2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
5 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
6 * and including many others, as listed in the AUTHORS file in the
7 * top-level source directory and at http://www.gromacs.org.
9 * GROMACS is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public License
11 * as published by the Free Software Foundation; either version 2.1
12 * of the License, or (at your option) any later version.
14 * GROMACS is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with GROMACS; if not, see
21 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
22 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 * If you want to redistribute modifications to GROMACS, please
25 * consider that scientific software is very special. Version
26 * control is crucial - bugs must be traceable. We will be happy to
27 * consider code for inclusion in the official distribution, but
28 * derived work must not be called official GROMACS. Details are found
29 * in the README & COPYING files - if they are missing, get the
30 * official version at http://www.gromacs.org.
32 * To help us fund GROMACS development, we humbly ask that you cite
33 * the research papers on the package. Check out http://www.gromacs.org.
36 * Note: this file was generated by the GROMACS avx_256_double kernel generator.
42 #include "../nb_kernel.h"
43 #include "types/simple.h"
44 #include "gromacs/math/vec.h"
47 #include "gromacs/simd/math_x86_avx_256_double.h"
48 #include "kernelutil_x86_avx_256_double.h"
51 * Gromacs nonbonded kernel: nb_kernel_ElecRF_VdwLJ_GeomW4W4_VF_avx_256_double
52 * Electrostatics interaction: ReactionField
53 * VdW interaction: LennardJones
54 * Geometry: Water4-Water4
55 * Calculate force/pot: PotentialAndForce
58 nb_kernel_ElecRF_VdwLJ_GeomW4W4_VF_avx_256_double
59 (t_nblist * gmx_restrict nlist,
60 rvec * gmx_restrict xx,
61 rvec * gmx_restrict ff,
62 t_forcerec * gmx_restrict fr,
63 t_mdatoms * gmx_restrict mdatoms,
64 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
65 t_nrnb * gmx_restrict nrnb)
67 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
68 * just 0 for non-waters.
69 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
70 * jnr indices corresponding to data put in the four positions in the SIMD register.
72 int i_shift_offset,i_coord_offset,outeriter,inneriter;
73 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
74 int jnrA,jnrB,jnrC,jnrD;
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 *iinr,*jindex,*jjnr,*shiftidx,*gid;
80 real *shiftvec,*fshift,*x,*f;
81 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
83 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
84 real * vdwioffsetptr0;
85 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
86 real * vdwioffsetptr1;
87 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
88 real * vdwioffsetptr2;
89 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
90 real * vdwioffsetptr3;
91 __m256d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
92 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
93 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
94 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
95 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
96 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
97 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
98 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
99 __m256d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
100 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
101 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
102 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
103 __m256d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
104 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
105 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
106 __m256d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
107 __m256d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
108 __m256d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
109 __m256d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
110 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
113 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
116 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
117 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
118 __m256d dummy_mask,cutoff_mask;
119 __m128 tmpmask0,tmpmask1;
120 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
121 __m256d one = _mm256_set1_pd(1.0);
122 __m256d two = _mm256_set1_pd(2.0);
128 jindex = nlist->jindex;
130 shiftidx = nlist->shift;
132 shiftvec = fr->shift_vec[0];
133 fshift = fr->fshift[0];
134 facel = _mm256_set1_pd(fr->epsfac);
135 charge = mdatoms->chargeA;
136 krf = _mm256_set1_pd(fr->ic->k_rf);
137 krf2 = _mm256_set1_pd(fr->ic->k_rf*2.0);
138 crf = _mm256_set1_pd(fr->ic->c_rf);
139 nvdwtype = fr->ntype;
141 vdwtype = mdatoms->typeA;
143 /* Setup water-specific parameters */
144 inr = nlist->iinr[0];
145 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
146 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
147 iq3 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
148 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
150 jq1 = _mm256_set1_pd(charge[inr+1]);
151 jq2 = _mm256_set1_pd(charge[inr+2]);
152 jq3 = _mm256_set1_pd(charge[inr+3]);
153 vdwjidx0A = 2*vdwtype[inr+0];
154 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
155 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
156 qq11 = _mm256_mul_pd(iq1,jq1);
157 qq12 = _mm256_mul_pd(iq1,jq2);
158 qq13 = _mm256_mul_pd(iq1,jq3);
159 qq21 = _mm256_mul_pd(iq2,jq1);
160 qq22 = _mm256_mul_pd(iq2,jq2);
161 qq23 = _mm256_mul_pd(iq2,jq3);
162 qq31 = _mm256_mul_pd(iq3,jq1);
163 qq32 = _mm256_mul_pd(iq3,jq2);
164 qq33 = _mm256_mul_pd(iq3,jq3);
166 /* Avoid stupid compiler warnings */
167 jnrA = jnrB = jnrC = jnrD = 0;
176 for(iidx=0;iidx<4*DIM;iidx++)
181 /* Start outer loop over neighborlists */
182 for(iidx=0; iidx<nri; iidx++)
184 /* Load shift vector for this list */
185 i_shift_offset = DIM*shiftidx[iidx];
187 /* Load limits for loop over neighbors */
188 j_index_start = jindex[iidx];
189 j_index_end = jindex[iidx+1];
191 /* Get outer coordinate index */
193 i_coord_offset = DIM*inr;
195 /* Load i particle coords and add shift vector */
196 gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
197 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
199 fix0 = _mm256_setzero_pd();
200 fiy0 = _mm256_setzero_pd();
201 fiz0 = _mm256_setzero_pd();
202 fix1 = _mm256_setzero_pd();
203 fiy1 = _mm256_setzero_pd();
204 fiz1 = _mm256_setzero_pd();
205 fix2 = _mm256_setzero_pd();
206 fiy2 = _mm256_setzero_pd();
207 fiz2 = _mm256_setzero_pd();
208 fix3 = _mm256_setzero_pd();
209 fiy3 = _mm256_setzero_pd();
210 fiz3 = _mm256_setzero_pd();
212 /* Reset potential sums */
213 velecsum = _mm256_setzero_pd();
214 vvdwsum = _mm256_setzero_pd();
216 /* Start inner kernel loop */
217 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
220 /* Get j neighbor index, and coordinate index */
225 j_coord_offsetA = DIM*jnrA;
226 j_coord_offsetB = DIM*jnrB;
227 j_coord_offsetC = DIM*jnrC;
228 j_coord_offsetD = DIM*jnrD;
230 /* load j atom coordinates */
231 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
232 x+j_coord_offsetC,x+j_coord_offsetD,
233 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
234 &jy2,&jz2,&jx3,&jy3,&jz3);
236 /* Calculate displacement vector */
237 dx00 = _mm256_sub_pd(ix0,jx0);
238 dy00 = _mm256_sub_pd(iy0,jy0);
239 dz00 = _mm256_sub_pd(iz0,jz0);
240 dx11 = _mm256_sub_pd(ix1,jx1);
241 dy11 = _mm256_sub_pd(iy1,jy1);
242 dz11 = _mm256_sub_pd(iz1,jz1);
243 dx12 = _mm256_sub_pd(ix1,jx2);
244 dy12 = _mm256_sub_pd(iy1,jy2);
245 dz12 = _mm256_sub_pd(iz1,jz2);
246 dx13 = _mm256_sub_pd(ix1,jx3);
247 dy13 = _mm256_sub_pd(iy1,jy3);
248 dz13 = _mm256_sub_pd(iz1,jz3);
249 dx21 = _mm256_sub_pd(ix2,jx1);
250 dy21 = _mm256_sub_pd(iy2,jy1);
