2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2012,2013, by the GROMACS development team, led by
5 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
6 * and including many others, as listed in the AUTHORS file in the
7 * top-level source directory and at http://www.gromacs.org.
9 * GROMACS is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public License
11 * as published by the Free Software Foundation; either version 2.1
12 * of the License, or (at your option) any later version.
14 * GROMACS is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with GROMACS; if not, see
21 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
22 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 * If you want to redistribute modifications to GROMACS, please
25 * consider that scientific software is very special. Version
26 * control is crucial - bugs must be traceable. We will be happy to
27 * consider code for inclusion in the official distribution, but
28 * derived work must not be called official GROMACS. Details are found
29 * in the README & COPYING files - if they are missing, get the
30 * official version at http://www.gromacs.org.
32 * To help us fund GROMACS development, we humbly ask that you cite
33 * the research papers on the package. Check out http://www.gromacs.org.
36 * Note: this file was generated by the GROMACS avx_256_double kernel generator.
44 #include "../nb_kernel.h"
45 #include "types/simple.h"
49 #include "gromacs/simd/math_x86_avx_256_double.h"
50 #include "kernelutil_x86_avx_256_double.h"
53 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJ_GeomW4W4_VF_avx_256_double
54 * Electrostatics interaction: Ewald
55 * VdW interaction: LennardJones
56 * Geometry: Water4-Water4
57 * Calculate force/pot: PotentialAndForce
60 nb_kernel_ElecEw_VdwLJ_GeomW4W4_VF_avx_256_double
61 (t_nblist * gmx_restrict nlist,
62 rvec * gmx_restrict xx,
63 rvec * gmx_restrict ff,
64 t_forcerec * gmx_restrict fr,
65 t_mdatoms * gmx_restrict mdatoms,
66 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
67 t_nrnb * gmx_restrict nrnb)
69 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
70 * just 0 for non-waters.
71 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
72 * jnr indices corresponding to data put in the four positions in the SIMD register.
74 int i_shift_offset,i_coord_offset,outeriter,inneriter;
75 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
76 int jnrA,jnrB,jnrC,jnrD;
77 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
78 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
79 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
80 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
82 real *shiftvec,*fshift,*x,*f;
83 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
85 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
86 real * vdwioffsetptr0;
87 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
88 real * vdwioffsetptr1;
89 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
90 real * vdwioffsetptr2;
91 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
92 real * vdwioffsetptr3;
93 __m256d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
94 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
95 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
96 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
97 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
98 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
99 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
100 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
101 __m256d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
102 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
103 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
104 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
105 __m256d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
106 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
107 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
108 __m256d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
109 __m256d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
110 __m256d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
111 __m256d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
112 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
115 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
118 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
119 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
121 __m256d ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
122 __m256d beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
124 __m256d dummy_mask,cutoff_mask;
125 __m128 tmpmask0,tmpmask1;
126 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
127 __m256d one = _mm256_set1_pd(1.0);
128 __m256d two = _mm256_set1_pd(2.0);
134 jindex = nlist->jindex;
136 shiftidx = nlist->shift;
138 shiftvec = fr->shift_vec[0];
139 fshift = fr->fshift[0];
140 facel = _mm256_set1_pd(fr->epsfac);
141 charge = mdatoms->chargeA;
142 nvdwtype = fr->ntype;
144 vdwtype = mdatoms->typeA;
146 sh_ewald = _mm256_set1_pd(fr->ic->sh_ewald);
147 beta = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
148 beta2 = _mm256_mul_pd(beta,beta);
149 beta3 = _mm256_mul_pd(beta,beta2);
151 ewtab = fr->ic->tabq_coul_FDV0;
152 ewtabscale = _mm256_set1_pd(fr->ic->tabq_scale);
153 ewtabhalfspace = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
155 /* Setup water-specific parameters */
156 inr = nlist->iinr[0];
157 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
158 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
159 iq3 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
160 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
162 jq1 = _mm256_set1_pd(charge[inr+1]);
163 jq2 = _mm256_set1_pd(charge[inr+2]);
164 jq3 = _mm256_set1_pd(charge[inr+3]);
165 vdwjidx0A = 2*vdwtype[inr+0];
166 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
167 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
168 qq11 = _mm256_mul_pd(iq1,jq1);
169 qq12 = _mm256_mul_pd(iq1,jq2);
170 qq13 = _mm256_mul_pd(iq1,jq3);
171 qq21 = _mm256_mul_pd(iq2,jq1);
172 qq22 = _mm256_mul_pd(iq2,jq2);
173 qq23 = _mm256_mul_pd(iq2,jq3);
174 qq31 = _mm256_mul_pd(iq3,jq1);
175 qq32 = _mm256_mul_pd(iq3,jq2);
176 qq33 = _mm256_mul_pd(iq3,jq3);
178 /* Avoid stupid compiler warnings */
179 jnrA = jnrB = jnrC = jnrD = 0;
188 for(iidx=0;iidx<4*DIM;iidx++)
193 /* Start outer loop over neighborlists */
194 for(iidx=0; iidx<nri; iidx++)
196 /* Load shift vector for this list */
197 i_shift_offset = DIM*shiftidx[iidx];
199 /* Load limits for loop over neighbors */
200 j_index_start = jindex[iidx];
201 j_index_end = jindex[iidx+1];
203 /* Get outer coordinate index */
205 i_coord_offset = DIM*inr;
207 /* Load i particle coords and add shift vector */
208 gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
209 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
211 fix0 = _mm256_setzero_pd();
212 fiy0 = _mm256_setzero_pd();
213 fiz0 = _mm256_setzero_pd();
214 fix1 = _mm256_setzero_pd();
215 fiy1 = _mm256_setzero_pd();
216 fiz1 = _mm256_setzero_pd();
217 fix2 = _mm256_setzero_pd();
218 fiy2 = _mm256_setzero_pd();
219 fiz2 = _mm256_setzero_pd();
220 fix3 = _mm256_setzero_pd();
221 fiy3 = _mm256_setzero_pd();
222 fiz3 = _mm256_setzero_pd();
224 /* Reset potential sums */
225 velecsum = _mm256_setzero_pd();
226 vvdwsum = _mm256_setzero_pd();
228 /* Start inner kernel loop */
229 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
232 /* Get j neighbor index, and coordinate index */
237 j_coord_offsetA = DIM*jnrA;
238 j_coord_offsetB = DIM*jnrB;
239 j_coord_offsetC = DIM*jnrC;
240 j_coord_offsetD = DIM*jnrD;
242 /* load j atom coordinates */
243 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
244 x+j_coord_offsetC,x+j_coord_offsetD,
245 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
246 &jy2,&jz2,&jx3,&jy3,&jz3);
248 /* Calculate displacement vector */
249 dx00 = _mm256_sub_pd(ix0,jx0);
250 dy00 = _mm256_sub_pd(iy0,jy0);
251 dz00 = _mm256_sub_pd(iz0,jz0);
252 dx11 = _mm256_sub_pd(ix1,jx1);
253 dy11 = _mm256_sub_pd(iy1,jy1);
254 dz11 = _mm256_sub_pd(iz1,jz1);
255 dx12 = _mm256_sub_pd(ix1,jx2);
256 dy12 = _mm256_sub_pd(iy1,jy2);
257 dz12 = _mm256_sub_pd(iz1,jz2);
258 dx13 = _mm256_sub_pd(ix1,jx3);
259 dy13 = _mm256_sub_pd(iy1,jy3);
260 dz13 = _mm256_sub_pd(iz1,jz3);
261 dx21 = _mm256_sub_pd(ix2,jx1);
262 dy21 = _mm256_sub_pd(iy2,jy1);
263 dz21 = _mm256_sub_pd(iz2,jz1);
264 dx22 = _mm256_sub_pd(ix2,jx2);
265 dy22 = _mm256_sub_pd(iy2,jy2);
266 dz22 = _mm256_sub_pd(iz2,jz2);
267 dx23 = _mm256_sub_pd(ix2,jx3);
268 dy23 = _mm256_sub_pd(iy2,jy3);
269 dz23 = _mm256_sub_pd(iz2,jz3);
270 dx31 = _mm256_sub_pd(ix3,jx1);
271 dy31 = _mm256_sub_pd(iy3,jy1);
272 dz31 = _mm256_sub_pd(iz3,jz1);
273 dx32 = _mm256_sub_pd(ix3,jx2);
274 dy32 = _mm256_sub_pd(iy3,jy2);
275 dz32 = _mm256_sub_pd(iz3,jz2);
276 dx33 = _mm256_sub_pd(ix3,jx3);
277 dy33 = _mm256_sub_pd(iy3,jy3);
278 dz33 = _mm256_sub_pd(iz3,jz3);
280 /* Calculate squared distance and things based on it */
281 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
282 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
283 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
284 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
285 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
286 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
287 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
288 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
289 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
290 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
292 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
293 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
294 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
295 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
296 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
297 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
298 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
299 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
300 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
302 rinvsq00 = gmx_mm256_inv_pd(rsq00);
303 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
304 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
305 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
306 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
307 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
308 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
309 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
310 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
311 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
313 fjx0 = _mm256_setzero_pd();
314 fjy0 = _mm256_setzero_pd();
315 fjz0 = _mm256_setzero_pd();
316 fjx1 = _mm256_setzero_pd();
317 fjy1 = _mm256_setzero_pd();
318 fjz1 = _mm256_setzero_pd();
319 fjx2 = _mm256_setzero_pd();
320 fjy2 = _mm256_setzero_pd();
321 fjz2 = _mm256_setzero_pd();
322 fjx3 = _mm256_setzero_pd();
323 fjy3 = _mm256_setzero_pd();
324 fjz3 = _mm256_setzero_pd();
326 /**************************
327 * CALCULATE INTERACTIONS *
328 **************************/
330 /* LENNARD-JONES DISPERSION/REPULSION */
332 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
333 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
334 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
335 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
336 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
338 /* Update potential sum for this i atom from the interaction with this j atom. */
339 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
343 /* Calculate temporary vectorial force */
344 tx = _mm256_mul_pd(fscal,dx00);
345 ty = _mm256_mul_pd(fscal,dy00);
346 tz = _mm256_mul_pd(fscal,dz00);
348 /* Update vectorial force */
349 fix0 = _mm256_add_pd(fix0,tx);
350 fiy0 = _mm256_add_pd(fiy0,ty);
351 fiz0 = _mm256_add_pd(fiz0,tz);
353 fjx0 = _mm256_add_pd(fjx0,tx);
354 fjy0 = _mm256_add_pd(fjy0,ty);
355 fjz0 = _mm256_add_pd(fjz0,tz);
357 /**************************
358 * CALCULATE INTERACTIONS *
359 **************************/
361 r11 = _mm256_mul_pd(rsq11,rinv11);
363 /* EWALD ELECTROSTATICS */
365 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
366 ewrt = _mm256_mul_pd(r11,ewtabscale);
367 ewitab = _mm256_cvttpd_epi32(ewrt);
368 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
369 ewitab = _mm_slli_epi32(ewitab,2);
370 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
371 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
372 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
373 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
374 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
