 j_coord_offsetH = DIM*jnrH;
1568 /* load j atom coordinates */
1569 gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
1570 x+j_coord_offsetC,x+j_coord_offsetD,
1571 x+j_coord_offsetE,x+j_coord_offsetF,
1572 x+j_coord_offsetG,x+j_coord_offsetH,
1573 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1574 &jy2,&jz2,&jx3,&jy3,&jz3);
1576 /* Calculate displacement vector */
1577 dx00 = _mm256_sub_ps(ix0,jx0);
1578 dy00 = _mm256_sub_ps(iy0,jy0);
1579 dz00 = _mm256_sub_ps(iz0,jz0);
1580 dx11 = _mm256_sub_ps(ix1,jx1);
1581 dy11 = _mm256_sub_ps(iy1,jy1);
1582 dz11 = _mm256_sub_ps(iz1,jz1);
1583 dx12 = _mm256_sub_ps(ix1,jx2);
1584 dy12 = _mm256_sub_ps(iy1,jy2);
1585 dz12 = _mm256_sub_ps(iz1,jz2);
1586 dx13 = _mm256_sub_ps(ix1,jx3);
1587 dy13 = _mm256_sub_ps(iy1,jy3);
1588 dz13 = _mm256_sub_ps(iz1,jz3);
1589 dx21 = _mm256_sub_ps(ix2,jx1);
1590 dy21 = _mm256_sub_ps(iy2,jy1);
1591 dz21 = _mm256_sub_ps(iz2,jz1);
1592 dx22 = _mm256_sub_ps(ix2,jx2);
1593 dy22 = _mm256_sub_ps(iy2,jy2);
1594 dz22 = _mm256_sub_ps(iz2,jz2);
1595 dx23 = _mm256_sub_ps(ix2,jx3);
1596 dy23 = _mm256_sub_ps(iy2,jy3);
1597 dz23 = _mm256_sub_ps(iz2,jz3);
1598 dx31 = _mm256_sub_ps(ix3,jx1);
1599 dy31 = _mm256_sub_ps(iy3,jy1);
1600 dz31 = _mm256_sub_ps(iz3,jz1);
1601 dx32 = _mm256_sub_ps(ix3,jx2);
1602 dy32 = _mm256_sub_ps(iy3,jy2);
1603 dz32 = _mm256_sub_ps(iz3,jz2);
1604 dx33 = _mm256_sub_ps(ix3,jx3);
1605 dy33 = _mm256_sub_ps(iy3,jy3);
1606 dz33 = _mm256_sub_ps(iz3,jz3);
1608 /* Calculate squared distance and things based on it */
1609 rsq00 = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
1610 rsq11 = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
1611 rsq12 = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
1612 rsq13 = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
1613 rsq21 = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
1614 rsq22 = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
1615 rsq23 = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
1616 rsq31 = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
1617 rsq32 = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
1618 rsq33 = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
1620 rinv00 = avx256_invsqrt_f(rsq00);
1621 rinv11 = avx256_invsqrt_f(rsq11);
1622 rinv12 = avx256_invsqrt_f(rsq12);
1623 rinv13 = avx256_invsqrt_f(rsq13);
1624 rinv21 = avx256_invsqrt_f(rsq21);
1625 rinv22 = avx256_invsqrt_f(rsq22);
1626 rinv23 = avx256_invsqrt_f(rsq23);
1627 rinv31 = avx256_invsqrt_f(rsq31);
1628 rinv32 = avx256_invsqrt_f(rsq32);
1629 rinv33 = avx256_invsqrt_f(rsq33);
1631 rinvsq00 = _mm256_mul_ps(rinv00,rinv00);
1632 rinvsq11 = _mm256_mul_ps(rinv11,rinv11);
1633 rinvsq12 = _mm256_mul_ps(rinv12,rinv12);
1634 rinvsq13 = _mm256_mul_ps(rinv13,rinv13);
1635 rinvsq21 = _mm256_mul_ps(rinv21,rinv21);
1636 rinvsq22 = _mm256_mul_ps(rinv22,rinv22);
1637 rinvsq23 = _mm256_mul_ps(rinv23,rinv23);
1638 rinvsq31 = _mm256_mul_ps(rinv31,rinv31);
1639 rinvsq32 = _mm256_mul_ps(rinv32,rinv32);
1640 rinvsq33 = _mm256_mul_ps(rinv33,rinv33);
1642 fjx0 = _mm256_setzero_ps();
1643 fjy0 = _mm256_setzero_ps();
1644 fjz0 = _mm256_setzero_ps();
1645 fjx1 = _mm256_setzero_ps();
1646 fjy1 = _mm256_setzero_ps();
1647 fjz1 = _mm256_setzero_ps();
1648 fjx2 = _mm256_setzero_ps();
1649 fjy2 = _mm256_setzero_ps();
1650 fjz2 = _mm256_setzero_ps();
1651 fjx3 = _mm256_setzero_ps();
1652 fjy3 = _mm256_setzero_ps();
1653 fjz3 = _mm256_setzero_ps();
1655 /**************************
1656 * CALCULATE INTERACTIONS *
1657 **************************/
1659 r00 = _mm256_mul_ps(rsq00,rinv00);
1661 /* Analytical LJ-PME */