251 dz21 = _mm256_sub_pd(iz2,jz1);
252 dx22 = _mm256_sub_pd(ix2,jx2);
253 dy22 = _mm256_sub_pd(iy2,jy2);
254 dz22 = _mm256_sub_pd(iz2,jz2);
255 dx23 = _mm256_sub_pd(ix2,jx3);
256 dy23 = _mm256_sub_pd(iy2,jy3);
257 dz23 = _mm256_sub_pd(iz2,jz3);
258 dx31 = _mm256_sub_pd(ix3,jx1);
259 dy31 = _mm256_sub_pd(iy3,jy1);
260 dz31 = _mm256_sub_pd(iz3,jz1);
261 dx32 = _mm256_sub_pd(ix3,jx2);
262 dy32 = _mm256_sub_pd(iy3,jy2);
263 dz32 = _mm256_sub_pd(iz3,jz2);
264 dx33 = _mm256_sub_pd(ix3,jx3);
265 dy33 = _mm256_sub_pd(iy3,jy3);
266 dz33 = _mm256_sub_pd(iz3,jz3);
268 /* Calculate squared distance and things based on it */
269 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
270 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
271 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
272 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
273 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
274 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
275 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
276 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
277 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
278 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
280 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
281 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
282 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
283 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
284 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
285 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
286 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
287 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
288 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
290 rinvsq00 = gmx_mm256_inv_pd(rsq00);
291 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
292 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
293 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
294 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
295 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
296 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
297 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
298 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
299 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
301 fjx0 = _mm256_setzero_pd();
302 fjy0 = _mm256_setzero_pd();
303 fjz0 = _mm256_setzero_pd();
304 fjx1 = _mm256_setzero_pd();
305 fjy1 = _mm256_setzero_pd();
306 fjz1 = _mm256_setzero_pd();
307 fjx2 = _mm256_setzero_pd();
308 fjy2 = _mm256_setzero_pd();
309 fjz2 = _mm256_setzero_pd();
310 fjx3 = _mm256_setzero_pd();
311 fjy3 = _mm256_setzero_pd();
312 fjz3 = _mm256_setzero_pd();
314 /**************************
315 * CALCULATE INTERACTIONS *
316 **************************/
318 /* LENNARD-JONES DISPERSION/REPULSION */
320 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
321 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
322 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
323 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
324 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
326 /* Update potential sum for this i atom from the interaction with this j atom. */
327 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
331 /* Calculate temporary vectorial force */
332 tx = _mm256_mul_pd(fscal,dx00);
333 ty = _mm256_mul_pd(fscal,dy00);
334 tz = _mm256_mul_pd(fscal,dz00);
336 /* Update vectorial force */
337 fix0 = _mm256_add_pd(fix0,tx);
338 fiy0 = _mm256_add_pd(fiy0,ty);
339 fiz0 = _mm256_add_pd(fiz0,tz);
341 fjx0 = _mm256_add_pd(fjx0,tx);
342 fjy0 = _mm256_add_pd(fjy0,ty);
343 fjz0 = _mm256_add_pd(fjz0,tz);
345 /**************************
346 * CALCULATE INTERACTIONS *
347 **************************/
349 /* REACTION-FIELD ELECTROSTATICS */
350 velec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_add_pd(rinv11,_mm256_mul_pd(krf,rsq11)),crf));
351 felec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_mul_pd(rinv11,rinvsq11),krf2));
353 /* Update potential sum for this i atom from the interaction with this j atom. */
354 velecsum = _mm256_add_pd(velecsum,velec);
358 /* Calculate temporary vectorial force */
359 tx = _mm256_mul_pd(fscal,dx11);
360 ty = _mm256_mul_pd(fscal,dy11);
361 tz = _mm256_mul_pd(fscal,dz11);
363 /* Update vectorial force */
364 fix1 = _mm256_add_pd(fix1,tx);
365 fiy1 = _mm256_add_pd(fiy1,ty);
366 fiz1 = _mm256_add_pd(fiz1,tz);
368 fjx1 = _mm256_add_pd(fjx1,tx);
369 fjy1 = _mm256_add_pd(fjy1,ty);
370 fjz1 = _mm256_add_pd(fjz1,tz);
372 /**************************
373 * CALCULATE INTERACTIONS *
374 **************************/
376 /* REACTION-FIELD ELECTROSTATICS */
377 velec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_add_pd(rinv12,_mm256_mul_pd(krf,rsq12)),crf));
378 felec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_mul_pd(rinv12,rinvsq12),krf2));
380 /* Update potential sum for this i atom from the interaction with this j atom. */
381 velecsum = _mm256_add_pd(velecsum,velec);
385 /* Calculate temporary vectorial force */
386 tx = _mm256_mul_pd(fscal,dx12);
387 ty = _mm256_mul_pd(fscal,dy12);
388 tz = _mm256_mul_pd(fscal,dz12);
390 /* Update vectorial force */
391 fix1 = _mm256_add_pd(fix1,tx);
392 fiy1 = _mm256_add_pd(fiy1,ty);
393 fiz1 = _mm256_add_pd(fiz1,tz);
395 fjx2 = _mm256_add_pd(fjx2,tx);
396 fjy2 = _mm256_add_pd(fjy2,ty);
397 fjz2 = _mm256_add_pd(fjz2,tz);
399 /**************************
400 * CALCULATE INTERACTIONS *
401 **************************/
403 /* REACTION-FIELD ELECTROSTATICS */
404 velec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_add_pd(rinv13,_mm256_mul_pd(krf,rsq13)),crf));
405 felec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_mul_pd(rinv13,rinvsq13),krf2));
407 /* Update potential sum for this i atom from the interaction with this j atom. */
408 velecsum = _mm256_add_pd(velecsum,velec);
412 /* Calculate temporary vectorial force */
413 tx = _mm256_mul_pd(fscal,dx13);
414 ty = _mm256_mul_pd(fscal,dy13);
415 tz = _mm256_mul_pd(fscal,dz13);
417 /* Update vectorial force */
418 fix1 = _mm256_add_pd(fix1,tx);
419 fiy1 = _mm256_add_pd(fiy1,ty);
420 fiz1 = _mm256_add_pd(fiz1,tz);
422 fjx3 = _mm256_add_pd(fjx3,tx);
423 fjy3 = _mm256_add_pd(fjy3,ty);
424 fjz3 = _mm256_add_pd(fjz3,tz);
426 /**************************
427 * CALCULATE INTERACTIONS *
428 **************************/
430 /* REACTION-FIELD ELECTROSTATICS */
431 velec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_add_pd(rinv21,_mm256_mul_pd(krf,rsq21)),crf));
432 felec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_mul_pd(rinv21,rinvsq21),krf2));
434 /* Update potential sum for this i atom from the interaction with this j atom. */
435 velecsum = _mm256_add_pd(velecsum,velec);
439 /* Calculate temporary vectorial force */
440 tx = _mm256_mul_pd(fscal,dx21);
441 ty = _mm256_mul_pd(fscal,dy21);
442 tz = _mm256_mul_pd(fscal,dz21);
444 /* Update vectorial force */
445 fix2 = _mm256_add_pd(fix2,tx);
446 fiy2 = _mm256_add_pd(fiy2,ty);
447 fiz2 = _mm256_add_pd(fiz2,tz);
449 fjx1 = _mm256_add_pd(fjx1,tx);
450 fjy1 = _mm256_add_pd(fjy1,ty);
451 fjz1 = _mm256_add_pd(fjz1,tz);
453 /**************************
454 * CALCULATE INTERACTIONS *
455 **************************/
457 /* REACTION-FIELD ELECTROSTATICS */
458 velec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_add_pd(rinv22,_mm256_mul_pd(krf,rsq22)),crf));
459 felec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_mul_pd(rinv22,rinvsq22),krf2));
461 /* Update potential sum for this i atom from the interaction with this j atom. */
462 velecsum = _mm256_add_pd(velecsum,velec);
466 /* Calculate temporary vectorial force */
467 tx = _mm256_mul_pd(fscal,dx22);
468 ty = _mm256_mul_pd(fscal,dy22);
469 tz = _mm256_mul_pd(fscal,dz22);
471 /* Update vectorial force */
472 fix2 = _mm256_add_pd(fix2,tx);
473 fiy2 = _mm256_add_pd(fiy2,ty);
474 fiz2 = _mm256_add_pd(fiz2,tz);
476 fjx2 = _mm256_add_pd(fjx2,tx);
477 fjy2 = _mm256_add_pd(fjy2,ty);