375 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
376 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
377 velec = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
378 felec = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
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,dx11);
387 ty = _mm256_mul_pd(fscal,dy11);
388 tz = _mm256_mul_pd(fscal,dz11);
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 fjx1 = _mm256_add_pd(fjx1,tx);
396 fjy1 = _mm256_add_pd(fjy1,ty);
397 fjz1 = _mm256_add_pd(fjz1,tz);
399 /**************************
400 * CALCULATE INTERACTIONS *
401 **************************/
403 r12 = _mm256_mul_pd(rsq12,rinv12);
405 /* EWALD ELECTROSTATICS */
407 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
408 ewrt = _mm256_mul_pd(r12,ewtabscale);
409 ewitab = _mm256_cvttpd_epi32(ewrt);
410 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
411 ewitab = _mm_slli_epi32(ewitab,2);
412 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
413 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
414 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
415 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
416 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
417 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
418 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
419 velec = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
420 felec = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
422 /* Update potential sum for this i atom from the interaction with this j atom. */
423 velecsum = _mm256_add_pd(velecsum,velec);
427 /* Calculate temporary vectorial force */
428 tx = _mm256_mul_pd(fscal,dx12);
429 ty = _mm256_mul_pd(fscal,dy12);
430 tz = _mm256_mul_pd(fscal,dz12);
432 /* Update vectorial force */
433 fix1 = _mm256_add_pd(fix1,tx);
434 fiy1 = _mm256_add_pd(fiy1,ty);
435 fiz1 = _mm256_add_pd(fiz1,tz);
437 fjx2 = _mm256_add_pd(fjx2,tx);
438 fjy2 = _mm256_add_pd(fjy2,ty);
439 fjz2 = _mm256_add_pd(fjz2,tz);
441 /**************************
442 * CALCULATE INTERACTIONS *
443 **************************/
445 r13 = _mm256_mul_pd(rsq13,rinv13);
447 /* EWALD ELECTROSTATICS */
449 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
450 ewrt = _mm256_mul_pd(r13,ewtabscale);
451 ewitab = _mm256_cvttpd_epi32(ewrt);
452 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
453 ewitab = _mm_slli_epi32(ewitab,2);
454 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
455 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
456 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
457 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
458 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
459 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
460 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
461 velec = _mm256_mul_pd(qq13,_mm256_sub_pd(rinv13,velec));
462 felec = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
464 /* Update potential sum for this i atom from the interaction with this j atom. */
465 velecsum = _mm256_add_pd(velecsum,velec);
469 /* Calculate temporary vectorial force */
470 tx = _mm256_mul_pd(fscal,dx13);
471 ty = _mm256_mul_pd(fscal,dy13);
472 tz = _mm256_mul_pd(fscal,dz13);
474 /* Update vectorial force */
475 fix1 = _mm256_add_pd(fix1,tx);
476 fiy1 = _mm256_add_pd(fiy1,ty);
477 fiz1 = _mm256_add_pd(fiz1,tz);
479 fjx3 = _mm256_add_pd(fjx3,tx);
480 fjy3 = _mm256_add_pd(fjy3,ty);
481 fjz3 = _mm256_add_pd(fjz3,tz);
483 /**************************
484 * CALCULATE INTERACTIONS *
485 **************************/
487 r21 = _mm256_mul_pd(rsq21,rinv21);
489 /* EWALD ELECTROSTATICS */
491 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
492 ewrt = _mm256_mul_pd(r21,ewtabscale);
493 ewitab = _mm256_cvttpd_epi32(ewrt);
494 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
495 ewitab = _mm_slli_epi32(ewitab,2);
496 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
497 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
498 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
499 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
500 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
501 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
502 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
503 velec = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
504 felec = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
506 /* Update potential sum for this i atom from the interaction with this j atom. */
507 velecsum = _mm256_add_pd(velecsum,velec);
511 /* Calculate temporary vectorial force */
512 tx = _mm256_mul_pd(fscal,dx21);
513 ty = _mm256_mul_pd(fscal,dy21);
514 tz = _mm256_mul_pd(fscal,dz21);
516 /* Update vectorial force */
517 fix2 = _mm256_add_pd(fix2,tx);
518 fiy2 = _mm256_add_pd(fiy2,ty);
519 fiz2 = _mm256_add_pd(fiz2,tz);
521 fjx1 = _mm256_add_pd(fjx1,tx);
522 fjy1 = _mm256_add_pd(fjy1,ty);
523 fjz1 = _mm256_add_pd(fjz1,tz);
525 /**************************
526 * CALCULATE INTERACTIONS *
527 **************************/
529 r22 = _mm256_mul_pd(rsq22,rinv22);
531 /* EWALD ELECTROSTATICS */
533 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
534 ewrt = _mm256_mul_pd(r22,ewtabscale);
535 ewitab = _mm256_cvttpd_epi32(ewrt);
536 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
537 ewitab = _mm_slli_epi32(ewitab,2);
538 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
539 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
540 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
541 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
542 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
543 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
544 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
545 velec = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
546 felec = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
548 /* Update potential sum for this i atom from the interaction with this j atom. */
549 velecsum = _mm256_add_pd(velecsum,velec);
553 /* Calculate temporary vectorial force */
554 tx = _mm256_mul_pd(fscal,dx22);
555 ty = _mm256_mul_pd(fscal,dy22);
556 tz = _mm256_mul_pd(fscal,dz22);
558 /* Update vectorial force */
559 fix2 = _mm256_add_pd(fix2,tx);
560 fiy2 = _mm256_add_pd(fiy2,ty);
561 fiz2 = _mm256_add_pd(fiz2,tz);
563 fjx2 = _mm256_add_pd(fjx2,tx);
564 fjy2 = _mm256_add_pd(fjy2,ty);
565 fjz2 = _mm256_add_pd(fjz2,tz);
567 /**************************
568 * CALCULATE INTERACTIONS *
569 **************************/
571 r23 = _mm256_mul_pd(rsq23,rinv23);
573 /* EWALD ELECTROSTATICS */
575 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
576 ewrt = _mm256_mul_pd(r23,ewtabscale);
577 ewitab = _mm256_cvttpd_epi32(ewrt);
578 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
579 ewitab = _mm_slli_epi32(ewitab,2);
580 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
581 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
582 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
583 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
584 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
585 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
586 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
587 velec = _mm256_mul_pd(qq23,_mm256_sub_pd(rinv23,velec));
588 felec = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
590 /* Update potential sum for this i atom from the interaction with this j atom. */
591 velecsum = _mm256_add_pd(velecsum,velec);
595 /* Calculate temporary vectorial force */
596 tx = _mm256_mul_pd(fscal,dx23);
597 ty = _mm256_mul_pd(fscal,dy23);
598 tz = _mm256_mul_pd(fscal,dz23);
600 /* Update vectorial force */
601 fix2 = _mm256_add_pd(fix2,tx);
602 fiy2 = _mm256_add_pd(fiy2,ty);
603 fiz2 = _mm256_add_pd(fiz2,tz);
605 fjx3 = _mm256_add_pd(fjx3,tx);
606 fjy3 = _mm256_add_pd(fjy3,ty);
607 fjz3 = _mm256_add_pd(fjz3,tz);
609 /**************************
610 * CALCULATE INTERACTIONS *
611 **************************/
613 r31 = _mm256_mul_pd(rsq31,rinv31);
615 /* EWALD ELECTROSTATICS */
617 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
618 ewrt = _mm256_mul_pd(r31,ewtabscale);
619 ewitab = _mm256_cvttpd_epi32(ewrt);
620 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
621 ewitab = _mm_slli_epi32(ewitab,2);
622 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
623 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
624 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
625 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
626 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
627 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
628 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
629 velec = _mm256_mul_pd(qq31,_mm256_sub_pd(rinv31,velec));
630 felec = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
632 /* Update potential sum for this i atom from the interaction with this j atom. */
633 velecsum = _mm256_add_pd(velecsum,velec);
637 /* Calculate temporary vectorial force */
638 tx = _mm256_mul_pd(fscal,dx31);
639 ty = _mm256_mul_pd(fscal,dy31);
640 tz = _mm256_mul_pd(fscal,dz31);
642 /* Update vectorial force */
643 fix3 = _mm256_add_pd(fix3,tx);
644 fiy3 = _mm256_add_pd(fiy3,ty);
645 fiz3 = _mm256_add_pd(fiz3,tz);
647 fjx1 = _mm256_add_pd(fjx1,tx);
648 fjy1 = _mm256_add_pd(fjy1,ty);
649 fjz1 = _mm256_add_pd(fjz1,tz);
651 /**************************
652 * CALCULATE INTERACTIONS *
653 **************************/
655 r32 = _mm256_mul_pd(rsq32,rinv32);
657 /* EWALD ELECTROSTATICS */
659 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
660 ewrt = _mm256_mul_pd(r32,ewtabscale);
661 ewitab = _mm256_cvttpd_epi32(ewrt);
662 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
663 ewitab = _mm_slli_epi32(ewitab,2);
664 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
665 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
666 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
667 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
668 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
669 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
670 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
671 velec = _mm256_mul_pd(qq32,_mm256_sub_pd(rinv32,velec));
672 felec = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
674 /* Update potential sum for this i atom from the interaction with this j atom. */
675 velecsum = _mm256_add_pd(velecsum,velec);
679 /* Calculate temporary vectorial force */
680 tx = _mm256_mul_pd(fscal,dx32);
681 ty = _mm256_mul_pd(fscal,dy32);
682 tz = _mm256_mul_pd(fscal,dz32);
684 /* Update vectorial force */
685 fix3 = _mm256_add_pd(fix3,tx);
686 fiy3 = _mm256_add_pd(fiy3,ty);
687 fiz3 = _mm256_add_pd(fiz3,tz);
689 fjx2 = _mm256_add_pd(fjx2,tx);
690 fjy2 = _mm256_add_pd(fjy2,ty);
691 fjz2 = _mm256_add_pd(fjz2,tz);
693 /**************************
694 * CALCULATE INTERACTIONS *
695 **************************/
697 r33 = _mm256_mul_pd(rsq33,rinv33);
699 /* EWALD ELECTROSTATICS */
701 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
702 ewrt = _mm256_mul_pd(r33,ewtabscale);
703 ewitab = _mm256_cvttpd_epi32(ewrt);
704 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
705 ewitab = _mm_slli_epi32(ewitab,2);
706 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
707 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
708 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
709 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
710 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
711 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
712 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
713 velec = _mm256_mul_pd(qq33,_mm256_sub_pd(rinv33,velec));
714 felec = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
716 /* Update potential sum for this i atom from the interaction with this j atom. */
717 velecsum = _mm256_add_pd(velecsum,velec);
721 /* Calculate temporary vectorial force */
722 tx = _mm256_mul_pd(fscal,dx33);
723 ty = _mm256_mul_pd(fscal,dy33);
724 tz = _mm256_mul_pd(fscal,dz33);
726 /* Update vectorial force */
727 fix3 = _mm256_add_pd(fix3,tx);
728 fiy3 = _mm256_add_pd(fiy3,ty);
729 fiz3 = _mm256_add_pd(fiz3,tz);
731 fjx3 = _mm256_add_pd(fjx3,tx);
732 fjy3 = _mm256_add_pd(fjy3,ty);
733 fjz3 = _mm256_add_pd(fjz3,tz);
735 fjptrA = f+j_coord_offsetA;
736 fjptrB = f+j_coord_offsetB;
737 fjptrC = f+j_coord_offsetC;
738 fjptrD = f+j_coord_offsetD;
740 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
741 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
742 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
744 /* Inner loop uses 404 flops */
750 /* Get j neighbor index, and coordinate index */
751 jnrlistA = jjnr[jidx];
752 jnrlistB = jjnr[jidx+1];
753 jnrlistC = jjnr[jidx+2];
754 jnrlistD = jjnr[jidx+3];
755 /* Sign of each element will be negative for non-real atoms.