1662 rinvsix = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
1663 ewcljrsq = _mm256_mul_ps(ewclj2,rsq00);
1664 ewclj6 = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
1665 exponent = avx256_exp_f(ewcljrsq);
1666 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
1667 poly = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
1668 /* f6A = 6 * C6grid * (1 - poly) */
1669 f6A = _mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly));
1670 /* f6B = C6grid * exponent * beta^6 */
1671 f6B = _mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6));
1672 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
1673 fvdw = _mm256_mul_ps(_mm256_add_ps(_mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),_mm256_sub_ps(c6_00,f6A)),rinvsix),f6B),rinvsq00);
1677 /* Calculate temporary vectorial force */
1678 tx = _mm256_mul_ps(fscal,dx00);
1679 ty = _mm256_mul_ps(fscal,dy00);
1680 tz = _mm256_mul_ps(fscal,dz00);
1682 /* Update vectorial force */
1683 fix0 = _mm256_add_ps(fix0,tx);
1684 fiy0 = _mm256_add_ps(fiy0,ty);
1685 fiz0 = _mm256_add_ps(fiz0,tz);
1687 fjx0 = _mm256_add_ps(fjx0,tx);
1688 fjy0 = _mm256_add_ps(fjy0,ty);
1689 fjz0 = _mm256_add_ps(fjz0,tz);
1691 /**************************
1692 * CALCULATE INTERACTIONS *
1693 **************************/
1695 r11 = _mm256_mul_ps(rsq11,rinv11);
1697 /* EWALD ELECTROSTATICS */
1699 /* Analytical PME correction */
1700 zeta2 = _mm256_mul_ps(beta2,rsq11);
1701 rinv3 = _mm256_mul_ps(rinvsq11,rinv11);
1702 pmecorrF = avx256_pmecorrF_f(zeta2);
1703 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1704 felec = _mm256_mul_ps(qq11,felec);
1708 /* Calculate temporary vectorial force */
1709 tx = _mm256_mul_ps(fscal,dx11);
1710 ty = _mm256_mul_ps(fscal,dy11);
1711 tz = _mm256_mul_ps(fscal,dz11);
1713 /* Update vectorial force */
1714 fix1 = _mm256_add_ps(fix1,tx);
1715 fiy1 = _mm256_add_ps(fiy1,ty);
1716 fiz1 = _mm256_add_ps(fiz1,tz);
1718 fjx1 = _mm256_add_ps(fjx1,tx);
1719 fjy1 = _mm256_add_ps(fjy1,ty);
1720 fjz1 = _mm256_add_ps(fjz1,tz);
1722 /**************************
1723 * CALCULATE INTERACTIONS *
1724 **************************/
1726 r12 = _mm256_mul_ps(rsq12,rinv12);
1728 /* EWALD ELECTROSTATICS */
1730 /* Analytical PME correction */
1731 zeta2 = _mm256_mul_ps(beta2,rsq12);
1732 rinv3 = _mm256_mul_ps(rinvsq12,rinv12);
1733 pmecorrF = avx256_pmecorrF_f(zeta2);
1734 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1735 felec = _mm256_mul_ps(qq12,felec);
1739 /* Calculate temporary vectorial force */
1740 tx = _mm256_mul_ps(fscal,dx12);
1741 ty = _mm256_mul_ps(fscal,dy12);
1742 tz = _mm256_mul_ps(fscal,dz12);
1744 /* Update vectorial force */
1745 fix1 = _mm256_add_ps(fix1,tx);
1746 fiy1 = _mm256_add_ps(fiy1,ty);
1747 fiz1 = _mm256_add_ps(fiz1,tz);
1749 fjx2 = _mm256_add_ps(fjx2,tx);
1750 fjy2 = _mm256_add_ps(fjy2,ty);
1751 fjz2 = _mm256_add_ps(fjz2,tz);
1753 /**************************
1754 * CALCULATE INTERACTIONS *
1755 **************************/
1757 r13 = _mm256_mul_ps(rsq13,rinv13);
1759 /* EWALD ELECTROSTATICS */
1761 /* Analytical PME correction */
1762 zeta2 = _mm256_mul_ps(beta2,rsq13);
1763 rinv3 = _mm256_mul_ps(rinvsq13,rinv13);
1764 pmecorrF = avx256_pmecorrF_f(zeta2);
1765 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1766 felec = _mm256_mul_ps(qq13,felec);
1770 /* Calculate temporary vectorial force */
1771 tx = _mm256_mul_ps(fscal,dx13);
1772 ty = _mm256_mul_ps(fscal,dy13);
1773 tz = _mm256_mul_ps(fscal,dz13);
1775 /* Update vectorial force */
1776 fix1 = _mm256_add_ps(fix1,tx);
1777 fiy1 = _mm256_add_ps(fiy1,ty);