478 fjz2 = _mm256_add_pd(fjz2,tz);
480 /**************************
481 * CALCULATE INTERACTIONS *
482 **************************/
484 /* REACTION-FIELD ELECTROSTATICS */
485 velec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_add_pd(rinv23,_mm256_mul_pd(krf,rsq23)),crf));
486 felec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_mul_pd(rinv23,rinvsq23),krf2));
488 /* Update potential sum for this i atom from the interaction with this j atom. */
489 velecsum = _mm256_add_pd(velecsum,velec);
493 /* Calculate temporary vectorial force */
494 tx = _mm256_mul_pd(fscal,dx23);
495 ty = _mm256_mul_pd(fscal,dy23);
496 tz = _mm256_mul_pd(fscal,dz23);
498 /* Update vectorial force */
499 fix2 = _mm256_add_pd(fix2,tx);
500 fiy2 = _mm256_add_pd(fiy2,ty);
501 fiz2 = _mm256_add_pd(fiz2,tz);
503 fjx3 = _mm256_add_pd(fjx3,tx);
504 fjy3 = _mm256_add_pd(fjy3,ty);
505 fjz3 = _mm256_add_pd(fjz3,tz);
507 /**************************
508 * CALCULATE INTERACTIONS *
509 **************************/
511 /* REACTION-FIELD ELECTROSTATICS */
512 velec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_add_pd(rinv31,_mm256_mul_pd(krf,rsq31)),crf));
513 felec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_mul_pd(rinv31,rinvsq31),krf2));
515 /* Update potential sum for this i atom from the interaction with this j atom. */
516 velecsum = _mm256_add_pd(velecsum,velec);
520 /* Calculate temporary vectorial force */
521 tx = _mm256_mul_pd(fscal,dx31);
522 ty = _mm256_mul_pd(fscal,dy31);
523 tz = _mm256_mul_pd(fscal,dz31);
525 /* Update vectorial force */
526 fix3 = _mm256_add_pd(fix3,tx);
527 fiy3 = _mm256_add_pd(fiy3,ty);
528 fiz3 = _mm256_add_pd(fiz3,tz);
530 fjx1 = _mm256_add_pd(fjx1,tx);
531 fjy1 = _mm256_add_pd(fjy1,ty);
532 fjz1 = _mm256_add_pd(fjz1,tz);
534 /**************************
535 * CALCULATE INTERACTIONS *
536 **************************/
538 /* REACTION-FIELD ELECTROSTATICS */
539 velec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_add_pd(rinv32,_mm256_mul_pd(krf,rsq32)),crf));
540 felec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_mul_pd(rinv32,rinvsq32),krf2));
542 /* Update potential sum for this i atom from the interaction with this j atom. */
543 velecsum = _mm256_add_pd(velecsum,velec);
547 /* Calculate temporary vectorial force */
548 tx = _mm256_mul_pd(fscal,dx32);
549 ty = _mm256_mul_pd(fscal,dy32);
550 tz = _mm256_mul_pd(fscal,dz32);
552 /* Update vectorial force */
553 fix3 = _mm256_add_pd(fix3,tx);
554 fiy3 = _mm256_add_pd(fiy3,ty);
555 fiz3 = _mm256_add_pd(fiz3,tz);
557 fjx2 = _mm256_add_pd(fjx2,tx);
558 fjy2 = _mm256_add_pd(fjy2,ty);
559 fjz2 = _mm256_add_pd(fjz2,tz);
561 /**************************
562 * CALCULATE INTERACTIONS *
563 **************************/
565 /* REACTION-FIELD ELECTROSTATICS */
566 velec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_add_pd(rinv33,_mm256_mul_pd(krf,rsq33)),crf));
567 felec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_mul_pd(rinv33,rinvsq33),krf2));
569 /* Update potential sum for this i atom from the interaction with this j atom. */
570 velecsum = _mm256_add_pd(velecsum,velec);
574 /* Calculate temporary vectorial force */
575 tx = _mm256_mul_pd(fscal,dx33);
576 ty = _mm256_mul_pd(fscal,dy33);
577 tz = _mm256_mul_pd(fscal,dz33);
579 /* Update vectorial force */
580 fix3 = _mm256_add_pd(fix3,tx);
581 fiy3 = _mm256_add_pd(fiy3,ty);
582 fiz3 = _mm256_add_pd(fiz3,tz);
584 fjx3 = _mm256_add_pd(fjx3,tx);
585 fjy3 = _mm256_add_pd(fjy3,ty);
586 fjz3 = _mm256_add_pd(fjz3,tz);
588 fjptrA = f+j_coord_offsetA;
589 fjptrB = f+j_coord_offsetB;
590 fjptrC = f+j_coord_offsetC;
591 fjptrD = f+j_coord_offsetD;
593 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
594 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
595 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
597 /* Inner loop uses 323 flops */
603 /* Get j neighbor index, and coordinate index */
604 jnrlistA = jjnr[jidx];
605 jnrlistB = jjnr[jidx+1];
606 jnrlistC = jjnr[jidx+2];
607 jnrlistD = jjnr[jidx+3];
608 /* Sign of each element will be negative for non-real atoms.
609 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
610 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
612 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
614 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
615 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
616 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
618 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
619 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
620 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
621 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
622 j_coord_offsetA = DIM*jnrA;
623 j_coord_offsetB = DIM*jnrB;
624 j_coord_offsetC = DIM*jnrC;
625 j_coord_offsetD = DIM*jnrD;
627 /* load j atom coordinates */
628 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
629 x+j_coord_offsetC,x+j_coord_offsetD,
630 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
631 &jy2,&jz2,&jx3,&jy3,&jz3);
633 /* Calculate displacement vector */
634 dx00 = _mm256_sub_pd(ix0,jx0);
635 dy00 = _mm256_sub_pd(iy0,jy0);
636 dz00 = _mm256_sub_pd(iz0,jz0);
637 dx11 = _mm256_sub_pd(ix1,jx1);
638 dy11 = _mm256_sub_pd(iy1,jy1);
639 dz11 = _mm256_sub_pd(iz1,jz1);
640 dx12 = _mm256_sub_pd(ix1,jx2);
641 dy12 = _mm256_sub_pd(iy1,jy2);
642 dz12 = _mm256_sub_pd(iz1,jz2);
643 dx13 = _mm256_sub_pd(ix1,jx3);
644 dy13 = _mm256_sub_pd(iy1,jy3);
645 dz13 = _mm256_sub_pd(iz1,jz3);
646 dx21 = _mm256_sub_pd(ix2,jx1);
647 dy21 = _mm256_sub_pd(iy2,jy1);
648 dz21 = _mm256_sub_pd(iz2,jz1);
649 dx22 = _mm256_sub_pd(ix2,jx2);
650 dy22 = _mm256_sub_pd(iy2,jy2);
651 dz22 = _mm256_sub_pd(iz2,jz2);
652 dx23 = _mm256_sub_pd(ix2,jx3);
653 dy23 = _mm256_sub_pd(iy2,jy3);
654 dz23 = _mm256_sub_pd(iz2,jz3);
655 dx31 = _mm256_sub_pd(ix3,jx1);
656 dy31 = _mm256_sub_pd(iy3,jy1);
657 dz31 = _mm256_sub_pd(iz3,jz1);
658 dx32 = _mm256_sub_pd(ix3,jx2);
659 dy32 = _mm256_sub_pd(iy3,jy2);
660 dz32 = _mm256_sub_pd(iz3,jz2);
661 dx33 = _mm256_sub_pd(ix3,jx3);
662 dy33 = _mm256_sub_pd(iy3,jy3);
663 dz33 = _mm256_sub_pd(iz3,jz3);
665 /* Calculate squared distance and things based on it */
666 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
667 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
668 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
669 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
670 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
671 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
672 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
673 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
674 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
675 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
677 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
678 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
679 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
680 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
681 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
682 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
683 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
684 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
685 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
687 rinvsq00 = gmx_mm256_inv_pd(rsq00);
688 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
689 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
690 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
691 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
692 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
693 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
694 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
695 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
696 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
698 fjx0 = _mm256_setzero_pd();
699 fjy0 = _mm256_setzero_pd();
700 fjz0 = _mm256_setzero_pd();
701 fjx1 = _mm256_setzero_pd();