756 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
757 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
759 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
761 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
762 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
763 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
765 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
766 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
767 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
768 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
769 j_coord_offsetA = DIM*jnrA;
770 j_coord_offsetB = DIM*jnrB;
771 j_coord_offsetC = DIM*jnrC;
772 j_coord_offsetD = DIM*jnrD;
774 /* load j atom coordinates */
775 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
776 x+j_coord_offsetC,x+j_coord_offsetD,
777 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
778 &jy2,&jz2,&jx3,&jy3,&jz3);
780 /* Calculate displacement vector */
781 dx00 = _mm256_sub_pd(ix0,jx0);
782 dy00 = _mm256_sub_pd(iy0,jy0);
783 dz00 = _mm256_sub_pd(iz0,jz0);
784 dx11 = _mm256_sub_pd(ix1,jx1);
785 dy11 = _mm256_sub_pd(iy1,jy1);
786 dz11 = _mm256_sub_pd(iz1,jz1);
787 dx12 = _mm256_sub_pd(ix1,jx2);
788 dy12 = _mm256_sub_pd(iy1,jy2);
789 dz12 = _mm256_sub_pd(iz1,jz2);
790 dx13 = _mm256_sub_pd(ix1,jx3);
791 dy13 = _mm256_sub_pd(iy1,jy3);
792 dz13 = _mm256_sub_pd(iz1,jz3);
793 dx21 = _mm256_sub_pd(ix2,jx1);
794 dy21 = _mm256_sub_pd(iy2,jy1);
795 dz21 = _mm256_sub_pd(iz2,jz1);
796 dx22 = _mm256_sub_pd(ix2,jx2);
797 dy22 = _mm256_sub_pd(iy2,jy2);
798 dz22 = _mm256_sub_pd(iz2,jz2);
799 dx23 = _mm256_sub_pd(ix2,jx3);
800 dy23 = _mm256_sub_pd(iy2,jy3);
801 dz23 = _mm256_sub_pd(iz2,jz3);
802 dx31 = _mm256_sub_pd(ix3,jx1);
803 dy31 = _mm256_sub_pd(iy3,jy1);
804 dz31 = _mm256_sub_pd(iz3,jz1);
805 dx32 = _mm256_sub_pd(ix3,jx2);
806 dy32 = _mm256_sub_pd(iy3,jy2);
807 dz32 = _mm256_sub_pd(iz3,jz2);
808 dx33 = _mm256_sub_pd(ix3,jx3);
809 dy33 = _mm256_sub_pd(iy3,jy3);
810 dz33 = _mm256_sub_pd(iz3,jz3);
812 /* Calculate squared distance and things based on it */
813 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
814 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
815 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
816 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
817 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
818 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
819 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
820 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
821 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
822 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
824 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
825 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
826 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
827 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
828 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
829 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
830 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
831 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
832 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
834 rinvsq00 = gmx_mm256_inv_pd(rsq00);
835 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
836 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
837 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
838 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
839 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
840 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
841 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
842 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
843 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
845 fjx0 = _mm256_setzero_pd();
846 fjy0 = _mm256_setzero_pd();
847 fjz0 = _mm256_setzero_pd();
848 fjx1 = _mm256_setzero_pd();
849 fjy1 = _mm256_setzero_pd();
850 fjz1 = _mm256_setzero_pd();
851 fjx2 = _mm256_setzero_pd();
852 fjy2 = _mm256_setzero_pd();
853 fjz2 = _mm256_setzero_pd();
854 fjx3 = _mm256_setzero_pd();
855 fjy3 = _mm256_setzero_pd();
856 fjz3 = _mm256_setzero_pd();
858 /**************************
859 * CALCULATE INTERACTIONS *
860 **************************/
862 /* LENNARD-JONES DISPERSION/REPULSION */
864 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
865 vvdw6 = _mm256_mul_pd(c6_00,rinvsix);
866 vvdw12 = _mm256_mul_pd(c12_00,_mm256_mul_pd(rinvsix,rinvsix));
867 vvdw = _mm256_sub_pd( _mm256_mul_pd(vvdw12,one_twelfth) , _mm256_mul_pd(vvdw6,one_sixth) );
868 fvdw = _mm256_mul_pd(_mm256_sub_pd(vvdw12,vvdw6),rinvsq00);
870 /* Update potential sum for this i atom from the interaction with this j atom. */
871 vvdw = _mm256_andnot_pd(dummy_mask,vvdw);
872 vvdwsum = _mm256_add_pd(vvdwsum,vvdw);
876 fscal = _mm256_andnot_pd(dummy_mask,fscal);
878 /* Calculate temporary vectorial force */
879 tx = _mm256_mul_pd(fscal,dx00);
880 ty = _mm256_mul_pd(fscal,dy00);
881 tz = _mm256_mul_pd(fscal,dz00);
883 /* Update vectorial force */
884 fix0 = _mm256_add_pd(fix0,tx);
885 fiy0 = _mm256_add_pd(fiy0,ty);
886 fiz0 = _mm256_add_pd(fiz0,tz);
888 fjx0 = _mm256_add_pd(fjx0,tx);
889 fjy0 = _mm256_add_pd(fjy0,ty);
890 fjz0 = _mm256_add_pd(fjz0,tz);
892 /**************************
893 * CALCULATE INTERACTIONS *
894 **************************/
896 r11 = _mm256_mul_pd(rsq11,rinv11);
897 r11 = _mm256_andnot_pd(dummy_mask,r11);
899 /* EWALD ELECTROSTATICS */
901 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
902 ewrt = _mm256_mul_pd(r11,ewtabscale);
903 ewitab = _mm256_cvttpd_epi32(ewrt);
904 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
905 ewitab = _mm_slli_epi32(ewitab,2);
906 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
907 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
908 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
909 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
910 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
911 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
912 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
913 velec = _mm256_mul_pd(qq11,_mm256_sub_pd(rinv11,velec));
914 felec = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
916 /* Update potential sum for this i atom from the interaction with this j atom. */
917 velec = _mm256_andnot_pd(dummy_mask,velec);
918 velecsum = _mm256_add_pd(velecsum,velec);
922 fscal = _mm256_andnot_pd(dummy_mask,fscal);
924 /* Calculate temporary vectorial force */
925 tx = _mm256_mul_pd(fscal,dx11);
926 ty = _mm256_mul_pd(fscal,dy11);
927 tz = _mm256_mul_pd(fscal,dz11);
929 /* Update vectorial force */
930 fix1 = _mm256_add_pd(fix1,tx);
931 fiy1 = _mm256_add_pd(fiy1,ty);
932 fiz1 = _mm256_add_pd(fiz1,tz);
934 fjx1 = _mm256_add_pd(fjx1,tx);
935 fjy1 = _mm256_add_pd(fjy1,ty);
936 fjz1 = _mm256_add_pd(fjz1,tz);
938 /**************************
939 * CALCULATE INTERACTIONS *
940 **************************/
942 r12 = _mm256_mul_pd(rsq12,rinv12);
943 r12 = _mm256_andnot_pd(dummy_mask,r12);
945 /* EWALD ELECTROSTATICS */
947 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
948 ewrt = _mm256_mul_pd(r12,ewtabscale);
949 ewitab = _mm256_cvttpd_epi32(ewrt);
950 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
951 ewitab = _mm_slli_epi32(ewitab,2);
952 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
953 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
954 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
955 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
956 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
957 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
958 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
959 velec = _mm256_mul_pd(qq12,_mm256_sub_pd(rinv12,velec));
960 felec = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
962 /* Update potential sum for this i atom from the interaction with this j atom. */
963 velec = _mm256_andnot_pd(dummy_mask,velec);
964 velecsum = _mm256_add_pd(velecsum,velec);
968 fscal = _mm256_andnot_pd(dummy_mask,fscal);
970 /* Calculate temporary vectorial force */
971 tx = _mm256_mul_pd(fscal,dx12);
972 ty = _mm256_mul_pd(fscal,dy12);
973 tz = _mm256_mul_pd(fscal,dz12);
975 /* Update vectorial force */
976 fix1 = _mm256_add_pd(fix1,tx);
977 fiy1 = _mm256_add_pd(fiy1,ty);
978 fiz1 = _mm256_add_pd(fiz1,tz);
980 fjx2 = _mm256_add_pd(fjx2,tx);
981 fjy2 = _mm256_add_pd(fjy2,ty);
982 fjz2 = _mm256_add_pd(fjz2,tz);
984 /**************************
985 * CALCULATE INTERACTIONS *
986 **************************/
988 r13 = _mm256_mul_pd(rsq13,rinv13);
989 r13 = _mm256_andnot_pd(dummy_mask,r13);
991 /* EWALD ELECTROSTATICS */
993 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
994 ewrt = _mm256_mul_pd(r13,ewtabscale);
995 ewitab = _mm256_cvttpd_epi32(ewrt);
996 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
997 ewitab = _mm_slli_epi32(ewitab,2);
998 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
999 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1000 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1001 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1002 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1003 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1004 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1005 velec = _mm256_mul_pd(qq13,_mm256_sub_pd(rinv13,velec));
1006 felec = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
1008 /* Update potential sum for this i atom from the interaction with this j atom. */
1009 velec = _mm256_andnot_pd(dummy_mask,velec);
1010 velecsum = _mm256_add_pd(velecsum,velec);
1014 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1016 /* Calculate temporary vectorial force */
1017 tx = _mm256_mul_pd(fscal,dx13);
1018 ty = _mm256_mul_pd(fscal,dy13);
1019 tz = _mm256_mul_pd(fscal,dz13);
1021 /* Update vectorial force */
1022 fix1 = _mm256_add_pd(fix1,tx);
1023 fiy1 = _mm256_add_pd(fiy1,ty);
1024 fiz1 = _mm256_add_pd(fiz1,tz);
1026 fjx3 = _mm256_add_pd(fjx3,tx);
1027 fjy3 = _mm256_add_pd(fjy3,ty);
1028 fjz3 = _mm256_add_pd(fjz3,tz);
1030 /**************************
1031 * CALCULATE INTERACTIONS *
1032 **************************/
1034 r21 = _mm256_mul_pd(rsq21,rinv21);
1035 r21 = _mm256_andnot_pd(dummy_mask,r21);
1037 /* EWALD ELECTROSTATICS */
1039 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1040 ewrt = _mm256_mul_pd(r21,ewtabscale);
1041 ewitab = _mm256_cvttpd_epi32(ewrt);
1042 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1043 ewitab = _mm_slli_epi32(ewitab,2);
1044 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1045 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1046 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1047 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1048 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1049 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1050 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1051 velec = _mm256_mul_pd(qq21,_mm256_sub_pd(rinv21,velec));
