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 sparc64_hpc_ace_double kernel generator.
42 #include "../nb_kernel.h"
43 #include "types/simple.h"
44 #include "gromacs/math/vec.h"
47 #include "kernelutil_sparc64_hpc_ace_double.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJEw_GeomW4P1_VF_sparc64_hpc_ace_double
51 * Electrostatics interaction: Ewald
52 * VdW interaction: LJEwald
53 * Geometry: Water4-Particle
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecEw_VdwLJEw_GeomW4P1_VF_sparc64_hpc_ace_double
58 (t_nblist * gmx_restrict nlist,
59 rvec * gmx_restrict xx,
60 rvec * gmx_restrict ff,
61 t_forcerec * gmx_restrict fr,
62 t_mdatoms * gmx_restrict mdatoms,
63 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
64 t_nrnb * gmx_restrict nrnb)
66 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
67 * just 0 for non-waters.
68 * Suffixes A,B refer to j loop unrolling done with double precision SIMD, e.g. for the two different
69 * jnr indices corresponding to data put in the four positions in the SIMD register.
71 int i_shift_offset,i_coord_offset,outeriter,inneriter;
72 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
74 int j_coord_offsetA,j_coord_offsetB;
75 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
77 real *shiftvec,*fshift,*x,*f;
78 _fjsp_v2r8 tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
80 _fjsp_v2r8 ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
82 _fjsp_v2r8 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
84 _fjsp_v2r8 ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
86 _fjsp_v2r8 ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
87 int vdwjidx0A,vdwjidx0B;
88 _fjsp_v2r8 jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
89 _fjsp_v2r8 dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
90 _fjsp_v2r8 dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
91 _fjsp_v2r8 dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
92 _fjsp_v2r8 dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
93 _fjsp_v2r8 velec,felec,velecsum,facel,crf,krf,krf2;
96 _fjsp_v2r8 rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
99 _fjsp_v2r8 one_sixth = gmx_fjsp_set1_v2r8(1.0/6.0);
100 _fjsp_v2r8 one_twelfth = gmx_fjsp_set1_v2r8(1.0/12.0);
101 _fjsp_v2r8 c6grid_00;
102 _fjsp_v2r8 c6grid_10;
103 _fjsp_v2r8 c6grid_20;
104 _fjsp_v2r8 c6grid_30;
106 _fjsp_v2r8 ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
107 _fjsp_v2r8 one_half = gmx_fjsp_set1_v2r8(0.5);
108 _fjsp_v2r8 minus_one = gmx_fjsp_set1_v2r8(-1.0);
109 _fjsp_v2r8 ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
112 _fjsp_v2r8 dummy_mask,cutoff_mask;
113 _fjsp_v2r8 one = gmx_fjsp_set1_v2r8(1.0);
114 _fjsp_v2r8 two = gmx_fjsp_set1_v2r8(2.0);
115 union { _fjsp_v2r8 simd; long long int i[2]; } vfconv,gbconv,ewconv;
122 jindex = nlist->jindex;
124 shiftidx = nlist->shift;
126 shiftvec = fr->shift_vec[0];
127 fshift = fr->fshift[0];
128 facel = gmx_fjsp_set1_v2r8(fr->epsfac);
129 charge = mdatoms->chargeA;
130 nvdwtype = fr->ntype;
132 vdwtype = mdatoms->typeA;
133 vdwgridparam = fr->ljpme_c6grid;
134 sh_lj_ewald = gmx_fjsp_set1_v2r8(fr->ic->sh_lj_ewald);
135 ewclj = gmx_fjsp_set1_v2r8(fr->ewaldcoeff_lj);
136 ewclj2 = _fjsp_mul_v2r8(minus_one,_fjsp_mul_v2r8(ewclj,ewclj));
138 sh_ewald = gmx_fjsp_set1_v2r8(fr->ic->sh_ewald);
139 ewtab = fr->ic->tabq_coul_FDV0;
140 ewtabscale = gmx_fjsp_set1_v2r8(fr->ic->tabq_scale);
141 ewtabhalfspace = gmx_fjsp_set1_v2r8(0.5/fr->ic->tabq_scale);
143 /* Setup water-specific parameters */
144 inr = nlist->iinr[0];