702 fjy1 = _mm256_setzero_pd();
703 fjz1 = _mm256_setzero_pd();
704 fjx2 = _mm256_setzero_pd();
705 fjy2 = _mm256_setzero_pd();
706 fjz2 = _mm256_setzero_pd();
707 fjx3 = _mm256_setzero_pd();
708 fjy3 = _mm256_setzero_pd();
709 fjz3 = _mm256_setzero_pd();
711 /**************************
712 * CALCULATE INTERACTIONS *
713 **************************/
715 /* LENNARD-JONES DISPERSION/REPULSION */
717 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
718 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
719 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
720 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
721 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
723 /* Update potential sum for this i atom from the interaction with this j atom. */
724 vvdw = _mm256_andnot_pd(dummy_mask,vvdw);
725 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
729 fscal = _mm256_andnot_pd(dummy_mask,fscal);
731 /* Calculate temporary vectorial force */
732 tx = _mm256_mul_pd(fscal,dx00);
733 ty = _mm256_mul_pd(fscal,dy00);
734 tz = _mm256_mul_pd(fscal,dz00);
736 /* Update vectorial force */
737 fix0 = _mm256_add_pd(fix0,tx);
738 fiy0 = _mm256_add_pd(fiy0,ty);
739 fiz0 = _mm256_add_pd(fiz0,tz);
741 fjx0 = _mm256_add_pd(fjx0,tx);
742 fjy0 = _mm256_add_pd(fjy0,ty);
743 fjz0 = _mm256_add_pd(fjz0,tz);
745 /**************************
746 * CALCULATE INTERACTIONS *
747 **************************/
749 /* REACTION-FIELD ELECTROSTATICS */
750 velec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_add_pd(rinv11,_mm256_mul_pd(krf,rsq11)),crf));
751 felec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_mul_pd(rinv11,rinvsq11),krf2));
753 /* Update potential sum for this i atom from the interaction with this j atom. */
754 velec = _mm256_andnot_pd(dummy_mask,velec);
755 velecsum = _mm256_add_pd(velecsum,velec);
759 fscal = _mm256_andnot_pd(dummy_mask,fscal);
761 /* Calculate temporary vectorial force */
762 tx = _mm256_mul_pd(fscal,dx11);
763 ty = _mm256_mul_pd(fscal,dy11);
764 tz = _mm256_mul_pd(fscal,dz11);
766 /* Update vectorial force */
767 fix1 = _mm256_add_pd(fix1,tx);
768 fiy1 = _mm256_add_pd(fiy1,ty);
769 fiz1 = _mm256_add_pd(fiz1,tz);
771 fjx1 = _mm256_add_pd(fjx1,tx);
772 fjy1 = _mm256_add_pd(fjy1,ty);
773 fjz1 = _mm256_add_pd(fjz1,tz);
775 /**************************
776 * CALCULATE INTERACTIONS *
777 **************************/
779 /* REACTION-FIELD ELECTROSTATICS */
780 velec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_add_pd(rinv12,_mm256_mul_pd(krf,rsq12)),crf));
781 felec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_mul_pd(rinv12,rinvsq12),krf2));
783 /* Update potential sum for this i atom from the interaction with this j atom. */
784 velec = _mm256_andnot_pd(dummy_mask,velec);
785 velecsum = _mm256_add_pd(velecsum,velec);
789 fscal = _mm256_andnot_pd(dummy_mask,fscal);
791 /* Calculate temporary vectorial force */
792 tx = _mm256_mul_pd(fscal,dx12);
793 ty = _mm256_mul_pd(fscal,dy12);
794 tz = _mm256_mul_pd(fscal,dz12);
796 /* Update vectorial force */
797 fix1 = _mm256_add_pd(fix1,tx);
798 fiy1 = _mm256_add_pd(fiy1,ty);
799 fiz1 = _mm256_add_pd(fiz1,tz);
801 fjx2 = _mm256_add_pd(fjx2,tx);
802 fjy2 = _mm256_add_pd(fjy2,ty);
803 fjz2 = _mm256_add_pd(fjz2,tz);
805 /**************************
806 * CALCULATE INTERACTIONS *
807 **************************/
809 /* REACTION-FIELD ELECTROSTATICS */
810 velec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_add_pd(rinv13,_mm256_mul_pd(krf,rsq13)),crf));
811 felec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_mul_pd(rinv13,rinvsq13),krf2));
813 /* Update potential sum for this i atom from the interaction with this j atom. */
814 velec = _mm256_andnot_pd(dummy_mask,velec);
815 velecsum = _mm256_add_pd(velecsum,velec);
819 fscal = _mm256_andnot_pd(dummy_mask,fscal);
821 /* Calculate temporary vectorial force */
822 tx = _mm256_mul_pd(fscal,dx13);
823 ty = _mm256_mul_pd(fscal,dy13);
824 tz = _mm256_mul_pd(fscal,dz13);
826 /* Update vectorial force */
827 fix1 = _mm256_add_pd(fix1,tx);
828 fiy1 = _mm256_add_pd(fiy1,ty);
829 fiz1 = _mm256_add_pd(fiz1,tz);
831 fjx3 = _mm256_add_pd(fjx3,tx);
832 fjy3 = _mm256_add_pd(fjy3,ty);
833 fjz3 = _mm256_add_pd(fjz3,tz);
835 /**************************
836 * CALCULATE INTERACTIONS *
837 **************************/
839 /* REACTION-FIELD ELECTROSTATICS */
840 velec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_add_pd(rinv21,_mm256_mul_pd(krf,rsq21)),crf));
841 felec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_mul_pd(rinv21,rinvsq21),krf2));
843 /* Update potential sum for this i atom from the interaction with this j atom. */
844 velec = _mm256_andnot_pd(dummy_mask,velec);
845 velecsum = _mm256_add_pd(velecsum,velec);
849 fscal = _mm256_andnot_pd(dummy_mask,fscal);
851 /* Calculate temporary vectorial force */
852 tx = _mm256_mul_pd(fscal,dx21);
853 ty = _mm256_mul_pd(fscal,dy21);
854 tz = _mm256_mul_pd(fscal,dz21);
856 /* Update vectorial force */
857 fix2 = _mm256_add_pd(fix2,tx);
858 fiy2 = _mm256_add_pd(fiy2,ty);
859 fiz2 = _mm256_add_pd(fiz2,tz);
861 fjx1 = _mm256_add_pd(fjx1,tx);
862 fjy1 = _mm256_add_pd(fjy1,ty);
863 fjz1 = _mm256_add_pd(fjz1,tz);
865 /**************************
866 * CALCULATE INTERACTIONS *
867 **************************/
869 /* REACTION-FIELD ELECTROSTATICS */
870 velec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_add_pd(rinv22,_mm256_mul_pd(krf,rsq22)),crf));
871 felec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_mul_pd(rinv22,rinvsq22),krf2));
873 /* Update potential sum for this i atom from the interaction with this j atom. */
874 velec = _mm256_andnot_pd(dummy_mask,velec);
875 velecsum = _mm256_add_pd(velecsum,velec);
879 fscal = _mm256_andnot_pd(dummy_mask,fscal);
881 /* Calculate temporary vectorial force */
882 tx = _mm256_mul_pd(fscal,dx22);
883 ty = _mm256_mul_pd(fscal,dy22);
884 tz = _mm256_mul_pd(fscal,dz22);
886 /* Update vectorial force */
887 fix2 = _mm256_add_pd(fix2,tx);
888 fiy2 = _mm256_add_pd(fiy2,ty);
889 fiz2 = _mm256_add_pd(fiz2,tz);
891 fjx2 = _mm256_add_pd(fjx2,tx);
892 fjy2 = _mm256_add_pd(fjy2,ty);
893 fjz2 = _mm256_add_pd(fjz2,tz);
895 /**************************
896 * CALCULATE INTERACTIONS *
897 **************************/
899 /* REACTION-FIELD ELECTROSTATICS */
900 velec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_add_pd(rinv23,_mm256_mul_pd(krf,rsq23)),crf));
901 felec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_mul_pd(rinv23,rinvsq23),krf2));
903 /* Update potential sum for this i atom from the interaction with this j atom. */
904 velec = _mm256_andnot_pd(dummy_mask,velec);
905 velecsum = _mm256_add_pd(velecsum,velec);
909 fscal = _mm256_andnot_pd(dummy_mask,fscal);
911 /* Calculate temporary vectorial force */
912 tx = _mm256_mul_pd(fscal,dx23);
913 ty = _mm256_mul_pd(fscal,dy23);
914 tz = _mm256_mul_pd(fscal,dz23);
916 /* Update vectorial force */
917 fix2 = _mm256_add_pd(fix2,tx);
918 fiy2 = _mm256_add_pd(fiy2,ty);
919 fiz2 = _mm256_add_pd(fiz2,tz);
921 fjx3 = _mm256_add_pd(fjx3,tx);
922 fjy3 = _mm256_add_pd(fjy3,ty);
923 fjz3 = _mm256_add_pd(fjz3,tz);
925 /**************************
926 * CALCULATE INTERACTIONS *
927 **************************/
929 /* REACTION-FIELD ELECTROSTATICS */
930 velec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_add_pd(rinv31,_mm256_mul_pd(krf,rsq31)),crf));
931 felec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_mul_pd(rinv31,rinvsq31),krf2));
933 /* Update potential sum for this i atom from the interaction with this j atom. */
934 velec = _mm256_andnot_pd(dummy_mask,velec);
935 velecsum = _mm256_add_pd(velecsum,velec);
939 fscal = _mm256_andnot_pd(dummy_mask,fscal);
941 /* Calculate temporary vectorial force */
942 tx = _mm256_mul_pd(fscal,dx31);
943 ty = _mm256_mul_pd(fscal,dy31);
944 tz = _mm256_mul_pd(fscal,dz31);
946 /* Update vectorial force */
947 fix3 = _mm256_add_pd(fix3,tx);
948 fiy3 = _mm256_add_pd(fiy3,ty);