1052 felec = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1054 /* Update potential sum for this i atom from the interaction with this j atom. */
1055 velec = _mm256_andnot_pd(dummy_mask,velec);
1056 velecsum = _mm256_add_pd(velecsum,velec);
1060 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1062 /* Calculate temporary vectorial force */
1063 tx = _mm256_mul_pd(fscal,dx21);
1064 ty = _mm256_mul_pd(fscal,dy21);
1065 tz = _mm256_mul_pd(fscal,dz21);
1067 /* Update vectorial force */
1068 fix2 = _mm256_add_pd(fix2,tx);
1069 fiy2 = _mm256_add_pd(fiy2,ty);
1070 fiz2 = _mm256_add_pd(fiz2,tz);
1072 fjx1 = _mm256_add_pd(fjx1,tx);
1073 fjy1 = _mm256_add_pd(fjy1,ty);
1074 fjz1 = _mm256_add_pd(fjz1,tz);
1076 /**************************
1077 * CALCULATE INTERACTIONS *
1078 **************************/
1080 r22 = _mm256_mul_pd(rsq22,rinv22);
1081 r22 = _mm256_andnot_pd(dummy_mask,r22);
1083 /* EWALD ELECTROSTATICS */
1085 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1086 ewrt = _mm256_mul_pd(r22,ewtabscale);
1087 ewitab = _mm256_cvttpd_epi32(ewrt);
1088 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1089 ewitab = _mm_slli_epi32(ewitab,2);
1090 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1091 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1092 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1093 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1094 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1095 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1096 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1097 velec = _mm256_mul_pd(qq22,_mm256_sub_pd(rinv22,velec));
1098 felec = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1100 /* Update potential sum for this i atom from the interaction with this j atom. */
1101 velec = _mm256_andnot_pd(dummy_mask,velec);
1102 velecsum = _mm256_add_pd(velecsum,velec);
1106 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1108 /* Calculate temporary vectorial force */
1109 tx = _mm256_mul_pd(fscal,dx22);
1110 ty = _mm256_mul_pd(fscal,dy22);
1111 tz = _mm256_mul_pd(fscal,dz22);
1113 /* Update vectorial force */
1114 fix2 = _mm256_add_pd(fix2,tx);
1115 fiy2 = _mm256_add_pd(fiy2,ty);
1116 fiz2 = _mm256_add_pd(fiz2,tz);
1118 fjx2 = _mm256_add_pd(fjx2,tx);
1119 fjy2 = _mm256_add_pd(fjy2,ty);
1120 fjz2 = _mm256_add_pd(fjz2,tz);
1122 /**************************
1123 * CALCULATE INTERACTIONS *
1124 **************************/
1126 r23 = _mm256_mul_pd(rsq23,rinv23);
1127 r23 = _mm256_andnot_pd(dummy_mask,r23);
1129 /* EWALD ELECTROSTATICS */
1131 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1132 ewrt = _mm256_mul_pd(r23,ewtabscale);
1133 ewitab = _mm256_cvttpd_epi32(ewrt);
1134 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1135 ewitab = _mm_slli_epi32(ewitab,2);
1136 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1137 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1138 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1139 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1140 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1141 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1142 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1143 velec = _mm256_mul_pd(qq23,_mm256_sub_pd(rinv23,velec));
1144 felec = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
1146 /* Update potential sum for this i atom from the interaction with this j atom. */
1147 velec = _mm256_andnot_pd(dummy_mask,velec);
1148 velecsum = _mm256_add_pd(velecsum,velec);
1152 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1154 /* Calculate temporary vectorial force */
1155 tx = _mm256_mul_pd(fscal,dx23);
1156 ty = _mm256_mul_pd(fscal,dy23);
1157 tz = _mm256_mul_pd(fscal,dz23);
1159 /* Update vectorial force */
1160 fix2 = _mm256_add_pd(fix2,tx);
1161 fiy2 = _mm256_add_pd(fiy2,ty);
1162 fiz2 = _mm256_add_pd(fiz2,tz);
1164 fjx3 = _mm256_add_pd(fjx3,tx);
1165 fjy3 = _mm256_add_pd(fjy3,ty);
1166 fjz3 = _mm256_add_pd(fjz3,tz);
1168 /**************************
1169 * CALCULATE INTERACTIONS *
1170 **************************/
1172 r31 = _mm256_mul_pd(rsq31,rinv31);
1173 r31 = _mm256_andnot_pd(dummy_mask,r31);
1175 /* EWALD ELECTROSTATICS */
1177 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1178 ewrt = _mm256_mul_pd(r31,ewtabscale);
1179 ewitab = _mm256_cvttpd_epi32(ewrt);
1180 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1181 ewitab = _mm_slli_epi32(ewitab,2);
1182 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1183 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1184 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1185 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1186 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1187 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1188 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1189 velec = _mm256_mul_pd(qq31,_mm256_sub_pd(rinv31,velec));
1190 felec = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
1192 /* Update potential sum for this i atom from the interaction with this j atom. */
1193 velec = _mm256_andnot_pd(dummy_mask,velec);
1194 velecsum = _mm256_add_pd(velecsum,velec);
1198 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1200 /* Calculate temporary vectorial force */
1201 tx = _mm256_mul_pd(fscal,dx31);
1202 ty = _mm256_mul_pd(fscal,dy31);
1203 tz = _mm256_mul_pd(fscal,dz31);
1205 /* Update vectorial force */
1206 fix3 = _mm256_add_pd(fix3,tx);
1207 fiy3 = _mm256_add_pd(fiy3,ty);
1208 fiz3 = _mm256_add_pd(fiz3,tz);
1210 fjx1 = _mm256_add_pd(fjx1,tx);
1211 fjy1 = _mm256_add_pd(fjy1,ty);
1212 fjz1 = _mm256_add_pd(fjz1,tz);
1214 /**************************
1215 * CALCULATE INTERACTIONS *
1216 **************************/
1218 r32 = _mm256_mul_pd(rsq32,rinv32);
1219 r32 = _mm256_andnot_pd(dummy_mask,r32);
1221 /* EWALD ELECTROSTATICS */
1223 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1224 ewrt = _mm256_mul_pd(r32,ewtabscale);
1225 ewitab = _mm256_cvttpd_epi32(ewrt);
1226 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1227 ewitab = _mm_slli_epi32(ewitab,2);
1228 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1229 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1230 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1231 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1232 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1233 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1234 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1235 velec = _mm256_mul_pd(qq32,_mm256_sub_pd(rinv32,velec));
1236 felec = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
1238 /* Update potential sum for this i atom from the interaction with this j atom. */
1239 velec = _mm256_andnot_pd(dummy_mask,velec);
1240 velecsum = _mm256_add_pd(velecsum,velec);
1244 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1246 /* Calculate temporary vectorial force */
1247 tx = _mm256_mul_pd(fscal,dx32);
1248 ty = _mm256_mul_pd(fscal,dy32);
1249 tz = _mm256_mul_pd(fscal,dz32);
1251 /* Update vectorial force */
1252 fix3 = _mm256_add_pd(fix3,tx);
1253 fiy3 = _mm256_add_pd(fiy3,ty);
1254 fiz3 = _mm256_add_pd(fiz3,tz);
1256 fjx2 = _mm256_add_pd(fjx2,tx);
1257 fjy2 = _mm256_add_pd(fjy2,ty);
1258 fjz2 = _mm256_add_pd(fjz2,tz);
1260 /**************************
1261 * CALCULATE INTERACTIONS *
1262 **************************/
1264 r33 = _mm256_mul_pd(rsq33,rinv33);
1265 r33 = _mm256_andnot_pd(dummy_mask,r33);
1267 /* EWALD ELECTROSTATICS */
1269 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1270 ewrt = _mm256_mul_pd(r33,ewtabscale);
1271 ewitab = _mm256_cvttpd_epi32(ewrt);
1272 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1273 ewitab = _mm_slli_epi32(ewitab,2);
1274 ewtabF = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
1275 ewtabD = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
1276 ewtabV = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,2) );
1277 ewtabFn = _mm256_load_pd( ewtab + _mm_extract_epi32(ewitab,3) );
1278 GMX_MM256_FULLTRANSPOSE4_PD(ewtabF,ewtabD,ewtabV,ewtabFn);
1279 felec = _mm256_add_pd(ewtabF,_mm256_mul_pd(eweps,ewtabD));
1280 velec = _mm256_sub_pd(ewtabV,_mm256_mul_pd(_mm256_mul_pd(ewtabhalfspace,eweps),_mm256_add_pd(ewtabF,felec)));
1281 velec = _mm256_mul_pd(qq33,_mm256_sub_pd(rinv33,velec));
1282 felec = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
1284 /* Update potential sum for this i atom from the interaction with this j atom. */
1285 velec = _mm256_andnot_pd(dummy_mask,velec);
1286 velecsum = _mm256_add_pd(velecsum,velec);
1290 fscal = _mm256_andnot_pd(dummy_mask,fscal);
1292 /* Calculate temporary vectorial force */
1293 tx = _mm256_mul_pd(fscal,dx33);
1294 ty = _mm256_mul_pd(fscal,dy33);
1295 tz = _mm256_mul_pd(fscal,dz33);
1297 /* Update vectorial force */
1298 fix3 = _mm256_add_pd(fix3,tx);
1299 fiy3 = _mm256_add_pd(fiy3,ty);
1300 fiz3 = _mm256_add_pd(fiz3,tz);
1302 fjx3 = _mm256_add_pd(fjx3,tx);
1303 fjy3 = _mm256_add_pd(fjy3,ty);
1304 fjz3 = _mm256_add_pd(fjz3,tz);
1306 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
1307 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
1308 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
1309 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
1311 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1312 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1313 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1315 /* Inner loop uses 413 flops */
1318 /* End of innermost loop */
1320 gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1321 f+i_coord_offset,fshift+i_shift_offset);
1324 /* Update potential energies */
1325 gmx_mm256_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1326 gmx_mm256_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1328 /* Increment number of inner iterations */
1329 inneriter += j_index_end - j_index_start;
1331 /* Outer loop uses 26 flops */
1334 /* Increment number of outer iterations */
1337 /* Update outer/inner flops */
1339 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*413);
1342 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJ_GeomW4W4_F_avx_256_double
1343 * Electrostatics interaction: Ewald
1344 * VdW interaction: LennardJones
1345 * Geometry: Water4-Water4
1346 * Calculate force/pot: Force
1349 nb_kernel_ElecEw_VdwLJ_GeomW4W4_F_avx_256_double
1350 (t_nblist * gmx_restrict nlist,
1351 rvec * gmx_restrict xx,
1352 rvec * gmx_restrict ff,
1353 t_forcerec * gmx_restrict fr,
1354 t_mdatoms * gmx_restrict mdatoms,
1355 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1356 t_nrnb * gmx_restrict nrnb)
1358 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1359 * just 0 for non-waters.