1778 fiz1 = _mm256_add_ps(fiz1,tz);
1780 fjx3 = _mm256_add_ps(fjx3,tx);
1781 fjy3 = _mm256_add_ps(fjy3,ty);
1782 fjz3 = _mm256_add_ps(fjz3,tz);
1784 /**************************
1785 * CALCULATE INTERACTIONS *
1786 **************************/
1788 r21 = _mm256_mul_ps(rsq21,rinv21);
1790 /* EWALD ELECTROSTATICS */
1792 /* Analytical PME correction */
1793 zeta2 = _mm256_mul_ps(beta2,rsq21);
1794 rinv3 = _mm256_mul_ps(rinvsq21,rinv21);
1795 pmecorrF = avx256_pmecorrF_f(zeta2);
1796 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1797 felec = _mm256_mul_ps(qq21,felec);
1801 /* Calculate temporary vectorial force */
1802 tx = _mm256_mul_ps(fscal,dx21);
1803 ty = _mm256_mul_ps(fscal,dy21);
1804 tz = _mm256_mul_ps(fscal,dz21);
1806 /* Update vectorial force */
1807 fix2 = _mm256_add_ps(fix2,tx);
1808 fiy2 = _mm256_add_ps(fiy2,ty);
1809 fiz2 = _mm256_add_ps(fiz2,tz);
1811 fjx1 = _mm256_add_ps(fjx1,tx);
1812 fjy1 = _mm256_add_ps(fjy1,ty);
1813 fjz1 = _mm256_add_ps(fjz1,tz);
1815 /**************************
1816 * CALCULATE INTERACTIONS *
1817 **************************/
1819 r22 = _mm256_mul_ps(rsq22,rinv22);
1821 /* EWALD ELECTROSTATICS */
1823 /* Analytical PME correction */
1824 zeta2 = _mm256_mul_ps(beta2,rsq22);
1825 rinv3 = _mm256_mul_ps(rinvsq22,rinv22);
1826 pmecorrF = avx256_pmecorrF_f(zeta2);
1827 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1828 felec = _mm256_mul_ps(qq22,felec);
1832 /* Calculate temporary vectorial force */
1833 tx = _mm256_mul_ps(fscal,dx22);
1834 ty = _mm256_mul_ps(fscal,dy22);
1835 tz = _mm256_mul_ps(fscal,dz22);
1837 /* Update vectorial force */
1838 fix2 = _mm256_add_ps(fix2,tx);
1839 fiy2 = _mm256_add_ps(fiy2,ty);
1840 fiz2 = _mm256_add_ps(fiz2,tz);
1842 fjx2 = _mm256_add_ps(fjx2,tx);
1843 fjy2 = _mm256_add_ps(fjy2,ty);
1844 fjz2 = _mm256_add_ps(fjz2,tz);
1846 /**************************
1847 * CALCULATE INTERACTIONS *
1848 **************************/
1850 r23 = _mm256_mul_ps(rsq23,rinv23);
1852 /* EWALD ELECTROSTATICS */
1854 /* Analytical PME correction */
1855 zeta2 = _mm256_mul_ps(beta2,rsq23);
1856 rinv3 = _mm256_mul_ps(rinvsq23,rinv23);
1857 pmecorrF = avx256_pmecorrF_f(zeta2);
1858 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1859 felec = _mm256_mul_ps(qq23,felec);
1863 /* Calculate temporary vectorial force */
1864 tx = _mm256_mul_ps(fscal,dx23);
1865 ty = _mm256_mul_ps(fscal,dy23);
1866 tz = _mm256_mul_ps(fscal,dz23);
1868 /* Update vectorial force */
1869 fix2 = _mm256_add_ps(fix2,tx);
1870 fiy2 = _mm256_add_ps(fiy2,ty);
1871 fiz2 = _mm256_add_ps(fiz2,tz);
1873 fjx3 = _mm256_add_ps(fjx3,tx);
1874 fjy3 = _mm256_add_ps(fjy3,ty);
1875 fjz3 = _mm256_add_ps(fjz3,tz);
1877 /**************************
1878 * CALCULATE INTERACTIONS *
1879 **************************/
1881 r31 = _mm256_mul_ps(rsq31,rinv31);
1883 /* EWALD ELECTROSTATICS */
1885 /* Analytical PME correction */
1886 zeta2 = _mm256_mul_ps(beta2,rsq31);
1887 rinv3 = _mm256_mul_ps(rinvsq31,rinv31);
1888 pmecorrF = avx256_pmecorrF_f(zeta2);
1889 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1890 felec = _mm256_mul_ps(qq31,felec);
1894 /* Calculate temporary vectorial force */
1895 tx = _mm256_mul_ps(fscal,dx31);
1896 ty = _mm256_mul_ps(fscal,dy31);
1897 tz = _mm256_mul_ps(fscal,dz31);
1899 /* Update vectorial force */
1900 fix3 = _mm256_add_ps(fix3,tx);
1901 fiy3 = _mm256_add_ps(fiy3,ty);
1902 fiz3 = _mm256_add_ps(fiz3,tz);
1904 fjx1 = _mm256_add_ps(fjx1,tx);
1905 fjy1 = _mm256_add_ps(fjy1,ty);
1906 fjz1 = _mm256_add_ps(fjz1,tz);
1908 /**************************
1909 * CALCULATE INTERACTIONS *
1910 **************************/