949 fiz3 = _mm256_add_pd(fiz3,tz);
951 fjx1 = _mm256_add_pd(fjx1,tx);
952 fjy1 = _mm256_add_pd(fjy1,ty);
953 fjz1 = _mm256_add_pd(fjz1,tz);
955 /**************************
956 * CALCULATE INTERACTIONS *
957 **************************/
959 /* REACTION-FIELD ELECTROSTATICS */
960 velec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_add_pd(rinv32,_mm256_mul_pd(krf,rsq32)),crf));
961 felec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_mul_pd(rinv32,rinvsq32),krf2));
963 /* Update potential sum for this i atom from the interaction with this j atom. */
964 velec = _mm256_andnot_pd(dummy_mask,velec);
965 velecsum = _mm256_add_pd(velecsum,velec);
969 fscal = _mm256_andnot_pd(dummy_mask,fscal);
971 /* Calculate temporary vectorial force */
972 tx = _mm256_mul_pd(fscal,dx32);
973 ty = _mm256_mul_pd(fscal,dy32);
974 tz = _mm256_mul_pd(fscal,dz32);
976 /* Update vectorial force */
977 fix3 = _mm256_add_pd(fix3,tx);
978 fiy3 = _mm256_add_pd(fiy3,ty);
979 fiz3 = _mm256_add_pd(fiz3,tz);
981 fjx2 = _mm256_add_pd(fjx2,tx);
982 fjy2 = _mm256_add_pd(fjy2,ty);
983 fjz2 = _mm256_add_pd(fjz2,tz);
985 /**************************
986 * CALCULATE INTERACTIONS *
987 **************************/
989 /* REACTION-FIELD ELECTROSTATICS */
990 velec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_add_pd(rinv33,_mm256_mul_pd(krf,rsq33)),crf));
991 felec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_mul_pd(rinv33,rinvsq33),krf2));
993 /* Update potential sum for this i atom from the interaction with this j atom. */
994 velec = _mm256_andnot_pd(dummy_mask,velec);
995 velecsum = _mm256_add_pd(velecsum,velec);
999 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1001 /* Calculate temporary vectorial force */
1002 tx = _mm256_mul_pd(fscal,dx33);
1003 ty = _mm256_mul_pd(fscal,dy33);
1004 tz = _mm256_mul_pd(fscal,dz33);
1006 /* Update vectorial force */
1007 fix3 = _mm256_add_pd(fix3,tx);
1008 fiy3 = _mm256_add_pd(fiy3,ty);
1009 fiz3 = _mm256_add_pd(fiz3,tz);
1011 fjx3 = _mm256_add_pd(fjx3,tx);
1012 fjy3 = _mm256_add_pd(fjy3,ty);
1013 fjz3 = _mm256_add_pd(fjz3,tz);
1015 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1016 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1017 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1018 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1020 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1021 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1022 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1024 /* Inner loop uses 323 flops */
1027 /* End of innermost loop */
1029 gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1030 f+i_coord_offset,fshift+i_shift_offset);
1033 /* Update potential energies */
1034 gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1035 gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1037 /* Increment number of inner iterations */
1038 inneriter += j_index_end - j_index_start;
1040 /* Outer loop uses 26 flops */
1043 /* Increment number of outer iterations */
1046 /* Update outer/inner flops */
1048 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*323);
1051 * Gromacs nonbonded kernel: nb_kernel_ElecRF_VdwLJ_GeomW4W4_F_avx_256_double
1052 * Electrostatics interaction: ReactionField
1053 * VdW interaction: LennardJones
1054 * Geometry: Water4-Water4
1055 * Calculate force/pot: Force
1058 nb_kernel_ElecRF_VdwLJ_GeomW4W4_F_avx_256_double
1059 (t_nblist * gmx_restrict nlist,
1060 rvec * gmx_restrict xx,
1061 rvec * gmx_restrict ff,
1062 t_forcerec * gmx_restrict fr,
1063 t_mdatoms * gmx_restrict mdatoms,
1064 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1065 t_nrnb * gmx_restrict nrnb)
1067 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1068 * just 0 for non-waters.
1069 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
1070 * jnr indices corresponding to data put in the four positions in the SIMD register.
1072 int i_shift_offset,i_coord_offset,outeriter,inneriter;
1073 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1074 int jnrA,jnrB,jnrC,jnrD;
1075 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1076 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1077 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1078 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
1079 real rcutoff_scalar;
1080 real *shiftvec,*fshift,*x,*f;
1081 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1082 real scratch[4*DIM];
1083 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1084 real * vdwioffsetptr0;
1085 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1086 real * vdwioffsetptr1;
1087 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1088 real * vdwioffsetptr2;
1089 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1090 real * vdwioffsetptr3;
1091 __m256d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1092 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1093 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1094 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1095 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1096 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1097 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1098 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
1099 __m256d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1100 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1101 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1102 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1103 __m256d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1104 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1105 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1106 __m256d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1107 __m256d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1108 __m256d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1109 __m256d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1110 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
1113 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1116 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
1117 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
1118 __m256d dummy_mask,cutoff_mask;
1119 __m128 tmpmask0,tmpmask1;
1120 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
1121 __m256d one = _mm256_set1_pd(1.0);
1122 __m256d two = _mm256_set1_pd(2.0);
1128 jindex = nlist->jindex;
1130 shiftidx = nlist->shift;
1132 shiftvec = fr->shift_vec[0];
1133 fshift = fr->fshift[0];
1134 facel = _mm256_set1_pd(fr->epsfac);
1135 charge = mdatoms->chargeA;
1136 krf = _mm256_set1_pd(fr->ic->k_rf);
1137 krf2 = _mm256_set1_pd(fr->ic->k_rf*2.0);
1138 crf = _mm256_set1_pd(fr->ic->c_rf);
1139 nvdwtype = fr->ntype;
1140 vdwparam = fr->nbfp;
1141 vdwtype = mdatoms->typeA;
1143 /* Setup water-specific parameters */
1144 inr = nlist->iinr[0];
1145 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
1146 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
1147 iq3 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
1148 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
1150 jq1 = _mm256_set1_pd(charge[inr+1]);
1151 jq2 = _mm256_set1_pd(charge[inr+2]);
1152 jq3 = _mm256_set1_pd(charge[inr+3]);
1153 vdwjidx0A = 2*vdwtype[inr+0];
1154 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
1155 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
1156 qq11 = _mm256_mul_pd(iq1,jq1);
1157 qq12 = _mm256_mul_pd(iq1,jq2);
1158 qq13 = _mm256_mul_pd(iq1,jq3);
1159 qq21 = _mm256_mul_pd(iq2,jq1);
1160 qq22 = _mm256_mul_pd(iq2,jq2);
1161 qq23 = _mm256_mul_pd(iq2,jq3);
1162 qq31 = _mm256_mul_pd(iq3,jq1);
1163 qq32 = _mm256_mul_pd(iq3,jq2);
1164 qq33 = _mm256_mul_pd(iq3,jq3);
1166 /* Avoid stupid compiler warnings */
1167 jnrA = jnrB = jnrC = jnrD = 0;
1168 j_coord_offsetA = 0;
1169 j_coord_offsetB = 0;
1170 j_coord_offsetC = 0;
1171 j_coord_offsetD = 0;
1176 for(iidx=0;iidx<4*DIM;iidx++)
1178 scratch[iidx] = 0.0;