1360 * Suffixes A,B,C,D refer to j loop unrolling done with AVX, e.g. for the four different
1361 * jnr indices corresponding to data put in the four positions in the SIMD register.
1363 int i_shift_offset,i_coord_offset,outeriter,inneriter;
1364 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1365 int jnrA,jnrB,jnrC,jnrD;
1366 int jnrlistA,jnrlistB,jnrlistC,jnrlistD;
1367 int jnrlistE,jnrlistF,jnrlistG,jnrlistH;
1368 int j_coord_offsetA,j_coord_offsetB,j_coord_offsetC,j_coord_offsetD;
1369 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
1370 real rcutoff_scalar;
1371 real *shiftvec,*fshift,*x,*f;
1372 real *fjptrA,*fjptrB,*fjptrC,*fjptrD;
1373 real scratch[4*DIM];
1374 __m256d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1375 real * vdwioffsetptr0;
1376 __m256d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1377 real * vdwioffsetptr1;
1378 __m256d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1379 real * vdwioffsetptr2;
1380 __m256d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1381 real * vdwioffsetptr3;
1382 __m256d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1383 int vdwjidx0A,vdwjidx0B,vdwjidx0C,vdwjidx0D;
1384 __m256d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1385 int vdwjidx1A,vdwjidx1B,vdwjidx1C,vdwjidx1D;
1386 __m256d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1387 int vdwjidx2A,vdwjidx2B,vdwjidx2C,vdwjidx2D;
1388 __m256d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1389 int vdwjidx3A,vdwjidx3B,vdwjidx3C,vdwjidx3D;
1390 __m256d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1391 __m256d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1392 __m256d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1393 __m256d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1394 __m256d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1395 __m256d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1396 __m256d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1397 __m256d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1398 __m256d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1399 __m256d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1400 __m256d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1401 __m256d velec,felec,velecsum,facel,crf,krf,krf2;
1404 __m256d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1407 __m256d one_sixth = _mm256_set1_pd(1.0/6.0);
1408 __m256d one_twelfth = _mm256_set1_pd(1.0/12.0);
1410 __m256d ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1411 __m256d beta,beta2,beta3,zeta2,pmecorrF,pmecorrV,rinv3;
1413 __m256d dummy_mask,cutoff_mask;
1414 __m128 tmpmask0,tmpmask1;
1415 __m256d signbit = _mm256_castsi256_pd( _mm256_set1_epi32(0x80000000) );
1416 __m256d one = _mm256_set1_pd(1.0);
1417 __m256d two = _mm256_set1_pd(2.0);
1423 jindex = nlist->jindex;
1425 shiftidx = nlist->shift;
1427 shiftvec = fr->shift_vec[0];
1428 fshift = fr->fshift[0];
1429 facel = _mm256_set1_pd(fr->epsfac);
1430 charge = mdatoms->chargeA;
1431 nvdwtype = fr->ntype;
1432 vdwparam = fr->nbfp;
1433 vdwtype = mdatoms->typeA;
1435 sh_ewald = _mm256_set1_pd(fr->ic->sh_ewald);
1436 beta = _mm256_set1_pd(fr->ic->ewaldcoeff_q);
1437 beta2 = _mm256_mul_pd(beta,beta);
1438 beta3 = _mm256_mul_pd(beta,beta2);
1440 ewtab = fr->ic->tabq_coul_F;
1441 ewtabscale = _mm256_set1_pd(fr->ic->tabq_scale);
1442 ewtabhalfspace = _mm256_set1_pd(0.5/fr->ic->tabq_scale);
1444 /* Setup water-specific parameters */
1445 inr = nlist->iinr[0];
1446 iq1 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+1]));
1447 iq2 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+2]));
1448 iq3 = _mm256_mul_pd(facel,_mm256_set1_pd(charge[inr+3]));
1449 vdwioffsetptr0 = vdwparam+2*nvdwtype*vdwtype[inr+0];
1451 jq1 = _mm256_set1_pd(charge[inr+1]);
1452 jq2 = _mm256_set1_pd(charge[inr+2]);
1453 jq3 = _mm256_set1_pd(charge[inr+3]);
1454 vdwjidx0A = 2*vdwtype[inr+0];
1455 c6_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A]);
1456 c12_00 = _mm256_set1_pd(vdwioffsetptr0[vdwjidx0A+1]);
1457 qq11 = _mm256_mul_pd(iq1,jq1);
1458 qq12 = _mm256_mul_pd(iq1,jq2);
1459 qq13 = _mm256_mul_pd(iq1,jq3);
1460 qq21 = _mm256_mul_pd(iq2,jq1);
1461 qq22 = _mm256_mul_pd(iq2,jq2);
1462 qq23 = _mm256_mul_pd(iq2,jq3);
1463 qq31 = _mm256_mul_pd(iq3,jq1);
1464 qq32 = _mm256_mul_pd(iq3,jq2);
1465 qq33 = _mm256_mul_pd(iq3,jq3);
1467 /* Avoid stupid compiler warnings */
1468 jnrA = jnrB = jnrC = jnrD = 0;
1469 j_coord_offsetA = 0;
1470 j_coord_offsetB = 0;
1471 j_coord_offsetC = 0;
1472 j_coord_offsetD = 0;
1477 for(iidx=0;iidx<4*DIM;iidx++)
1479 scratch[iidx] = 0.0;
1482 /* Start outer loop over neighborlists */
1483 for(iidx=0; iidx<nri; iidx++)
1485 /* Load shift vector for this list */
1486 i_shift_offset = DIM*shiftidx[iidx];
1488 /* Load limits for loop over neighbors */
1489 j_index_start = jindex[iidx];
1490 j_index_end = jindex[iidx+1];
1492 /* Get outer coordinate index */
1494 i_coord_offset = DIM*inr;
1496 /* Load i particle coords and add shift vector */
1497 gmx_mm256_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1498 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1500 fix0 = _mm256_setzero_pd();
1501 fiy0 = _mm256_setzero_pd();
1502 fiz0 = _mm256_setzero_pd();
1503 fix1 = _mm256_setzero_pd();
1504 fiy1 = _mm256_setzero_pd();
1505 fiz1 = _mm256_setzero_pd();
1506 fix2 = _mm256_setzero_pd();
1507 fiy2 = _mm256_setzero_pd();
1508 fiz2 = _mm256_setzero_pd();
1509 fix3 = _mm256_setzero_pd();
1510 fiy3 = _mm256_setzero_pd();
1511 fiz3 = _mm256_setzero_pd();
1513 /* Start inner kernel loop */
1514 for(jidx=j_index_start; jidx<j_index_end && jjnr[jidx+3]>=0; jidx+=4)
1517 /* Get j neighbor index, and coordinate index */
1519 jnrB = jjnr[jidx+1];
1520 jnrC = jjnr[jidx+2];
1521 jnrD = jjnr[jidx+3];
1522 j_coord_offsetA = DIM*jnrA;
1523 j_coord_offsetB = DIM*jnrB;
1524 j_coord_offsetC = DIM*jnrC;
1525 j_coord_offsetD = DIM*jnrD;
1527 /* load j atom coordinates */
1528 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1529 x+j_coord_offsetC,x+j_coord_offsetD,
1530 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1531 &jy2,&jz2,&jx3,&jy3,&jz3);
1533 /* Calculate displacement vector */
1534 dx00 = _mm256_sub_pd(ix0,jx0);
1535 dy00 = _mm256_sub_pd(iy0,jy0);
1536 dz00 = _mm256_sub_pd(iz0,jz0);
1537 dx11 = _mm256_sub_pd(ix1,jx1);
1538 dy11 = _mm256_sub_pd(iy1,jy1);
1539 dz11 = _mm256_sub_pd(iz1,jz1);
1540 dx12 = _mm256_sub_pd(ix1,jx2);
1541 dy12 = _mm256_sub_pd(iy1,jy2);
1542 dz12 = _mm256_sub_pd(iz1,jz2);
1543 dx13 = _mm256_sub_pd(ix1,jx3);
1544 dy13 = _mm256_sub_pd(iy1,jy3);
1545 dz13 = _mm256_sub_pd(iz1,jz3);
1546 dx21 = _mm256_sub_pd(ix2,jx1);
1547 dy21 = _mm256_sub_pd(iy2,jy1);
1548 dz21 = _mm256_sub_pd(iz2,jz1);
1549 dx22 = _mm256_sub_pd(ix2,jx2);
1550 dy22 = _mm256_sub_pd(iy2,jy2);
1551 dz22 = _mm256_sub_pd(iz2,jz2);
1552 dx23 = _mm256_sub_pd(ix2,jx3);
1553 dy23 = _mm256_sub_pd(iy2,jy3);
1554 dz23 = _mm256_sub_pd(iz2,jz3);
1555 dx31 = _mm256_sub_pd(ix3,jx1);
1556 dy31 = _mm256_sub_pd(iy3,jy1);
1557 dz31 = _mm256_sub_pd(iz3,jz1);
1558 dx32 = _mm256_sub_pd(ix3,jx2);
1559 dy32 = _mm256_sub_pd(iy3,jy2);
1560 dz32 = _mm256_sub_pd(iz3,jz2);
1561 dx33 = _mm256_sub_pd(ix3,jx3);
1562 dy33 = _mm256_sub_pd(iy3,jy3);
1563 dz33 = _mm256_sub_pd(iz3,jz3);
1565 /* Calculate squared distance and things based on it */
1566 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
1567 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
1568 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
1569 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
1570 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
1571 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
1572 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
1573 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
1574 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
1575 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
1577 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
1578 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
1579 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
1580 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
1581 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
1582 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
1583 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
1584 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
1585 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
1587 rinvsq00 = gmx_mm256_inv_pd(rsq00);
1588 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
1589 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
1590 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
1591 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
1592 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
1593 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
1594 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
1595 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
1596 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
1598 fjx0 = _mm256_setzero_pd();
1599 fjy0 = _mm256_setzero_pd();
1600 fjz0 = _mm256_setzero_pd();
1601 fjx1 = _mm256_setzero_pd();
1602 fjy1 = _mm256_setzero_pd();
1603 fjz1 = _mm256_setzero_pd();
1604 fjx2 = _mm256_setzero_pd();
1605 fjy2 = _mm256_setzero_pd();
1606 fjz2 = _mm256_setzero_pd();
1607 fjx3 = _mm256_setzero_pd();
1608 fjy3 = _mm256_setzero_pd();
1609 fjz3 = _mm256_setzero_pd();
1611 /**************************
1612 * CALCULATE INTERACTIONS *
1613 **************************/
1615 /* LENNARD-JONES DISPERSION/REPULSION */
1617 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1618 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
1622 /* Calculate temporary vectorial force */
1623 tx = _mm256_mul_pd(fscal,dx00);
1624 ty = _mm256_mul_pd(fscal,dy00);
1625 tz = _mm256_mul_pd(fscal,dz00);
1627 /* Update vectorial force */
1628 fix0 = _mm256_add_pd(fix0,tx);
1629 fiy0 = _mm256_add_pd(fiy0,ty);