1912 r32 = _mm256_mul_ps(rsq32,rinv32);
1914 /* EWALD ELECTROSTATICS */
1916 /* Analytical PME correction */
1917 zeta2 = _mm256_mul_ps(beta2,rsq32);
1918 rinv3 = _mm256_mul_ps(rinvsq32,rinv32);
1919 pmecorrF = avx256_pmecorrF_f(zeta2);
1920 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1921 felec = _mm256_mul_ps(qq32,felec);
1925 /* Calculate temporary vectorial force */
1926 tx = _mm256_mul_ps(fscal,dx32);
1927 ty = _mm256_mul_ps(fscal,dy32);
1928 tz = _mm256_mul_ps(fscal,dz32);
1930 /* Update vectorial force */
1931 fix3 = _mm256_add_ps(fix3,tx);
1932 fiy3 = _mm256_add_ps(fiy3,ty);
1933 fiz3 = _mm256_add_ps(fiz3,tz);
1935 fjx2 = _mm256_add_ps(fjx2,tx);
1936 fjy2 = _mm256_add_ps(fjy2,ty);
1937 fjz2 = _mm256_add_ps(fjz2,tz);
1939 /**************************
1940 * CALCULATE INTERACTIONS *
1941 **************************/
1943 r33 = _mm256_mul_ps(rsq33,rinv33);
1945 /* EWALD ELECTROSTATICS */
1947 /* Analytical PME correction */
1948 zeta2 = _mm256_mul_ps(beta2,rsq33);
1949 rinv3 = _mm256_mul_ps(rinvsq33,rinv33);
1950 pmecorrF = avx256_pmecorrF_f(zeta2);
1951 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
1952 felec = _mm256_mul_ps(qq33,felec);
1956 /* Calculate temporary vectorial force */
1957 tx = _mm256_mul_ps(fscal,dx33);
1958 ty = _mm256_mul_ps(fscal,dy33);
1959 tz = _mm256_mul_ps(fscal,dz33);
1961 /* Update vectorial force */
1962 fix3 = _mm256_add_ps(fix3,tx);
1963 fiy3 = _mm256_add_ps(fiy3,ty);
1964 fiz3 = _mm256_add_ps(fiz3,tz);
1966 fjx3 = _mm256_add_ps(fjx3,tx);
1967 fjy3 = _mm256_add_ps(fjy3,ty);
1968 fjz3 = _mm256_add_ps(fjz3,tz);
1970 fjptrA = f+j_coord_offsetA;
1971 fjptrB = f+j_coord_offsetB;
1972 fjptrC = f+j_coord_offsetC;
1973 fjptrD = f+j_coord_offsetD;
1974 fjptrE = f+j_coord_offsetE;
1975 fjptrF = f+j_coord_offsetF;
1976 fjptrG = f+j_coord_offsetG;
1977 fjptrH = f+j_coord_offsetH;
1979 gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
1980 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1981 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1983 /* Inner loop uses 553 flops */
1986 if(jidx<j_index_end)
1989 /* Get j neighbor index, and coordinate index */
1990 jnrlistA = jjnr[jidx];
1991 jnrlistB = jjnr[jidx+1];
1992 jnrlistC = jjnr[jidx+2];
1993 jnrlistD = jjnr[jidx+3];
1994 jnrlistE = jjnr[jidx+4];
1995 jnrlistF = jjnr[jidx+5];
1996 jnrlistG = jjnr[jidx+6];
1997 jnrlistH = jjnr[jidx+7];
1998 /* Sign of each element will be negative for non-real atoms.
1999 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
2000 * so use it as val = _mm_andnot_ps(mask,val) to clear dummy entries.
2002 dummy_mask = gmx_mm256_set_m128(gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx+4)),_mm_setzero_si128())),
2003 gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128())));
2005 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
2006 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
2007 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
2008 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
2009 jnrE = (jnrlistE>=0) ? jnrlistE : 0;
2010 jnrF = (jnrlistF>=0) ? jnrlistF : 0;
2011 jnrG = (jnrlistG>=0) ? jnrlistG : 0;
2012 jnrH = (jnrlistH>=0) ? jnrlistH : 0;
2013 j_coord_offsetA = DIM*jnrA;
2014 j_coord_offsetB = DIM*jnrB;
2015 j_coord_offsetC = DIM*jnrC;
2016 j_coord_offsetD = DIM*jnrD;
2017 j_coord_offsetE = DIM*jnrE;
2018 j_coord_offsetF = DIM*jnrF;
2019 j_coord_offsetG = DIM*jnrG;
2020 j_coord_offsetH = DIM*jnrH;
2022 /* load j atom coordinates */
2023 gmx_mm256_load_4rvec_8ptr_swizzle_ps(x+j_coord_offsetA,x+j_coord_offsetB,
2024 x+j_coord_offsetC,x+j_coord_offsetD,