1181 /* Start outer loop over neighborlists */
1182 for(iidx=0; iidx<nri; iidx++)
1184 /* Load shift vector for this list */
1185 i_shift_offset = DIM*shiftidx[iidx];
1187 /* Load limits for loop over neighbors */
1188 j_index_start = jindex[iidx];
1189 j_index_end = jindex[iidx+1];
1191 /* Get outer coordinate index */
1193 i_coord_offset = DIM*inr;
1195 /* Load i particle coords and add shift vector */
1196 gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1197 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1199 fix0 = _mm256_setzero_pd();
1200 fiy0 = _mm256_setzero_pd();
1201 fiz0 = _mm256_setzero_pd();
1202 fix1 = _mm256_setzero_pd();
1203 fiy1 = _mm256_setzero_pd();
1204 fiz1 = _mm256_setzero_pd();
1205 fix2 = _mm256_setzero_pd();
1206 fiy2 = _mm256_setzero_pd();
1207 fiz2 = _mm256_setzero_pd();
1208 fix3 = _mm256_setzero_pd();
1209 fiy3 = _mm256_setzero_pd();
1210 fiz3 = _mm256_setzero_pd();
1212 /* Start inner kernel loop */
1213 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1216 /* Get j neighbor index, and coordinate index */
1218 jnrB = jjnr[jidx+1];
1219 jnrC = jjnr[jidx+2];
1220 jnrD = jjnr[jidx+3];
1221 j_coord_offsetA = DIM*jnrA;
1222 j_coord_offsetB = DIM*jnrB;
1223 j_coord_offsetC = DIM*jnrC;
1224 j_coord_offsetD = DIM*jnrD;
1226 /* load j atom coordinates */
1227 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1228 x+j_coord_offsetC,x+j_coord_offsetD,
1229 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1230 &jy2,&jz2,&jx3,&jy3,&jz3);
1232 /* Calculate displacement vector */
1233 dx00 = _mm256_sub_pd(ix0,jx0);
1234 dy00 = _mm256_sub_pd(iy0,jy0);
1235 dz00 = _mm256_sub_pd(iz0,jz0);
1236 dx11 = _mm256_sub_pd(ix1,jx1);
1237 dy11 = _mm256_sub_pd(iy1,jy1);
1238 dz11 = _mm256_sub_pd(iz1,jz1);
1239 dx12 = _mm256_sub_pd(ix1,jx2);
1240 dy12 = _mm256_sub_pd(iy1,jy2);
1241 dz12 = _mm256_sub_pd(iz1,jz2);
1242 dx13 = _mm256_sub_pd(ix1,jx3);
1243 dy13 = _mm256_sub_pd(iy1,jy3);
1244 dz13 = _mm256_sub_pd(iz1,jz3);
1245 dx21 = _mm256_sub_pd(ix2,jx1);
1246 dy21 = _mm256_sub_pd(iy2,jy1);
1247 dz21 = _mm256_sub_pd(iz2,jz1);
1248 dx22 = _mm256_sub_pd(ix2,jx2);
1249 dy22 = _mm256_sub_pd(iy2,jy2);
1250 dz22 = _mm256_sub_pd(iz2,jz2);
1251 dx23 = _mm256_sub_pd(ix2,jx3);
1252 dy23 = _mm256_sub_pd(iy2,jy3);
1253 dz23 = _mm256_sub_pd(iz2,jz3);
1254 dx31 = _mm256_sub_pd(ix3,jx1);
1255 dy31 = _mm256_sub_pd(iy3,jy1);
1256 dz31 = _mm256_sub_pd(iz3,jz1);
1257 dx32 = _mm256_sub_pd(ix3,jx2);
1258 dy32 = _mm256_sub_pd(iy3,jy2);
1259 dz32 = _mm256_sub_pd(iz3,jz2);
1260 dx33 = _mm256_sub_pd(ix3,jx3);
1261 dy33 = _mm256_sub_pd(iy3,jy3);
1262 dz33 = _mm256_sub_pd(iz3,jz3);
1264 /* Calculate squared distance and things based on it */
1265 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1266 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1267 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1268 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
1269 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1270 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1271 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
1272 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
1273 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
1274 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
1276 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
1277 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
1278 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
1279 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
1280 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
1281 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
1282 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
1283 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
1284 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
1286 rinvsq00 = gmx_mm256_inv_pd(rsq00);
1287 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
1288 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
1289 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
1290 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
1291 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
1292 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
1293 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
1294 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
1295 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
1297 fjx0 = _mm256_setzero_pd();
1298 fjy0 = _mm256_setzero_pd();
1299 fjz0 = _mm256_setzero_pd();
1300 fjx1 = _mm256_setzero_pd();
1301 fjy1 = _mm256_setzero_pd();
1302 fjz1 = _mm256_setzero_pd();
1303 fjx2 = _mm256_setzero_pd();
1304 fjy2 = _mm256_setzero_pd();
1305 fjz2 = _mm256_setzero_pd();
1306 fjx3 = _mm256_setzero_pd();
1307 fjy3 = _mm256_setzero_pd();
1308 fjz3 = _mm256_setzero_pd();
1310 /**************************
1311 * CALCULATE INTERACTIONS *
1312 **************************/
1314 /* LENNARD-JONES DISPERSION/REPULSION */
1316 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1317 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1321 /* Calculate temporary vectorial force */
1322 tx = _mm256_mul_pd(fscal,dx00);
1323 ty = _mm256_mul_pd(fscal,dy00);
1324 tz = _mm256_mul_pd(fscal,dz00);
1326 /* Update vectorial force */
1327 fix0 = _mm256_add_pd(fix0,tx);
1328 fiy0 = _mm256_add_pd(fiy0,ty);
1329 fiz0 = _mm256_add_pd(fiz0,tz);
1331 fjx0 = _mm256_add_pd(fjx0,tx);
1332 fjy0 = _mm256_add_pd(fjy0,ty);
1333 fjz0 = _mm256_add_pd(fjz0,tz);
1335 /**************************
1336 * CALCULATE INTERACTIONS *
1337 **************************/
1339 /* REACTION-FIELD ELECTROSTATICS */
1340 felec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_mul_pd(rinv11,rinvsq11),krf2));
1344 /* Calculate temporary vectorial force */
1345 tx = _mm256_mul_pd(fscal,dx11);
1346 ty = _mm256_mul_pd(fscal,dy11);
1347 tz = _mm256_mul_pd(fscal,dz11);
1349 /* Update vectorial force */
1350 fix1 = _mm256_add_pd(fix1,tx);
1351 fiy1 = _mm256_add_pd(fiy1,ty);
1352 fiz1 = _mm256_add_pd(fiz1,tz);
1354 fjx1 = _mm256_add_pd(fjx1,tx);
1355 fjy1 = _mm256_add_pd(fjy1,ty);
1356 fjz1 = _mm256_add_pd(fjz1,tz);
1358 /**************************
1359 * CALCULATE INTERACTIONS *
1360 **************************/
1362 /* REACTION-FIELD ELECTROSTATICS */
1363 felec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_mul_pd(rinv12,rinvsq12),krf2));
1367 /* Calculate temporary vectorial force */
1368 tx = _mm256_mul_pd(fscal,dx12);
1369 ty = _mm256_mul_pd(fscal,dy12);
1370 tz = _mm256_mul_pd(fscal,dz12);
1372 /* Update vectorial force */
1373 fix1 = _mm256_add_pd(fix1,tx);
1374 fiy1 = _mm256_add_pd(fiy1,ty);
1375 fiz1 = _mm256_add_pd(fiz1,tz);
1377 fjx2 = _mm256_add_pd(fjx2,tx);
1378 fjy2 = _mm256_add_pd(fjy2,ty);
1379 fjz2 = _mm256_add_pd(fjz2,tz);
1381 /**************************
1382 * CALCULATE INTERACTIONS *
1383 **************************/
1385 /* REACTION-FIELD ELECTROSTATICS */
1386 felec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_mul_pd(rinv13,rinvsq13),krf2));
1390 /* Calculate temporary vectorial force */
1391 tx = _mm256_mul_pd(fscal,dx13);
1392 ty = _mm256_mul_pd(fscal,dy13);
1393 tz = _mm256_mul_pd(fscal,dz13);
1395 /* Update vectorial force */
1396 fix1 = _mm256_add_pd(fix1,tx);
1397 fiy1 = _mm256_add_pd(fiy1,ty);
1398 fiz1 = _mm256_add_pd(fiz1,tz);
1400 fjx3 = _mm256_add_pd(fjx3,tx);
1401 fjy3 = _mm256_add_pd(fjy3,ty);
1402 fjz3 = _mm256_add_pd(fjz3,tz);
1404 /**************************
1405 * CALCULATE INTERACTIONS *
1406 **************************/
1408 /* REACTION-FIELD ELECTROSTATICS */
1409 felec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_mul_pd(rinv21,rinvsq21),krf2));
1413 /* Calculate temporary vectorial force */
1414 tx = _mm256_mul_pd(fscal,dx21);
1415 ty = _mm256_mul_pd(fscal,dy21);
1416 tz = _mm256_mul_pd(fscal,dz21);
1418 /* Update vectorial force */
1419 fix2 = _mm256_add_pd(fix2,tx);
1420 fiy2 = _mm256_add_pd(fiy2,ty);
1421 fiz2 = _mm256_add_pd(fiz2,tz);
1423 fjx1 = _mm256_add_pd(fjx1,tx);
1424 fjy1 = _mm256_add_pd(fjy1,ty);