145 iq1 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+1]));
146 iq2 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+2]));
147 iq3 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+3]));
148 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
150 /* Avoid stupid compiler warnings */
158 /* Start outer loop over neighborlists */
159 for(iidx=0; iidx<nri; iidx++)
161 /* Load shift vector for this list */
162 i_shift_offset = DIM*shiftidx[iidx];
164 /* Load limits for loop over neighbors */
165 j_index_start = jindex[iidx];
166 j_index_end = jindex[iidx+1];
168 /* Get outer coordinate index */
170 i_coord_offset = DIM*inr;
172 /* Load i particle coords and add shift vector */
173 gmx_fjsp_load_shift_and_4rvec_broadcast_v2r8(shiftvec+i_shift_offset,x+i_coord_offset,
174 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
176 fix0 = _fjsp_setzero_v2r8();
177 fiy0 = _fjsp_setzero_v2r8();
178 fiz0 = _fjsp_setzero_v2r8();
179 fix1 = _fjsp_setzero_v2r8();
180 fiy1 = _fjsp_setzero_v2r8();
181 fiz1 = _fjsp_setzero_v2r8();
182 fix2 = _fjsp_setzero_v2r8();
183 fiy2 = _fjsp_setzero_v2r8();
184 fiz2 = _fjsp_setzero_v2r8();
185 fix3 = _fjsp_setzero_v2r8();
186 fiy3 = _fjsp_setzero_v2r8();
187 fiz3 = _fjsp_setzero_v2r8();
189 /* Reset potential sums */
190 velecsum = _fjsp_setzero_v2r8();
191 vvdwsum = _fjsp_setzero_v2r8();
193 /* Start inner kernel loop */
194 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
197 /* Get j neighbor index, and coordinate index */
200 j_coord_offsetA = DIM*jnrA;
201 j_coord_offsetB = DIM*jnrB;
203 /* load j atom coordinates */
204 gmx_fjsp_load_1rvec_2ptr_swizzle_v2r8(x+j_coord_offsetA,x+j_coord_offsetB,
207 /* Calculate displacement vector */
208 dx00 = _fjsp_sub_v2r8(ix0,jx0);
209 dy00 = _fjsp_sub_v2r8(iy0,jy0);
210 dz00 = _fjsp_sub_v2r8(iz0,jz0);
211 dx10 = _fjsp_sub_v2r8(ix1,jx0);
212 dy10 = _fjsp_sub_v2r8(iy1,jy0);
213 dz10 = _fjsp_sub_v2r8(iz1,jz0);
214 dx20 = _fjsp_sub_v2r8(ix2,jx0);
215 dy20 = _fjsp_sub_v2r8(iy2,jy0);
216 dz20 = _fjsp_sub_v2r8(iz2,jz0);
217 dx30 = _fjsp_sub_v2r8(ix3,jx0);
218 dy30 = _fjsp_sub_v2r8(iy3,jy0);
219 dz30 = _fjsp_sub_v2r8(iz3,jz0);
221 /* Calculate squared distance and things based on it */
222 rsq00 = gmx_fjsp_calc_rsq_v2r8(dx00,dy00,dz00);
223 rsq10 = gmx_fjsp_calc_rsq_v2r8(dx10,dy10,dz10);
224 rsq20 = gmx_fjsp_calc_rsq_v2r8(dx20,dy20,dz20);
225 rsq30 = gmx_fjsp_calc_rsq_v2r8(dx30,dy30,dz30);
227 rinv00 = gmx_fjsp_invsqrt_v2r8(rsq00);
228 rinv10 = gmx_fjsp_invsqrt_v2r8(rsq10);
229 rinv20 = gmx_fjsp_invsqrt_v2r8(rsq20);
230 rinv30 = gmx_fjsp_invsqrt_v2r8(rsq30);
232 rinvsq00 = _fjsp_mul_v2r8(rinv00,rinv00);
233 rinvsq10 = _fjsp_mul_v2r8(rinv10,rinv10);
234 rinvsq20 = _fjsp_mul_v2r8(rinv20,rinv20);
235 rinvsq30 = _fjsp_mul_v2r8(rinv30,rinv30);
237 /* Load parameters for j particles */
238 jq0 = gmx_fjsp_load_2real_swizzle_v2r8(charge+jnrA+0,charge+jnrB+0);
239 vdwjidx0A = 2*vdwtype[jnrA+0];
240 vdwjidx0B = 2*vdwtype[jnrB+0];
242 fjx0 = _fjsp_setzero_v2r8();
243 fjy0 = _fjsp_setzero_v2r8();
244 fjz0 = _fjsp_setzero_v2r8();
246 /**************************
247 * CALCULATE INTERACTIONS *
248 **************************/
250 r00 = _fjsp_mul_v2r8(rsq00,rinv00);
252 /* Compute parameters for interactions between i and j atoms */
253 gmx_fjsp_load_2pair_swizzle_v2r8(vdwparam+vdwioffset0+vdwjidx0A,
254 vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
256 c6grid_00 = gmx_fjsp_load_2real_swizzle_v2r8(vdwgridparam+vdwioffset0+vdwjidx0A,
257 vdwgridparam+vdwioffset0+vdwjidx0B);
259 /* Analytical LJ-PME */