2025 x+j_coord_offsetE,x+j_coord_offsetF,
2026 x+j_coord_offsetG,x+j_coord_offsetH,
2027 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
2028 &jy2,&jz2,&jx3,&jy3,&jz3);
2030 /* Calculate displacement vector */
2031 dx00 = _mm256_sub_ps(ix0,jx0);
2032 dy00 = _mm256_sub_ps(iy0,jy0);
2033 dz00 = _mm256_sub_ps(iz0,jz0);
2034 dx11 = _mm256_sub_ps(ix1,jx1);
2035 dy11 = _mm256_sub_ps(iy1,jy1);
2036 dz11 = _mm256_sub_ps(iz1,jz1);
2037 dx12 = _mm256_sub_ps(ix1,jx2);
2038 dy12 = _mm256_sub_ps(iy1,jy2);
2039 dz12 = _mm256_sub_ps(iz1,jz2);
2040 dx13 = _mm256_sub_ps(ix1,jx3);
2041 dy13 = _mm256_sub_ps(iy1,jy3);
2042 dz13 = _mm256_sub_ps(iz1,jz3);
2043 dx21 = _mm256_sub_ps(ix2,jx1);
2044 dy21 = _mm256_sub_ps(iy2,jy1);
2045 dz21 = _mm256_sub_ps(iz2,jz1);
2046 dx22 = _mm256_sub_ps(ix2,jx2);
2047 dy22 = _mm256_sub_ps(iy2,jy2);
2048 dz22 = _mm256_sub_ps(iz2,jz2);
2049 dx23 = _mm256_sub_ps(ix2,jx3);
2050 dy23 = _mm256_sub_ps(iy2,jy3);
2051 dz23 = _mm256_sub_ps(iz2,jz3);
2052 dx31 = _mm256_sub_ps(ix3,jx1);
2053 dy31 = _mm256_sub_ps(iy3,jy1);
2054 dz31 = _mm256_sub_ps(iz3,jz1);
2055 dx32 = _mm256_sub_ps(ix3,jx2);
2056 dy32 = _mm256_sub_ps(iy3,jy2);
2057 dz32 = _mm256_sub_ps(iz3,jz2);
2058 dx33 = _mm256_sub_ps(ix3,jx3);
2059 dy33 = _mm256_sub_ps(iy3,jy3);
2060 dz33 = _mm256_sub_ps(iz3,jz3);
2062 /* Calculate squared distance and things based on it */
2063 rsq00 = gmx_mm256_calc_rsq_ps(dx00,dy00,dz00);
2064 rsq11 = gmx_mm256_calc_rsq_ps(dx11,dy11,dz11);
2065 rsq12 = gmx_mm256_calc_rsq_ps(dx12,dy12,dz12);
2066 rsq13 = gmx_mm256_calc_rsq_ps(dx13,dy13,dz13);
2067 rsq21 = gmx_mm256_calc_rsq_ps(dx21,dy21,dz21);
2068 rsq22 = gmx_mm256_calc_rsq_ps(dx22,dy22,dz22);
2069 rsq23 = gmx_mm256_calc_rsq_ps(dx23,dy23,dz23);
2070 rsq31 = gmx_mm256_calc_rsq_ps(dx31,dy31,dz31);
2071 rsq32 = gmx_mm256_calc_rsq_ps(dx32,dy32,dz32);
2072 rsq33 = gmx_mm256_calc_rsq_ps(dx33,dy33,dz33);
2074 rinv00 = avx256_invsqrt_f(rsq00);
2075 rinv11 = avx256_invsqrt_f(rsq11);
2076 rinv12 = avx256_invsqrt_f(rsq12);
2077 rinv13 = avx256_invsqrt_f(rsq13);
2078 rinv21 = avx256_invsqrt_f(rsq21);
2079 rinv22 = avx256_invsqrt_f(rsq22);
2080 rinv23 = avx256_invsqrt_f(rsq23);
2081 rinv31 = avx256_invsqrt_f(rsq31);
2082 rinv32 = avx256_invsqrt_f(rsq32);
2083 rinv33 = avx256_invsqrt_f(rsq33);
2085 rinvsq00 = _mm256_mul_ps(rinv00,rinv00);
2086 rinvsq11 = _mm256_mul_ps(rinv11,rinv11);
2087 rinvsq12 = _mm256_mul_ps(rinv12,rinv12);
2088 rinvsq13 = _mm256_mul_ps(rinv13,rinv13);
2089 rinvsq21 = _mm256_mul_ps(rinv21,rinv21);
2090 rinvsq22 = _mm256_mul_ps(rinv22,rinv22);
2091 rinvsq23 = _mm256_mul_ps(rinv23,rinv23);
2092 rinvsq31 = _mm256_mul_ps(rinv31,rinv31);
2093 rinvsq32 = _mm256_mul_ps(rinv32,rinv32);
2094 rinvsq33 = _mm256_mul_ps(rinv33,rinv33);
2096 fjx0 = _mm256_setzero_ps();
2097 fjy0 = _mm256_setzero_ps();
2098 fjz0 = _mm256_setzero_ps();
2099 fjx1 = _mm256_setzero_ps();
2100 fjy1 = _mm256_setzero_ps();
2101 fjz1 = _mm256_setzero_ps();
2102 fjx2 = _mm256_setzero_ps();
2103 fjy2 = _mm256_setzero_ps();
2104 fjz2 = _mm256_setzero_ps();
2105 fjx3 = _mm256_setzero_ps();
2106 fjy3 = _mm256_setzero_ps();
2107 fjz3 = _mm256_setzero_ps();
2109 /**************************
2110 * CALCULATE INTERACTIONS *
2111 **************************/
2113 r00 = _mm256_mul_ps(rsq00,rinv00);
2114 r00 = _mm256_andnot_ps(dummy_mask,r00);
2116 /* Analytical LJ-PME */
2117 rinvsix = _mm256_mul_ps(_mm256_mul_ps(rinvsq00,rinvsq00),rinvsq00);
2118 ewcljrsq = _mm256_mul_ps(ewclj2,rsq00);
2119 ewclj6 = _mm256_mul_ps(ewclj2,_mm256_mul_ps(ewclj2,ewclj2));
2120 exponent = avx256_exp_f(ewcljrsq);