1425 fjz1 = _mm256_add_pd(fjz1,tz);
1427 /**************************
1428 * CALCULATE INTERACTIONS *
1429 **************************/
1431 /* REACTION-FIELD ELECTROSTATICS */
1432 felec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_mul_pd(rinv22,rinvsq22),krf2));
1436 /* Calculate temporary vectorial force */
1437 tx = _mm256_mul_pd(fscal,dx22);
1438 ty = _mm256_mul_pd(fscal,dy22);
1439 tz = _mm256_mul_pd(fscal,dz22);
1441 /* Update vectorial force */
1442 fix2 = _mm256_add_pd(fix2,tx);
1443 fiy2 = _mm256_add_pd(fiy2,ty);
1444 fiz2 = _mm256_add_pd(fiz2,tz);
1446 fjx2 = _mm256_add_pd(fjx2,tx);
1447 fjy2 = _mm256_add_pd(fjy2,ty);
1448 fjz2 = _mm256_add_pd(fjz2,tz);
1450 /**************************
1451 * CALCULATE INTERACTIONS *
1452 **************************/
1454 /* REACTION-FIELD ELECTROSTATICS */
1455 felec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_mul_pd(rinv23,rinvsq23),krf2));
1459 /* Calculate temporary vectorial force */
1460 tx = _mm256_mul_pd(fscal,dx23);
1461 ty = _mm256_mul_pd(fscal,dy23);
1462 tz = _mm256_mul_pd(fscal,dz23);
1464 /* Update vectorial force */
1465 fix2 = _mm256_add_pd(fix2,tx);
1466 fiy2 = _mm256_add_pd(fiy2,ty);
1467 fiz2 = _mm256_add_pd(fiz2,tz);
1469 fjx3 = _mm256_add_pd(fjx3,tx);
1470 fjy3 = _mm256_add_pd(fjy3,ty);
1471 fjz3 = _mm256_add_pd(fjz3,tz);
1473 /**************************
1474 * CALCULATE INTERACTIONS *
1475 **************************/
1477 /* REACTION-FIELD ELECTROSTATICS */
1478 felec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_mul_pd(rinv31,rinvsq31),krf2));
1482 /* Calculate temporary vectorial force */
1483 tx = _mm256_mul_pd(fscal,dx31);
1484 ty = _mm256_mul_pd(fscal,dy31);
1485 tz = _mm256_mul_pd(fscal,dz31);
1487 /* Update vectorial force */
1488 fix3 = _mm256_add_pd(fix3,tx);
1489 fiy3 = _mm256_add_pd(fiy3,ty);
1490 fiz3 = _mm256_add_pd(fiz3,tz);
1492 fjx1 = _mm256_add_pd(fjx1,tx);
1493 fjy1 = _mm256_add_pd(fjy1,ty);
1494 fjz1 = _mm256_add_pd(fjz1,tz);
1496 /**************************
1497 * CALCULATE INTERACTIONS *
1498 **************************/
1500 /* REACTION-FIELD ELECTROSTATICS */
1501 felec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_mul_pd(rinv32,rinvsq32),krf2));
1505 /* Calculate temporary vectorial force */
1506 tx = _mm256_mul_pd(fscal,dx32);
1507 ty = _mm256_mul_pd(fscal,dy32);
1508 tz = _mm256_mul_pd(fscal,dz32);
1510 /* Update vectorial force */
1511 fix3 = _mm256_add_pd(fix3,tx);
1512 fiy3 = _mm256_add_pd(fiy3,ty);
1513 fiz3 = _mm256_add_pd(fiz3,tz);
1515 fjx2 = _mm256_add_pd(fjx2,tx);
1516 fjy2 = _mm256_add_pd(fjy2,ty);
1517 fjz2 = _mm256_add_pd(fjz2,tz);
1519 /**************************
1520 * CALCULATE INTERACTIONS *
1521 **************************/
1523 /* REACTION-FIELD ELECTROSTATICS */
1524 felec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_mul_pd(rinv33,rinvsq33),krf2));
1528 /* Calculate temporary vectorial force */
1529 tx = _mm256_mul_pd(fscal,dx33);
1530 ty = _mm256_mul_pd(fscal,dy33);
1531 tz = _mm256_mul_pd(fscal,dz33);
1533 /* Update vectorial force */
1534 fix3 = _mm256_add_pd(fix3,tx);
1535 fiy3 = _mm256_add_pd(fiy3,ty);
1536 fiz3 = _mm256_add_pd(fiz3,tz);
1538 fjx3 = _mm256_add_pd(fjx3,tx);
1539 fjy3 = _mm256_add_pd(fjy3,ty);
1540 fjz3 = _mm256_add_pd(fjz3,tz);
1542 fjptrA = f+j_coord_offsetA;
1543 fjptrB = f+j_coord_offsetB;
1544 fjptrC = f+j_coord_offsetC;
1545 fjptrD = f+j_coord_offsetD;
1547 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1548 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1549 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1551 /* Inner loop uses 273 flops */
1554 if(jidx<j_index_end)
1557 /* Get j neighbor index, and coordinate index */
1558 jnrlistA = jjnr[jidx];
1559 jnrlistB = jjnr[jidx+1];
1560 jnrlistC = jjnr[jidx+2];
1561 jnrlistD = jjnr[jidx+3];
1562 /* Sign of each element will be negative for non-real atoms.
1563 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1564 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1566 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1568 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1569 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1570 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1572 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
1573 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
1574 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
1575 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
1576 j_coord_offsetA = DIM*jnrA;
1577 j_coord_offsetB = DIM*jnrB;
1578 j_coord_offsetC = DIM*jnrC;
1579 j_coord_offsetD = DIM*jnrD;
1581 /* load j atom coordinates */
1582 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1583 x+j_coord_offsetC,x+j_coord_offsetD,
1584 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1585 &jy2,&jz2,&jx3,&jy3,&jz3);
1587 /* Calculate displacement vector */
1588 dx00 = _mm256_sub_pd(ix0,jx0);
1589 dy00 = _mm256_sub_pd(iy0,jy0);
1590 dz00 = _mm256_sub_pd(iz0,jz0);
1591 dx11 = _mm256_sub_pd(ix1,jx1);
1592 dy11 = _mm256_sub_pd(iy1,jy1);
1593 dz11 = _mm256_sub_pd(iz1,jz1);
1594 dx12 = _mm256_sub_pd(ix1,jx2);
1595 dy12 = _mm256_sub_pd(iy1,jy2);
1596 dz12 = _mm256_sub_pd(iz1,jz2);
1597 dx13 = _mm256_sub_pd(ix1,jx3);
1598 dy13 = _mm256_sub_pd(iy1,jy3);
1599 dz13 = _mm256_sub_pd(iz1,jz3);
1600 dx21 = _mm256_sub_pd(ix2,jx1);
1601 dy21 = _mm256_sub_pd(iy2,jy1);
1602 dz21 = _mm256_sub_pd(iz2,jz1);
1603 dx22 = _mm256_sub_pd(ix2,jx2);
1604 dy22 = _mm256_sub_pd(iy2,jy2);
1605 dz22 = _mm256_sub_pd(iz2,jz2);
1606 dx23 = _mm256_sub_pd(ix2,jx3);
1607 dy23 = _mm256_sub_pd(iy2,jy3);
1608 dz23 = _mm256_sub_pd(iz2,jz3);
1609 dx31 = _mm256_sub_pd(ix3,jx1);
1610 dy31 = _mm256_sub_pd(iy3,jy1);
1611 dz31 = _mm256_sub_pd(iz3,jz1);
1612 dx32 = _mm256_sub_pd(ix3,jx2);
1613 dy32 = _mm256_sub_pd(iy3,jy2);
1614 dz32 = _mm256_sub_pd(iz3,jz2);
1615 dx33 = _mm256_sub_pd(ix3,jx3);
1616 dy33 = _mm256_sub_pd(iy3,jy3);
1617 dz33 = _mm256_sub_pd(iz3,jz3);
1619 /* Calculate squared distance and things based on it */
1620 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1621 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1622 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1623 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
1624 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1625 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1626 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
1627 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
1628 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
1629 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
1631 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
1632 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
1633 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
1634 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
1635 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
1636 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
1637 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
1638 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
1639 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
1641 rinvsq00 = gmx_mm256_inv_pd(rsq00);
1642 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
1643 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
1644 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
1645 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
1646 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
1647 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
1648 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
1649 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
1650 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
1652 fjx0 = _mm256_setzero_pd();
1653 fjy0 = _mm256_setzero_pd();
1654 fjz0 = _mm256_setzero_pd();
1655 fjx1 = _mm256_setzero_pd();
1656 fjy1 = _mm256_setzero_pd();
1657 fjz1 = _mm256_setzero_pd();
1658 fjx2 = _mm256_setzero_pd();
1659 fjy2 = _mm256_setzero_pd();