260 rinvsix = _fjsp_mul_v2r8(_fjsp_mul_v2r8(rinvsq00,rinvsq00),rinvsq00);
261 ewcljrsq = _fjsp_mul_v2r8(ewclj2,rsq00);
262 ewclj6 = _fjsp_mul_v2r8(ewclj2,_fjsp_mul_v2r8(ewclj2,ewclj2));
263 exponent = gmx_simd_exp_d(-ewcljrsq);
264 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
265 poly = _fjsp_mul_v2r8(exponent,_fjsp_madd_v2r8(_fjsp_mul_v2r8(ewcljrsq,ewcljrsq),one_half,_fjsp_sub_v2r8(one,ewcljrsq)));
266 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
267 vvdw6 = _fjsp_mul_v2r8(_fjsp_madd_v2r8(-c6grid_00,_fjsp_sub_v2r8(one,poly),c6_00),rinvsix);
268 vvdw12 = _fjsp_mul_v2r8(c12_00,_fjsp_mul_v2r8(rinvsix,rinvsix));
269 vvdw = _fjsp_msub_v2r8(vvdw12,one_twelfth,_fjsp_mul_v2r8(vvdw6,one_sixth));
270 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
271 fvdw = _fjsp_mul_v2r8(_fjsp_add_v2r8(vvdw12,_fjsp_msub_v2r8(_fjsp_mul_v2r8(c6grid_00,one_sixth),_fjsp_mul_v2r8(exponent,ewclj6),vvdw6)),rinvsq00);
273 /* Update potential sum for this i atom from the interaction with this j atom. */
274 vvdwsum = _fjsp_add_v2r8(vvdwsum,vvdw);
278 /* Update vectorial force */
279 fix0 = _fjsp_madd_v2r8(dx00,fscal,fix0);
280 fiy0 = _fjsp_madd_v2r8(dy00,fscal,fiy0);
281 fiz0 = _fjsp_madd_v2r8(dz00,fscal,fiz0);
283 fjx0 = _fjsp_madd_v2r8(dx00,fscal,fjx0);
284 fjy0 = _fjsp_madd_v2r8(dy00,fscal,fjy0);
285 fjz0 = _fjsp_madd_v2r8(dz00,fscal,fjz0);
287 /**************************
288 * CALCULATE INTERACTIONS *
289 **************************/
291 r10 = _fjsp_mul_v2r8(rsq10,rinv10);
293 /* Compute parameters for interactions between i and j atoms */
294 qq10 = _fjsp_mul_v2r8(iq1,jq0);
296 /* EWALD ELECTROSTATICS */
298 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
299 ewrt = _fjsp_mul_v2r8(r10,ewtabscale);
300 itab_tmp = _fjsp_dtox_v2r8(ewrt);
301 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
302 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
304 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
305 ewtabD = _fjsp_load_v2r8( ewtab + 4*ewconv.i[1] );
306 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
307 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
308 ewtabFn = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[1] +2);
309 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
310 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
311 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
312 velec = _fjsp_mul_v2r8(qq10,_fjsp_sub_v2r8(rinv10,velec));
313 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq10,rinv10),_fjsp_sub_v2r8(rinvsq10,felec));
315 /* Update potential sum for this i atom from the interaction with this j atom. */
316 velecsum = _fjsp_add_v2r8(velecsum,velec);
320 /* Update vectorial force */
321 fix1 = _fjsp_madd_v2r8(dx10,fscal,fix1);
322 fiy1 = _fjsp_madd_v2r8(dy10,fscal,fiy1);
323 fiz1 = _fjsp_madd_v2r8(dz10,fscal,fiz1);
325 fjx0 = _fjsp_madd_v2r8(dx10,fscal,fjx0);
326 fjy0 = _fjsp_madd_v2r8(dy10,fscal,fjy0);
327 fjz0 = _fjsp_madd_v2r8(dz10,fscal,fjz0);
329 /**************************
330 * CALCULATE INTERACTIONS *
331 **************************/
333 r20 = _fjsp_mul_v2r8(rsq20,rinv20);
335 /* Compute parameters for interactions between i and j atoms */
336 qq20 = _fjsp_mul_v2r8(iq2,jq0);
338 /* EWALD ELECTROSTATICS */
340 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
341 ewrt = _fjsp_mul_v2r8(r20,ewtabscale);
342 itab_tmp = _fjsp_dtox_v2r8(ewrt);
343 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
344 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
346 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