2121 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
2122 poly = _mm256_mul_ps(exponent,_mm256_add_ps(_mm256_sub_ps(one,ewcljrsq),_mm256_mul_ps(_mm256_mul_ps(ewcljrsq,ewcljrsq),one_half)));
2123 /* f6A = 6 * C6grid * (1 - poly) */
2124 f6A = _mm256_mul_ps(c6grid_00,_mm256_sub_ps(one,poly));
2125 /* f6B = C6grid * exponent * beta^6 */
2126 f6B = _mm256_mul_ps(_mm256_mul_ps(c6grid_00,one_sixth),_mm256_mul_ps(exponent,ewclj6));
2127 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
2128 fvdw = _mm256_mul_ps(_mm256_add_ps(_mm256_mul_ps(_mm256_sub_ps(_mm256_mul_ps(c12_00,rinvsix),_mm256_sub_ps(c6_00,f6A)),rinvsix),f6B),rinvsq00);
2132 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2134 /* Calculate temporary vectorial force */
2135 tx = _mm256_mul_ps(fscal,dx00);
2136 ty = _mm256_mul_ps(fscal,dy00);
2137 tz = _mm256_mul_ps(fscal,dz00);
2139 /* Update vectorial force */
2140 fix0 = _mm256_add_ps(fix0,tx);
2141 fiy0 = _mm256_add_ps(fiy0,ty);
2142 fiz0 = _mm256_add_ps(fiz0,tz);
2144 fjx0 = _mm256_add_ps(fjx0,tx);
2145 fjy0 = _mm256_add_ps(fjy0,ty);
2146 fjz0 = _mm256_add_ps(fjz0,tz);
2148 /**************************
2149 * CALCULATE INTERACTIONS *
2150 **************************/
2152 r11 = _mm256_mul_ps(rsq11,rinv11);
2153 r11 = _mm256_andnot_ps(dummy_mask,r11);
2155 /* EWALD ELECTROSTATICS */
2157 /* Analytical PME correction */
2158 zeta2 = _mm256_mul_ps(beta2,rsq11);
2159 rinv3 = _mm256_mul_ps(rinvsq11,rinv11);
2160 pmecorrF = avx256_pmecorrF_f(zeta2);
2161 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2162 felec = _mm256_mul_ps(qq11,felec);
2166 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2168 /* Calculate temporary vectorial force */
2169 tx = _mm256_mul_ps(fscal,dx11);
2170 ty = _mm256_mul_ps(fscal,dy11);
2171 tz = _mm256_mul_ps(fscal,dz11);
2173 /* Update vectorial force */
2174 fix1 = _mm256_add_ps(fix1,tx);
2175 fiy1 = _mm256_add_ps(fiy1,ty);
2176 fiz1 = _mm256_add_ps(fiz1,tz);
2178 fjx1 = _mm256_add_ps(fjx1,tx);
2179 fjy1 = _mm256_add_ps(fjy1,ty);
2180 fjz1 = _mm256_add_ps(fjz1,tz);
2182 /**************************
2183 * CALCULATE INTERACTIONS *
2184 **************************/
2186 r12 = _mm256_mul_ps(rsq12,rinv12);
2187 r12 = _mm256_andnot_ps(dummy_mask,r12);
2189 /* EWALD ELECTROSTATICS */
2191 /* Analytical PME correction */
2192 zeta2 = _mm256_mul_ps(beta2,rsq12);
2193 rinv3 = _mm256_mul_ps(rinvsq12,rinv12);
2194 pmecorrF = avx256_pmecorrF_f(zeta2);
2195 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2196 felec = _mm256_mul_ps(qq12,felec);
2200 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2202 /* Calculate temporary vectorial force */
2203 tx = _mm256_mul_ps(fscal,dx12);
2204 ty = _mm256_mul_ps(fscal,dy12);
2205 tz = _mm256_mul_ps(fscal,dz12);
2207 /* Update vectorial force */
2208 fix1 = _mm256_add_ps(fix1,tx);
2209 fiy1 = _mm256_add_ps(fiy1,ty);
2210 fiz1 = _mm256_add_ps(fiz1,tz);
2212 fjx2 = _mm256_add_ps(fjx2,tx);
2213 fjy2 = _mm256_add_ps(fjy2,ty);
2214 fjz2 = _mm256_add_ps(fjz2,tz);
2216 /**************************
2217 * CALCULATE INTERACTIONS *
2218 **************************/
2220 r13 = _mm256_mul_ps(rsq13,rinv13);
2221 r13 = _mm256_andnot_ps(dummy_mask,r13);
2223 /* EWALD ELECTROSTATICS */
2225 /* Analytical PME correction */
2226 zeta2 = _mm256_mul_ps(beta2,rsq13);
2227 rinv3 = _mm256_mul_ps(rinvsq13,rinv13);
2228 pmecorrF = avx256_pmecorrF_f(zeta2);
2229 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2230 felec = _mm256_mul_ps(qq13,felec);
2234 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2236 /* Calculate temporary vectorial force */
2237 tx = _mm256_mul_ps(fscal,dx13);