1660 fjz2 = _mm256_setzero_pd();
1661 fjx3 = _mm256_setzero_pd();
1662 fjy3 = _mm256_setzero_pd();
1663 fjz3 = _mm256_setzero_pd();
1665 /**************************
1666 * CALCULATE INTERACTIONS *
1667 **************************/
1669 /* LENNARD-JONES DISPERSION/REPULSION */
1671 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1672 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1676 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1678 /* Calculate temporary vectorial force */
1679 tx = _mm256_mul_pd(fscal,dx00);
1680 ty = _mm256_mul_pd(fscal,dy00);
1681 tz = _mm256_mul_pd(fscal,dz00);
1683 /* Update vectorial force */
1684 fix0 = _mm256_add_pd(fix0,tx);
1685 fiy0 = _mm256_add_pd(fiy0,ty);
1686 fiz0 = _mm256_add_pd(fiz0,tz);
1688 fjx0 = _mm256_add_pd(fjx0,tx);
1689 fjy0 = _mm256_add_pd(fjy0,ty);
1690 fjz0 = _mm256_add_pd(fjz0,tz);
1692 /**************************
1693 * CALCULATE INTERACTIONS *
1694 **************************/
1696 /* REACTION-FIELD ELECTROSTATICS */
1697 felec = _mm256_mul_pd(qq11,_mm256_sub_pd(_mm256_mul_pd(rinv11,rinvsq11),krf2));
1701 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1703 /* Calculate temporary vectorial force */
1704 tx = _mm256_mul_pd(fscal,dx11);
1705 ty = _mm256_mul_pd(fscal,dy11);
1706 tz = _mm256_mul_pd(fscal,dz11);
1708 /* Update vectorial force */
1709 fix1 = _mm256_add_pd(fix1,tx);
1710 fiy1 = _mm256_add_pd(fiy1,ty);
1711 fiz1 = _mm256_add_pd(fiz1,tz);
1713 fjx1 = _mm256_add_pd(fjx1,tx);
1714 fjy1 = _mm256_add_pd(fjy1,ty);
1715 fjz1 = _mm256_add_pd(fjz1,tz);
1717 /**************************
1718 * CALCULATE INTERACTIONS *
1719 **************************/
1721 /* REACTION-FIELD ELECTROSTATICS */
1722 felec = _mm256_mul_pd(qq12,_mm256_sub_pd(_mm256_mul_pd(rinv12,rinvsq12),krf2));
1726 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1728 /* Calculate temporary vectorial force */
1729 tx = _mm256_mul_pd(fscal,dx12);
1730 ty = _mm256_mul_pd(fscal,dy12);
1731 tz = _mm256_mul_pd(fscal,dz12);
1733 /* Update vectorial force */
1734 fix1 = _mm256_add_pd(fix1,tx);
1735 fiy1 = _mm256_add_pd(fiy1,ty);
1736 fiz1 = _mm256_add_pd(fiz1,tz);
1738 fjx2 = _mm256_add_pd(fjx2,tx);
1739 fjy2 = _mm256_add_pd(fjy2,ty);
1740 fjz2 = _mm256_add_pd(fjz2,tz);
1742 /**************************
1743 * CALCULATE INTERACTIONS *
1744 **************************/
1746 /* REACTION-FIELD ELECTROSTATICS */
1747 felec = _mm256_mul_pd(qq13,_mm256_sub_pd(_mm256_mul_pd(rinv13,rinvsq13),krf2));
1751 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1753 /* Calculate temporary vectorial force */
1754 tx = _mm256_mul_pd(fscal,dx13);
1755 ty = _mm256_mul_pd(fscal,dy13);
1756 tz = _mm256_mul_pd(fscal,dz13);
1758 /* Update vectorial force */
1759 fix1 = _mm256_add_pd(fix1,tx);
1760 fiy1 = _mm256_add_pd(fiy1,ty);
1761 fiz1 = _mm256_add_pd(fiz1,tz);
1763 fjx3 = _mm256_add_pd(fjx3,tx);
1764 fjy3 = _mm256_add_pd(fjy3,ty);
1765 fjz3 = _mm256_add_pd(fjz3,tz);
1767 /**************************
1768 * CALCULATE INTERACTIONS *
1769 **************************/
1771 /* REACTION-FIELD ELECTROSTATICS */
1772 felec = _mm256_mul_pd(qq21,_mm256_sub_pd(_mm256_mul_pd(rinv21,rinvsq21),krf2));
1776 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1778 /* Calculate temporary vectorial force */
1779 tx = _mm256_mul_pd(fscal,dx21);
1780 ty = _mm256_mul_pd(fscal,dy21);
1781 tz = _mm256_mul_pd(fscal,dz21);
1783 /* Update vectorial force */
1784 fix2 = _mm256_add_pd(fix2,tx);
1785 fiy2 = _mm256_add_pd(fiy2,ty);
1786 fiz2 = _mm256_add_pd(fiz2,tz);
1788 fjx1 = _mm256_add_pd(fjx1,tx);
1789 fjy1 = _mm256_add_pd(fjy1,ty);
1790 fjz1 = _mm256_add_pd(fjz1,tz);
1792 /**************************
1793 * CALCULATE INTERACTIONS *
1794 **************************/
1796 /* REACTION-FIELD ELECTROSTATICS */
1797 felec = _mm256_mul_pd(qq22,_mm256_sub_pd(_mm256_mul_pd(rinv22,rinvsq22),krf2));
1801 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1803 /* Calculate temporary vectorial force */
1804 tx = _mm256_mul_pd(fscal,dx22);
1805 ty = _mm256_mul_pd(fscal,dy22);
1806 tz = _mm256_mul_pd(fscal,dz22);
1808 /* Update vectorial force */
1809 fix2 = _mm256_add_pd(fix2,tx);
1810 fiy2 = _mm256_add_pd(fiy2,ty);
1811 fiz2 = _mm256_add_pd(fiz2,tz);
1813 fjx2 = _mm256_add_pd(fjx2,tx);
1814 fjy2 = _mm256_add_pd(fjy2,ty);
1815 fjz2 = _mm256_add_pd(fjz2,tz);
1817 /**************************
1818 * CALCULATE INTERACTIONS *
1819 **************************/
1821 /* REACTION-FIELD ELECTROSTATICS */
1822 felec = _mm256_mul_pd(qq23,_mm256_sub_pd(_mm256_mul_pd(rinv23,rinvsq23),krf2));
1826 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1828 /* Calculate temporary vectorial force */
1829 tx = _mm256_mul_pd(fscal,dx23);
1830 ty = _mm256_mul_pd(fscal,dy23);
1831 tz = _mm256_mul_pd(fscal,dz23);
1833 /* Update vectorial force */
1834 fix2 = _mm256_add_pd(fix2,tx);
1835 fiy2 = _mm256_add_pd(fiy2,ty);
1836 fiz2 = _mm256_add_pd(fiz2,tz);
1838 fjx3 = _mm256_add_pd(fjx3,tx);
1839 fjy3 = _mm256_add_pd(fjy3,ty);
1840 fjz3 = _mm256_add_pd(fjz3,tz);
1842 /**************************
1843 * CALCULATE INTERACTIONS *
1844 **************************/
1846 /* REACTION-FIELD ELECTROSTATICS */
1847 felec = _mm256_mul_pd(qq31,_mm256_sub_pd(_mm256_mul_pd(rinv31,rinvsq31),krf2));
1851 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1853 /* Calculate temporary vectorial force */
1854 tx = _mm256_mul_pd(fscal,dx31);
1855 ty = _mm256_mul_pd(fscal,dy31);
1856 tz = _mm256_mul_pd(fscal,dz31);
1858 /* Update vectorial force */
1859 fix3 = _mm256_add_pd(fix3,tx);
1860 fiy3 = _mm256_add_pd(fiy3,ty);
1861 fiz3 = _mm256_add_pd(fiz3,tz);
1863 fjx1 = _mm256_add_pd(fjx1,tx);
1864 fjy1 = _mm256_add_pd(fjy1,ty);
1865 fjz1 = _mm256_add_pd(fjz1,tz);
1867 /**************************
1868 * CALCULATE INTERACTIONS *
1869 **************************/
1871 /* REACTION-FIELD ELECTROSTATICS */
1872 felec = _mm256_mul_pd(qq32,_mm256_sub_pd(_mm256_mul_pd(rinv32,rinvsq32),krf2));
1876 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1878 /* Calculate temporary vectorial force */
1879 tx = _mm256_mul_pd(fscal,dx32);
1880 ty = _mm256_mul_pd(fscal,dy32);
1881 tz = _mm256_mul_pd(fscal,dz32);
1883 /* Update vectorial force */
1884 fix3 = _mm256_add_pd(fix3,tx);
1885 fiy3 = _mm256_add_pd(fiy3,ty);
1886 fiz3 = _mm256_add_pd(fiz3,tz);
1888 fjx2 = _mm256_add_pd(fjx2,tx);
1889 fjy2 = _mm256_add_pd(fjy2,ty);
1890 fjz2 = _mm256_add_pd(fjz2,tz);
1892 /**************************
1893 * CALCULATE INTERACTIONS *
1894 **************************/
1896 /* REACTION-FIELD ELECTROSTATICS */
1897 felec = _mm256_mul_pd(qq33,_mm256_sub_pd(_mm256_mul_pd(rinv33,rinvsq33),krf2));
1901 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1903 /* Calculate temporary vectorial force */
1904 tx = _mm256_mul_pd(fscal,dx33);
1905 ty = _mm256_mul_pd(fscal,dy33);
1906 tz = _mm256_mul_pd(fscal,dz33);
1908 /* Update vectorial force */
1909 fix3 = _mm256_add_pd(fix3,tx);
1910 fiy3 = _mm256_add_pd(fiy3,ty);
1911 fiz3 = _mm256_add_pd(fiz3,tz);
1913 fjx3 = _mm256_add_pd(fjx3,tx);
1914 fjy3 = _mm256_add_pd(fjy3,ty);
1915 fjz3 = _mm256_add_pd(fjz3,tz);
1917 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1918 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1919 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1920 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1922 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1923 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1924 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1926 /* Inner loop uses 273 flops */
1929 /* End of innermost loop */
1931 gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1932 f+i_coord_offset,fshift+i_shift_offset);
1934 /* Increment number of inner iterations */
1935 inneriter += j_index_end - j_index_start;
1937 /* Outer loop uses 24 flops */
1940 /* Increment number of outer iterations */
1943 /* Update outer/inner flops */
1945 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*273);