1630 fiz0 = _mm256_add_pd(fiz0,tz);
1632 fjx0 = _mm256_add_pd(fjx0,tx);
1633 fjy0 = _mm256_add_pd(fjy0,ty);
1634 fjz0 = _mm256_add_pd(fjz0,tz);
1636 /**************************
1637 * CALCULATE INTERACTIONS *
1638 **************************/
1640 r11 = _mm256_mul_pd(rsq11,rinv11);
1642 /* EWALD ELECTROSTATICS */
1644 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1645 ewrt = _mm256_mul_pd(r11,ewtabscale);
1646 ewitab = _mm256_cvttpd_epi32(ewrt);
1647 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1648 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1649 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1651 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1652 felec = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
1656 /* Calculate temporary vectorial force */
1657 tx = _mm256_mul_pd(fscal,dx11);
1658 ty = _mm256_mul_pd(fscal,dy11);
1659 tz = _mm256_mul_pd(fscal,dz11);
1661 /* Update vectorial force */
1662 fix1 = _mm256_add_pd(fix1,tx);
1663 fiy1 = _mm256_add_pd(fiy1,ty);
1664 fiz1 = _mm256_add_pd(fiz1,tz);
1666 fjx1 = _mm256_add_pd(fjx1,tx);
1667 fjy1 = _mm256_add_pd(fjy1,ty);
1668 fjz1 = _mm256_add_pd(fjz1,tz);
1670 /**************************
1671 * CALCULATE INTERACTIONS *
1672 **************************/
1674 r12 = _mm256_mul_pd(rsq12,rinv12);
1676 /* EWALD ELECTROSTATICS */
1678 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1679 ewrt = _mm256_mul_pd(r12,ewtabscale);
1680 ewitab = _mm256_cvttpd_epi32(ewrt);
1681 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1682 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1683 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1685 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1686 felec = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
1690 /* Calculate temporary vectorial force */
1691 tx = _mm256_mul_pd(fscal,dx12);
1692 ty = _mm256_mul_pd(fscal,dy12);
1693 tz = _mm256_mul_pd(fscal,dz12);
1695 /* Update vectorial force */
1696 fix1 = _mm256_add_pd(fix1,tx);
1697 fiy1 = _mm256_add_pd(fiy1,ty);
1698 fiz1 = _mm256_add_pd(fiz1,tz);
1700 fjx2 = _mm256_add_pd(fjx2,tx);
1701 fjy2 = _mm256_add_pd(fjy2,ty);
1702 fjz2 = _mm256_add_pd(fjz2,tz);
1704 /**************************
1705 * CALCULATE INTERACTIONS *
1706 **************************/
1708 r13 = _mm256_mul_pd(rsq13,rinv13);
1710 /* EWALD ELECTROSTATICS */
1712 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1713 ewrt = _mm256_mul_pd(r13,ewtabscale);
1714 ewitab = _mm256_cvttpd_epi32(ewrt);
1715 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1716 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1717 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1719 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1720 felec = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
1724 /* Calculate temporary vectorial force */
1725 tx = _mm256_mul_pd(fscal,dx13);
1726 ty = _mm256_mul_pd(fscal,dy13);
1727 tz = _mm256_mul_pd(fscal,dz13);
1729 /* Update vectorial force */
1730 fix1 = _mm256_add_pd(fix1,tx);
1731 fiy1 = _mm256_add_pd(fiy1,ty);
1732 fiz1 = _mm256_add_pd(fiz1,tz);
1734 fjx3 = _mm256_add_pd(fjx3,tx);
1735 fjy3 = _mm256_add_pd(fjy3,ty);
1736 fjz3 = _mm256_add_pd(fjz3,tz);
1738 /**************************
1739 * CALCULATE INTERACTIONS *
1740 **************************/
1742 r21 = _mm256_mul_pd(rsq21,rinv21);
1744 /* EWALD ELECTROSTATICS */
1746 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1747 ewrt = _mm256_mul_pd(r21,ewtabscale);
1748 ewitab = _mm256_cvttpd_epi32(ewrt);
1749 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1750 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1751 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1753 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1754 felec = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
1758 /* Calculate temporary vectorial force */
1759 tx = _mm256_mul_pd(fscal,dx21);
1760 ty = _mm256_mul_pd(fscal,dy21);
1761 tz = _mm256_mul_pd(fscal,dz21);
1763 /* Update vectorial force */
1764 fix2 = _mm256_add_pd(fix2,tx);
1765 fiy2 = _mm256_add_pd(fiy2,ty);
1766 fiz2 = _mm256_add_pd(fiz2,tz);
1768 fjx1 = _mm256_add_pd(fjx1,tx);
1769 fjy1 = _mm256_add_pd(fjy1,ty);
1770 fjz1 = _mm256_add_pd(fjz1,tz);
1772 /**************************
1773 * CALCULATE INTERACTIONS *
1774 **************************/
1776 r22 = _mm256_mul_pd(rsq22,rinv22);
1778 /* EWALD ELECTROSTATICS */
1780 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1781 ewrt = _mm256_mul_pd(r22,ewtabscale);
1782 ewitab = _mm256_cvttpd_epi32(ewrt);
1783 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1784 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1785 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1787 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1788 felec = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
1792 /* Calculate temporary vectorial force */
1793 tx = _mm256_mul_pd(fscal,dx22);
1794 ty = _mm256_mul_pd(fscal,dy22);
1795 tz = _mm256_mul_pd(fscal,dz22);
1797 /* Update vectorial force */
1798 fix2 = _mm256_add_pd(fix2,tx);
1799 fiy2 = _mm256_add_pd(fiy2,ty);
1800 fiz2 = _mm256_add_pd(fiz2,tz);
1802 fjx2 = _mm256_add_pd(fjx2,tx);
1803 fjy2 = _mm256_add_pd(fjy2,ty);
1804 fjz2 = _mm256_add_pd(fjz2,tz);
1806 /**************************
1807 * CALCULATE INTERACTIONS *
1808 **************************/
1810 r23 = _mm256_mul_pd(rsq23,rinv23);
1812 /* EWALD ELECTROSTATICS */
1814 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1815 ewrt = _mm256_mul_pd(r23,ewtabscale);
1816 ewitab = _mm256_cvttpd_epi32(ewrt);
1817 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1818 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1819 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1821 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1822 felec = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
1826 /* Calculate temporary vectorial force */
1827 tx = _mm256_mul_pd(fscal,dx23);
1828 ty = _mm256_mul_pd(fscal,dy23);
1829 tz = _mm256_mul_pd(fscal,dz23);
1831 /* Update vectorial force */
1832 fix2 = _mm256_add_pd(fix2,tx);
1833 fiy2 = _mm256_add_pd(fiy2,ty);
1834 fiz2 = _mm256_add_pd(fiz2,tz);
1836 fjx3 = _mm256_add_pd(fjx3,tx);
1837 fjy3 = _mm256_add_pd(fjy3,ty);
1838 fjz3 = _mm256_add_pd(fjz3,tz);
1840 /**************************
1841 * CALCULATE INTERACTIONS *
1842 **************************/
1844 r31 = _mm256_mul_pd(rsq31,rinv31);
1846 /* EWALD ELECTROSTATICS */
1848 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1849 ewrt = _mm256_mul_pd(r31,ewtabscale);
1850 ewitab = _mm256_cvttpd_epi32(ewrt);
1851 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1852 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1853 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1855 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1856 felec = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
1860 /* Calculate temporary vectorial force */
1861 tx = _mm256_mul_pd(fscal,dx31);
1862 ty = _mm256_mul_pd(fscal,dy31);
1863 tz = _mm256_mul_pd(fscal,dz31);
1865 /* Update vectorial force */
1866 fix3 = _mm256_add_pd(fix3,tx);
1867 fiy3 = _mm256_add_pd(fiy3,ty);
1868 fiz3 = _mm256_add_pd(fiz3,tz);
1870 fjx1 = _mm256_add_pd(fjx1,tx);
1871 fjy1 = _mm256_add_pd(fjy1,ty);
1872 fjz1 = _mm256_add_pd(fjz1,tz);
1874 /**************************
1875 * CALCULATE INTERACTIONS *
1876 **************************/
1878 r32 = _mm256_mul_pd(rsq32,rinv32);
1880 /* EWALD ELECTROSTATICS */
1882 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1883 ewrt = _mm256_mul_pd(r32,ewtabscale);
1884 ewitab = _mm256_cvttpd_epi32(ewrt);
1885 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1886 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1887 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1889 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1890 felec = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
1894 /* Calculate temporary vectorial force */
1895 tx = _mm256_mul_pd(fscal,dx32);
1896 ty = _mm256_mul_pd(fscal,dy32);
1897 tz = _mm256_mul_pd(fscal,dz32);
1899 /* Update vectorial force */
1900 fix3 = _mm256_add_pd(fix3,tx);
1901 fiy3 = _mm256_add_pd(fiy3,ty);
1902 fiz3 = _mm256_add_pd(fiz3,tz);
1904 fjx2 = _mm256_add_pd(fjx2,tx);
1905 fjy2 = _mm256_add_pd(fjy2,ty);
1906 fjz2 = _mm256_add_pd(fjz2,tz);
1908 /**************************
1909 * CALCULATE INTERACTIONS *
1910 **************************/
1912 r33 = _mm256_mul_pd(rsq33,rinv33);
1914 /* EWALD ELECTROSTATICS */
1916 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1917 ewrt = _mm256_mul_pd(r33,ewtabscale);
1918 ewitab = _mm256_cvttpd_epi32(ewrt);
1919 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
1920 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
1921 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
1923 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
1924 felec = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
1928 /* Calculate temporary vectorial force */
1929 tx = _mm256_mul_pd(fscal,dx33);
1930 ty = _mm256_mul_pd(fscal,dy33);
1931 tz = _mm256_mul_pd(fscal,dz33);
1933 /* Update vectorial force */
1934 fix3 = _mm256_add_pd(fix3,tx);
1935 fiy3 = _mm256_add_pd(fiy3,ty);
1936 fiz3 = _mm256_add_pd(fiz3,tz);
1938 fjx3 = _mm256_add_pd(fjx3,tx);
1939 fjy3 = _mm256_add_pd(fjy3,ty);
1940 fjz3 = _mm256_add_pd(fjz3,tz);
1942 fjptrA = f+j_coord_offsetA;
1943 fjptrB = f+j_coord_offsetB;
1944 fjptrC = f+j_coord_offsetC;
1945 fjptrD = f+j_coord_offsetD;
1947 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
1948 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
1949 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1951 /* Inner loop uses 354 flops */
1954 if(jidx<j_index_end)
1957 /* Get j neighbor index, and coordinate index */
1958 jnrlistA = jjnr[jidx];
1959 jnrlistB = jjnr[jidx+1];
1960 jnrlistC = jjnr[jidx+2];
1961 jnrlistD = jjnr[jidx+3];
1962 /* Sign of each element will be negative for non-real atoms.