2238 ty = _mm256_mul_ps(fscal,dy13);
2239 tz = _mm256_mul_ps(fscal,dz13);
2241 /* Update vectorial force */
2242 fix1 = _mm256_add_ps(fix1,tx);
2243 fiy1 = _mm256_add_ps(fiy1,ty);
2244 fiz1 = _mm256_add_ps(fiz1,tz);
2246 fjx3 = _mm256_add_ps(fjx3,tx);
2247 fjy3 = _mm256_add_ps(fjy3,ty);
2248 fjz3 = _mm256_add_ps(fjz3,tz);
2250 /**************************
2251 * CALCULATE INTERACTIONS *
2252 **************************/
2254 r21 = _mm256_mul_ps(rsq21,rinv21);
2255 r21 = _mm256_andnot_ps(dummy_mask,r21);
2257 /* EWALD ELECTROSTATICS */
2259 /* Analytical PME correction */
2260 zeta2 = _mm256_mul_ps(beta2,rsq21);
2261 rinv3 = _mm256_mul_ps(rinvsq21,rinv21);
2262 pmecorrF = avx256_pmecorrF_f(zeta2);
2263 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2264 felec = _mm256_mul_ps(qq21,felec);
2268 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2270 /* Calculate temporary vectorial force */
2271 tx = _mm256_mul_ps(fscal,dx21);
2272 ty = _mm256_mul_ps(fscal,dy21);
2273 tz = _mm256_mul_ps(fscal,dz21);
2275 /* Update vectorial force */
2276 fix2 = _mm256_add_ps(fix2,tx);
2277 fiy2 = _mm256_add_ps(fiy2,ty);
2278 fiz2 = _mm256_add_ps(fiz2,tz);
2280 fjx1 = _mm256_add_ps(fjx1,tx);
2281 fjy1 = _mm256_add_ps(fjy1,ty);
2282 fjz1 = _mm256_add_ps(fjz1,tz);
2284 /**************************
2285 * CALCULATE INTERACTIONS *
2286 **************************/
2288 r22 = _mm256_mul_ps(rsq22,rinv22);
2289 r22 = _mm256_andnot_ps(dummy_mask,r22);
2291 /* EWALD ELECTROSTATICS */
2293 /* Analytical PME correction */
2294 zeta2 = _mm256_mul_ps(beta2,rsq22);
2295 rinv3 = _mm256_mul_ps(rinvsq22,rinv22);
2296 pmecorrF = avx256_pmecorrF_f(zeta2);
2297 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2298 felec = _mm256_mul_ps(qq22,felec);
2302 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2304 /* Calculate temporary vectorial force */
2305 tx = _mm256_mul_ps(fscal,dx22);
2306 ty = _mm256_mul_ps(fscal,dy22);
2307 tz = _mm256_mul_ps(fscal,dz22);
2309 /* Update vectorial force */
2310 fix2 = _mm256_add_ps(fix2,tx);
2311 fiy2 = _mm256_add_ps(fiy2,ty);
2312 fiz2 = _mm256_add_ps(fiz2,tz);
2314 fjx2 = _mm256_add_ps(fjx2,tx);
2315 fjy2 = _mm256_add_ps(fjy2,ty);
2316 fjz2 = _mm256_add_ps(fjz2,tz);
2318 /**************************
2319 * CALCULATE INTERACTIONS *
2320 **************************/
2322 r23 = _mm256_mul_ps(rsq23,rinv23);
2323 r23 = _mm256_andnot_ps(dummy_mask,r23);
2325 /* EWALD ELECTROSTATICS */
2327 /* Analytical PME correction */
2328 zeta2 = _mm256_mul_ps(beta2,rsq23);
2329 rinv3 = _mm256_mul_ps(rinvsq23,rinv23);
2330 pmecorrF = avx256_pmecorrF_f(zeta2);
2331 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2332 felec = _mm256_mul_ps(qq23,felec);
2336 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2338 /* Calculate temporary vectorial force */
2339 tx = _mm256_mul_ps(fscal,dx23);
2340 ty = _mm256_mul_ps(fscal,dy23);
2341 tz = _mm256_mul_ps(fscal,dz23);
2343 /* Update vectorial force */
2344 fix2 = _mm256_add_ps(fix2,tx);
2345 fiy2 = _mm256_add_ps(fiy2,ty);
2346 fiz2 = _mm256_add_ps(fiz2,tz);
2348 fjx3 = _mm256_add_ps(fjx3,tx);
2349 fjy3 = _mm256_add_ps(fjy3,ty);
2350 fjz3 = _mm256_add_ps(fjz3,tz);
2352 /**************************
2353 * CALCULATE INTERACTIONS *
2354 **************************/
2356 r31 = _mm256_mul_ps(rsq31,rinv31);
2357 r31 = _mm256_andnot_ps(dummy_mask,r31);
2359 /* EWALD ELECTROSTATICS */
2361 /* Analytical PME correction */
2362 zeta2 = _mm256_mul_ps(beta2,rsq31);
2363 rinv3 = _mm256_mul_ps(rinvsq31,rinv31);
2364 pmecorrF = avx256_pmecorrF_f(zeta2);
2365 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2366 felec = _mm256_mul_ps(qq31,felec);