1963 * This mask will be 0xFFFFFFFF for dummy entries and 0x0 for real ones,
1964 * so use it as val = _mm_andnot_pd(mask,val) to clear dummy entries.
1966 tmpmask0 = gmx_mm_castsi128_ps(_mm_cmplt_epi32(_mm_loadu_si128((const __m128i *)(jjnr+jidx)),_mm_setzero_si128()));
1968 tmpmask1 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(3,3,2,2));
1969 tmpmask0 = _mm_permute_ps(tmpmask0,_GMX_MM_PERMUTE(1,1,0,0));
1970 dummy_mask = _mm256_castps_pd(gmx_mm256_set_m128(tmpmask1,tmpmask0));
1972 jnrA = (jnrlistA>=0) ? jnrlistA : 0;
1973 jnrB = (jnrlistB>=0) ? jnrlistB : 0;
1974 jnrC = (jnrlistC>=0) ? jnrlistC : 0;
1975 jnrD = (jnrlistD>=0) ? jnrlistD : 0;
1976 j_coord_offsetA = DIM*jnrA;
1977 j_coord_offsetB = DIM*jnrB;
1978 j_coord_offsetC = DIM*jnrC;
1979 j_coord_offsetD = DIM*jnrD;
1981 /* load j atom coordinates */
1982 gmx_mm256_load_4rvec_4ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1983 x+j_coord_offsetC,x+j_coord_offsetD,
1984 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1985 &jy2,&jz2,&jx3,&jy3,&jz3);
1987 /* Calculate displacement vector */
1988 dx00 = _mm256_sub_pd(ix0,jx0);
1989 dy00 = _mm256_sub_pd(iy0,jy0);
1990 dz00 = _mm256_sub_pd(iz0,jz0);
1991 dx11 = _mm256_sub_pd(ix1,jx1);
1992 dy11 = _mm256_sub_pd(iy1,jy1);
1993 dz11 = _mm256_sub_pd(iz1,jz1);
1994 dx12 = _mm256_sub_pd(ix1,jx2);
1995 dy12 = _mm256_sub_pd(iy1,jy2);
1996 dz12 = _mm256_sub_pd(iz1,jz2);
1997 dx13 = _mm256_sub_pd(ix1,jx3);
1998 dy13 = _mm256_sub_pd(iy1,jy3);
1999 dz13 = _mm256_sub_pd(iz1,jz3);
2000 dx21 = _mm256_sub_pd(ix2,jx1);
2001 dy21 = _mm256_sub_pd(iy2,jy1);
2002 dz21 = _mm256_sub_pd(iz2,jz1);
2003 dx22 = _mm256_sub_pd(ix2,jx2);
2004 dy22 = _mm256_sub_pd(iy2,jy2);
2005 dz22 = _mm256_sub_pd(iz2,jz2);
2006 dx23 = _mm256_sub_pd(ix2,jx3);
2007 dy23 = _mm256_sub_pd(iy2,jy3);
2008 dz23 = _mm256_sub_pd(iz2,jz3);
2009 dx31 = _mm256_sub_pd(ix3,jx1);
2010 dy31 = _mm256_sub_pd(iy3,jy1);
2011 dz31 = _mm256_sub_pd(iz3,jz1);
2012 dx32 = _mm256_sub_pd(ix3,jx2);
2013 dy32 = _mm256_sub_pd(iy3,jy2);
2014 dz32 = _mm256_sub_pd(iz3,jz2);
2015 dx33 = _mm256_sub_pd(ix3,jx3);
2016 dy33 = _mm256_sub_pd(iy3,jy3);
2017 dz33 = _mm256_sub_pd(iz3,jz3);
2019 /* Calculate squared distance and things based on it */
2020 rsq00 = gmx_mm256_calc_rsq_pd(dx00,dy00,dz00);
2021 rsq11 = gmx_mm256_calc_rsq_pd(dx11,dy11,dz11);
2022 rsq12 = gmx_mm256_calc_rsq_pd(dx12,dy12,dz12);
2023 rsq13 = gmx_mm256_calc_rsq_pd(dx13,dy13,dz13);
2024 rsq21 = gmx_mm256_calc_rsq_pd(dx21,dy21,dz21);
2025 rsq22 = gmx_mm256_calc_rsq_pd(dx22,dy22,dz22);
2026 rsq23 = gmx_mm256_calc_rsq_pd(dx23,dy23,dz23);
2027 rsq31 = gmx_mm256_calc_rsq_pd(dx31,dy31,dz31);
2028 rsq32 = gmx_mm256_calc_rsq_pd(dx32,dy32,dz32);
2029 rsq33 = gmx_mm256_calc_rsq_pd(dx33,dy33,dz33);
2031 rinv11 = gmx_mm256_invsqrt_pd(rsq11);
2032 rinv12 = gmx_mm256_invsqrt_pd(rsq12);
2033 rinv13 = gmx_mm256_invsqrt_pd(rsq13);
2034 rinv21 = gmx_mm256_invsqrt_pd(rsq21);
2035 rinv22 = gmx_mm256_invsqrt_pd(rsq22);
2036 rinv23 = gmx_mm256_invsqrt_pd(rsq23);
2037 rinv31 = gmx_mm256_invsqrt_pd(rsq31);
2038 rinv32 = gmx_mm256_invsqrt_pd(rsq32);
2039 rinv33 = gmx_mm256_invsqrt_pd(rsq33);
2041 rinvsq00 = gmx_mm256_inv_pd(rsq00);
2042 rinvsq11 = _mm256_mul_pd(rinv11,rinv11);
2043 rinvsq12 = _mm256_mul_pd(rinv12,rinv12);
2044 rinvsq13 = _mm256_mul_pd(rinv13,rinv13);
2045 rinvsq21 = _mm256_mul_pd(rinv21,rinv21);
2046 rinvsq22 = _mm256_mul_pd(rinv22,rinv22);
2047 rinvsq23 = _mm256_mul_pd(rinv23,rinv23);
2048 rinvsq31 = _mm256_mul_pd(rinv31,rinv31);
2049 rinvsq32 = _mm256_mul_pd(rinv32,rinv32);
2050 rinvsq33 = _mm256_mul_pd(rinv33,rinv33);
2052 fjx0 = _mm256_setzero_pd();
2053 fjy0 = _mm256_setzero_pd();
2054 fjz0 = _mm256_setzero_pd();
2055 fjx1 = _mm256_setzero_pd();
2056 fjy1 = _mm256_setzero_pd();
2057 fjz1 = _mm256_setzero_pd();
2058 fjx2 = _mm256_setzero_pd();
2059 fjy2 = _mm256_setzero_pd();
2060 fjz2 = _mm256_setzero_pd();
2061 fjx3 = _mm256_setzero_pd();
2062 fjy3 = _mm256_setzero_pd();
2063 fjz3 = _mm256_setzero_pd();
2065 /**************************
2066 * CALCULATE INTERACTIONS *
2067 **************************/
2069 /* LENNARD-JONES DISPERSION/REPULSION */
2071 rinvsix = _mm256_mul_pd(_mm256_mul_pd(rinvsq00,rinvsq00),rinvsq00);
2072 fvdw = _mm256_mul_pd(_mm256_sub_pd(_mm256_mul_pd(c12_00,rinvsix),c6_00),_mm256_mul_pd(rinvsix,rinvsq00));
2076 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2078 /* Calculate temporary vectorial force */
2079 tx = _mm256_mul_pd(fscal,dx00);
2080 ty = _mm256_mul_pd(fscal,dy00);
2081 tz = _mm256_mul_pd(fscal,dz00);
2083 /* Update vectorial force */
2084 fix0 = _mm256_add_pd(fix0,tx);
2085 fiy0 = _mm256_add_pd(fiy0,ty);
2086 fiz0 = _mm256_add_pd(fiz0,tz);
2088 fjx0 = _mm256_add_pd(fjx0,tx);
2089 fjy0 = _mm256_add_pd(fjy0,ty);
2090 fjz0 = _mm256_add_pd(fjz0,tz);
2092 /**************************
2093 * CALCULATE INTERACTIONS *
2094 **************************/
2096 r11 = _mm256_mul_pd(rsq11,rinv11);
2097 r11 = _mm256_andnot_pd(dummy_mask,r11);
2099 /* EWALD ELECTROSTATICS */
2101 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2102 ewrt = _mm256_mul_pd(r11,ewtabscale);
2103 ewitab = _mm256_cvttpd_epi32(ewrt);
2104 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2105 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2106 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2108 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2109 felec = _mm256_mul_pd(_mm256_mul_pd(qq11,rinv11),_mm256_sub_pd(rinvsq11,felec));
2113 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2115 /* Calculate temporary vectorial force */
2116 tx = _mm256_mul_pd(fscal,dx11);
2117 ty = _mm256_mul_pd(fscal,dy11);
2118 tz = _mm256_mul_pd(fscal,dz11);
2120 /* Update vectorial force */
2121 fix1 = _mm256_add_pd(fix1,tx);
2122 fiy1 = _mm256_add_pd(fiy1,ty);
2123 fiz1 = _mm256_add_pd(fiz1,tz);
2125 fjx1 = _mm256_add_pd(fjx1,tx);
2126 fjy1 = _mm256_add_pd(fjy1,ty);
2127 fjz1 = _mm256_add_pd(fjz1,tz);
2129 /**************************
2130 * CALCULATE INTERACTIONS *
2131 **************************/
2133 r12 = _mm256_mul_pd(rsq12,rinv12);
2134 r12 = _mm256_andnot_pd(dummy_mask,r12);
2136 /* EWALD ELECTROSTATICS */
2138 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2139 ewrt = _mm256_mul_pd(r12,ewtabscale);
2140 ewitab = _mm256_cvttpd_epi32(ewrt);
2141 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2142 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2143 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2145 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2146 felec = _mm256_mul_pd(_mm256_mul_pd(qq12,rinv12),_mm256_sub_pd(rinvsq12,felec));
2150 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2152 /* Calculate temporary vectorial force */
2153 tx = _mm256_mul_pd(fscal,dx12);
2154 ty = _mm256_mul_pd(fscal,dy12);
2155 tz = _mm256_mul_pd(fscal,dz12);
2157 /* Update vectorial force */
2158 fix1 = _mm256_add_pd(fix1,tx);
2159 fiy1 = _mm256_add_pd(fiy1,ty);
2160 fiz1 = _mm256_add_pd(fiz1,tz);
2162 fjx2 = _mm256_add_pd(fjx2,tx);
2163 fjy2 = _mm256_add_pd(fjy2,ty);
2164 fjz2 = _mm256_add_pd(fjz2,tz);
2166 /**************************
2167 * CALCULATE INTERACTIONS *
2168 **************************/
2170 r13 = _mm256_mul_pd(rsq13,rinv13);
2171 r13 = _mm256_andnot_pd(dummy_mask,r13);
2173 /* EWALD ELECTROSTATICS */
2175 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2176 ewrt = _mm256_mul_pd(r13,ewtabscale);
2177 ewitab = _mm256_cvttpd_epi32(ewrt);
2178 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2179 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2180 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2182 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2183 felec = _mm256_mul_pd(_mm256_mul_pd(qq13,rinv13),_mm256_sub_pd(rinvsq13,felec));
2187 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2189 /* Calculate temporary vectorial force */
2190 tx = _mm256_mul_pd(fscal,dx13);
2191 ty = _mm256_mul_pd(fscal,dy13);
2192 tz = _mm256_mul_pd(fscal,dz13);
2194 /* Update vectorial force */
2195 fix1 = _mm256_add_pd(fix1,tx);