2370 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2372 /* Calculate temporary vectorial force */
2373 tx = _mm256_mul_ps(fscal,dx31);
2374 ty = _mm256_mul_ps(fscal,dy31);
2375 tz = _mm256_mul_ps(fscal,dz31);
2377 /* Update vectorial force */
2378 fix3 = _mm256_add_ps(fix3,tx);
2379 fiy3 = _mm256_add_ps(fiy3,ty);
2380 fiz3 = _mm256_add_ps(fiz3,tz);
2382 fjx1 = _mm256_add_ps(fjx1,tx);
2383 fjy1 = _mm256_add_ps(fjy1,ty);
2384 fjz1 = _mm256_add_ps(fjz1,tz);
2386 /**************************
2387 * CALCULATE INTERACTIONS *
2388 **************************/
2390 r32 = _mm256_mul_ps(rsq32,rinv32);
2391 r32 = _mm256_andnot_ps(dummy_mask,r32);
2393 /* EWALD ELECTROSTATICS */
2395 /* Analytical PME correction */
2396 zeta2 = _mm256_mul_ps(beta2,rsq32);
2397 rinv3 = _mm256_mul_ps(rinvsq32,rinv32);
2398 pmecorrF = avx256_pmecorrF_f(zeta2);
2399 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2400 felec = _mm256_mul_ps(qq32,felec);
2404 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2406 /* Calculate temporary vectorial force */
2407 tx = _mm256_mul_ps(fscal,dx32);
2408 ty = _mm256_mul_ps(fscal,dy32);
2409 tz = _mm256_mul_ps(fscal,dz32);
2411 /* Update vectorial force */
2412 fix3 = _mm256_add_ps(fix3,tx);
2413 fiy3 = _mm256_add_ps(fiy3,ty);
2414 fiz3 = _mm256_add_ps(fiz3,tz);
2416 fjx2 = _mm256_add_ps(fjx2,tx);
2417 fjy2 = _mm256_add_ps(fjy2,ty);
2418 fjz2 = _mm256_add_ps(fjz2,tz);
2420 /**************************
2421 * CALCULATE INTERACTIONS *
2422 **************************/
2424 r33 = _mm256_mul_ps(rsq33,rinv33);
2425 r33 = _mm256_andnot_ps(dummy_mask,r33);
2427 /* EWALD ELECTROSTATICS */
2429 /* Analytical PME correction */
2430 zeta2 = _mm256_mul_ps(beta2,rsq33);
2431 rinv3 = _mm256_mul_ps(rinvsq33,rinv33);
2432 pmecorrF = avx256_pmecorrF_f(zeta2);
2433 felec = _mm256_add_ps( _mm256_mul_ps(pmecorrF,beta3), rinv3);
2434 felec = _mm256_mul_ps(qq33,felec);
2438 fscal = _mm256_andnot_ps(dummy_mask,fscal);
2440 /* Calculate temporary vectorial force */
2441 tx = _mm256_mul_ps(fscal,dx33);
2442 ty = _mm256_mul_ps(fscal,dy33);
2443 tz = _mm256_mul_ps(fscal,dz33);
2445 /* Update vectorial force */
2446 fix3 = _mm256_add_ps(fix3,tx);
2447 fiy3 = _mm256_add_ps(fiy3,ty);
2448 fiz3 = _mm256_add_ps(fiz3,tz);
2450 fjx3 = _mm256_add_ps(fjx3,tx);
2451 fjy3 = _mm256_add_ps(fjy3,ty);
2452 fjz3 = _mm256_add_ps(fjz3,tz);
2454 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
2455 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
2456 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
2457 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
2458 fjptrE = (jnrlistE>=0) ? f+j_coord_offsetE : scratch;
2459 fjptrF = (jnrlistF>=0) ? f+j_coord_offsetF : scratch;
2460 fjptrG = (jnrlistG>=0) ? f+j_coord_offsetG : scratch;
2461 fjptrH = (jnrlistH>=0) ? f+j_coord_offsetH : scratch;
2463 gmx_mm256_decrement_4rvec_8ptr_swizzle_ps(fjptrA,fjptrB,fjptrC,fjptrD,fjptrE,fjptrF,fjptrG,fjptrH,
2464 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
2465 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2467 /* Inner loop uses 563 flops */
2470 /* End of innermost loop */
2472 gmx_mm256_update_iforce_4atom_swizzle_ps(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
2473 f+i_coord_offset,fshift+i_shift_offset);
2475 /* Increment number of inner iterations */
2476 inneriter += j_index_end - j_index_start;
2478 /* Outer loop uses 24 flops */
2481 /* Increment number of outer iterations */
2484 /* Update outer/inner flops */
2486 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*563);