2196 fiy1 = _mm256_add_pd(fiy1,ty);
2197 fiz1 = _mm256_add_pd(fiz1,tz);
2199 fjx3 = _mm256_add_pd(fjx3,tx);
2200 fjy3 = _mm256_add_pd(fjy3,ty);
2201 fjz3 = _mm256_add_pd(fjz3,tz);
2203 /**************************
2204 * CALCULATE INTERACTIONS *
2205 **************************/
2207 r21 = _mm256_mul_pd(rsq21,rinv21);
2208 r21 = _mm256_andnot_pd(dummy_mask,r21);
2210 /* EWALD ELECTROSTATICS */
2212 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2213 ewrt = _mm256_mul_pd(r21,ewtabscale);
2214 ewitab = _mm256_cvttpd_epi32(ewrt);
2215 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2216 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2217 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2219 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2220 felec = _mm256_mul_pd(_mm256_mul_pd(qq21,rinv21),_mm256_sub_pd(rinvsq21,felec));
2224 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2226 /* Calculate temporary vectorial force */
2227 tx = _mm256_mul_pd(fscal,dx21);
2228 ty = _mm256_mul_pd(fscal,dy21);
2229 tz = _mm256_mul_pd(fscal,dz21);
2231 /* Update vectorial force */
2232 fix2 = _mm256_add_pd(fix2,tx);
2233 fiy2 = _mm256_add_pd(fiy2,ty);
2234 fiz2 = _mm256_add_pd(fiz2,tz);
2236 fjx1 = _mm256_add_pd(fjx1,tx);
2237 fjy1 = _mm256_add_pd(fjy1,ty);
2238 fjz1 = _mm256_add_pd(fjz1,tz);
2240 /**************************
2241 * CALCULATE INTERACTIONS *
2242 **************************/
2244 r22 = _mm256_mul_pd(rsq22,rinv22);
2245 r22 = _mm256_andnot_pd(dummy_mask,r22);
2247 /* EWALD ELECTROSTATICS */
2249 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2250 ewrt = _mm256_mul_pd(r22,ewtabscale);
2251 ewitab = _mm256_cvttpd_epi32(ewrt);
2252 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2253 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2254 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2256 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2257 felec = _mm256_mul_pd(_mm256_mul_pd(qq22,rinv22),_mm256_sub_pd(rinvsq22,felec));
2261 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2263 /* Calculate temporary vectorial force */
2264 tx = _mm256_mul_pd(fscal,dx22);
2265 ty = _mm256_mul_pd(fscal,dy22);
2266 tz = _mm256_mul_pd(fscal,dz22);
2268 /* Update vectorial force */
2269 fix2 = _mm256_add_pd(fix2,tx);
2270 fiy2 = _mm256_add_pd(fiy2,ty);
2271 fiz2 = _mm256_add_pd(fiz2,tz);
2273 fjx2 = _mm256_add_pd(fjx2,tx);
2274 fjy2 = _mm256_add_pd(fjy2,ty);
2275 fjz2 = _mm256_add_pd(fjz2,tz);
2277 /**************************
2278 * CALCULATE INTERACTIONS *
2279 **************************/
2281 r23 = _mm256_mul_pd(rsq23,rinv23);
2282 r23 = _mm256_andnot_pd(dummy_mask,r23);
2284 /* EWALD ELECTROSTATICS */
2286 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2287 ewrt = _mm256_mul_pd(r23,ewtabscale);
2288 ewitab = _mm256_cvttpd_epi32(ewrt);
2289 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2290 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2291 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2293 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2294 felec = _mm256_mul_pd(_mm256_mul_pd(qq23,rinv23),_mm256_sub_pd(rinvsq23,felec));
2298 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2300 /* Calculate temporary vectorial force */
2301 tx = _mm256_mul_pd(fscal,dx23);
2302 ty = _mm256_mul_pd(fscal,dy23);
2303 tz = _mm256_mul_pd(fscal,dz23);
2305 /* Update vectorial force */
2306 fix2 = _mm256_add_pd(fix2,tx);
2307 fiy2 = _mm256_add_pd(fiy2,ty);
2308 fiz2 = _mm256_add_pd(fiz2,tz);
2310 fjx3 = _mm256_add_pd(fjx3,tx);
2311 fjy3 = _mm256_add_pd(fjy3,ty);
2312 fjz3 = _mm256_add_pd(fjz3,tz);
2314 /**************************
2315 * CALCULATE INTERACTIONS *
2316 **************************/
2318 r31 = _mm256_mul_pd(rsq31,rinv31);
2319 r31 = _mm256_andnot_pd(dummy_mask,r31);
2321 /* EWALD ELECTROSTATICS */
2323 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2324 ewrt = _mm256_mul_pd(r31,ewtabscale);
2325 ewitab = _mm256_cvttpd_epi32(ewrt);
2326 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2327 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2328 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2330 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2331 felec = _mm256_mul_pd(_mm256_mul_pd(qq31,rinv31),_mm256_sub_pd(rinvsq31,felec));
2335 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2337 /* Calculate temporary vectorial force */
2338 tx = _mm256_mul_pd(fscal,dx31);
2339 ty = _mm256_mul_pd(fscal,dy31);
2340 tz = _mm256_mul_pd(fscal,dz31);
2342 /* Update vectorial force */
2343 fix3 = _mm256_add_pd(fix3,tx);
2344 fiy3 = _mm256_add_pd(fiy3,ty);
2345 fiz3 = _mm256_add_pd(fiz3,tz);
2347 fjx1 = _mm256_add_pd(fjx1,tx);
2348 fjy1 = _mm256_add_pd(fjy1,ty);
2349 fjz1 = _mm256_add_pd(fjz1,tz);
2351 /**************************
2352 * CALCULATE INTERACTIONS *
2353 **************************/
2355 r32 = _mm256_mul_pd(rsq32,rinv32);
2356 r32 = _mm256_andnot_pd(dummy_mask,r32);
2358 /* EWALD ELECTROSTATICS */
2360 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2361 ewrt = _mm256_mul_pd(r32,ewtabscale);
2362 ewitab = _mm256_cvttpd_epi32(ewrt);
2363 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2364 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2365 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2367 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2368 felec = _mm256_mul_pd(_mm256_mul_pd(qq32,rinv32),_mm256_sub_pd(rinvsq32,felec));
2372 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2374 /* Calculate temporary vectorial force */
2375 tx = _mm256_mul_pd(fscal,dx32);
2376 ty = _mm256_mul_pd(fscal,dy32);
2377 tz = _mm256_mul_pd(fscal,dz32);
2379 /* Update vectorial force */
2380 fix3 = _mm256_add_pd(fix3,tx);
2381 fiy3 = _mm256_add_pd(fiy3,ty);
2382 fiz3 = _mm256_add_pd(fiz3,tz);
2384 fjx2 = _mm256_add_pd(fjx2,tx);
2385 fjy2 = _mm256_add_pd(fjy2,ty);
2386 fjz2 = _mm256_add_pd(fjz2,tz);
2388 /**************************
2389 * CALCULATE INTERACTIONS *
2390 **************************/
2392 r33 = _mm256_mul_pd(rsq33,rinv33);
2393 r33 = _mm256_andnot_pd(dummy_mask,r33);
2395 /* EWALD ELECTROSTATICS */
2397 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2398 ewrt = _mm256_mul_pd(r33,ewtabscale);
2399 ewitab = _mm256_cvttpd_epi32(ewrt);
2400 eweps = _mm256_sub_pd(ewrt,_mm256_round_pd(ewrt, _MM_FROUND_FLOOR));
2401 gmx_mm256_load_4pair_swizzle_pd(ewtab + _mm_extract_epi32(ewitab,0),ewtab + _mm_extract_epi32(ewitab,1),
2402 ewtab + _mm_extract_epi32(ewitab,2),ewtab + _mm_extract_epi32(ewitab,3),
2404 felec = _mm256_add_pd(_mm256_mul_pd( _mm256_sub_pd(one,eweps),ewtabF),_mm256_mul_pd(eweps,ewtabFn));
2405 felec = _mm256_mul_pd(_mm256_mul_pd(qq33,rinv33),_mm256_sub_pd(rinvsq33,felec));
2409 fscal = _mm256_andnot_pd(dummy_mask,fscal);
2411 /* Calculate temporary vectorial force */
2412 tx = _mm256_mul_pd(fscal,dx33);
2413 ty = _mm256_mul_pd(fscal,dy33);
2414 tz = _mm256_mul_pd(fscal,dz33);
2416 /* Update vectorial force */
2417 fix3 = _mm256_add_pd(fix3,tx);
2418 fiy3 = _mm256_add_pd(fiy3,ty);
2419 fiz3 = _mm256_add_pd(fiz3,tz);
2421 fjx3 = _mm256_add_pd(fjx3,tx);
2422 fjy3 = _mm256_add_pd(fjy3,ty);
2423 fjz3 = _mm256_add_pd(fjz3,tz);
2425 fjptrA = (jnrlistA>=0) ? f+j_coord_offsetA : scratch;
2426 fjptrB = (jnrlistB>=0) ? f+j_coord_offsetB : scratch;
2427 fjptrC = (jnrlistC>=0) ? f+j_coord_offsetC : scratch;
2428 fjptrD = (jnrlistD>=0) ? f+j_coord_offsetD : scratch;
2430 gmx_mm256_decrement_4rvec_4ptr_swizzle_pd(fjptrA,fjptrB,fjptrC,fjptrD,
2431 fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,
2432 fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2434 /* Inner loop uses 363 flops */
2437 /* End of innermost loop */
2439 gmx_mm256_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
2440 f+i_coord_offset,fshift+i_shift_offset);
2442 /* Increment number of inner iterations */
2443 inneriter += j_index_end - j_index_start;
2445 /* Outer loop uses 24 flops */
2448 /* Increment number of outer iterations */
2451 /* Update outer/inner flops */
2453 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*363);