347 ewtabD = _fjsp_load_v2r8( ewtab + 4*ewconv.i[1] );
348 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
349 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
350 ewtabFn = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[1] +2);
351 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
352 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
353 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
354 velec = _fjsp_mul_v2r8(qq20,_fjsp_sub_v2r8(rinv20,velec));
355 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq20,rinv20),_fjsp_sub_v2r8(rinvsq20,felec));
357 /* Update potential sum for this i atom from the interaction with this j atom. */
358 velecsum = _fjsp_add_v2r8(velecsum,velec);
362 /* Update vectorial force */
363 fix2 = _fjsp_madd_v2r8(dx20,fscal,fix2);
364 fiy2 = _fjsp_madd_v2r8(dy20,fscal,fiy2);
365 fiz2 = _fjsp_madd_v2r8(dz20,fscal,fiz2);
367 fjx0 = _fjsp_madd_v2r8(dx20,fscal,fjx0);
368 fjy0 = _fjsp_madd_v2r8(dy20,fscal,fjy0);
369 fjz0 = _fjsp_madd_v2r8(dz20,fscal,fjz0);
371 /**************************
372 * CALCULATE INTERACTIONS *
373 **************************/
375 r30 = _fjsp_mul_v2r8(rsq30,rinv30);
377 /* Compute parameters for interactions between i and j atoms */
378 qq30 = _fjsp_mul_v2r8(iq3,jq0);
380 /* EWALD ELECTROSTATICS */
382 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
383 ewrt = _fjsp_mul_v2r8(r30,ewtabscale);
384 itab_tmp = _fjsp_dtox_v2r8(ewrt);
385 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
386 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
388 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
389 ewtabD = _fjsp_load_v2r8( ewtab + 4*ewconv.i[1] );
390 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
391 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
392 ewtabFn = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[1] +2);
393 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
394 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
395 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
396 velec = _fjsp_mul_v2r8(qq30,_fjsp_sub_v2r8(rinv30,velec));
397 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq30,rinv30),_fjsp_sub_v2r8(rinvsq30,felec));
399 /* Update potential sum for this i atom from the interaction with this j atom. */
400 velecsum = _fjsp_add_v2r8(velecsum,velec);
404 /* Update vectorial force */
405 fix3 = _fjsp_madd_v2r8(dx30,fscal,fix3);
406 fiy3 = _fjsp_madd_v2r8(dy30,fscal,fiy3);
407 fiz3 = _fjsp_madd_v2r8(dz30,fscal,fiz3);
409 fjx0 = _fjsp_madd_v2r8(dx30,fscal,fjx0);
410 fjy0 = _fjsp_madd_v2r8(dy30,fscal,fjy0);
411 fjz0 = _fjsp_madd_v2r8(dz30,fscal,fjz0);
413 gmx_fjsp_decrement_1rvec_2ptr_swizzle_v2r8(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
415 /* Inner loop uses 185 flops */
422 j_coord_offsetA = DIM*jnrA;
424 /* load j atom coordinates */
425 gmx_fjsp_load_1rvec_1ptr_swizzle_v2r8(x+j_coord_offsetA,
428 /* Calculate displacement vector */
429 dx00 = _fjsp_sub_v2r8(ix0,jx0);
430 dy00 = _fjsp_sub_v2r8(iy0,jy0);
431 dz00 = _fjsp_sub_v2r8(iz0,jz0);
432 dx10 = _fjsp_sub_v2r8(ix1,jx0);
433 dy10 = _fjsp_sub_v2r8(iy1,jy0);
434 dz10 = _fjsp_sub_v2r8(iz1,jz0);
435 dx20 = _fjsp_sub_v2r8(ix2,jx0);
436 dy20 = _fjsp_sub_v2r8(iy2,jy0);
437 dz20 = _fjsp_sub_v2r8(iz2,jz0);
438 dx30 = _fjsp_sub_v2r8(ix3,jx0);
439 dy30 = _fjsp_sub_v2r8(iy3,jy0);
440 dz30 = _fjsp_sub_v2r8(iz3,jz0);
442 /* Calculate squared distance and things based on it */
443 rsq00 = gmx_fjsp_calc_rsq_v2r8(dx00,dy00,dz00);
444 rsq10 = gmx_fjsp_calc_rsq_v2r8(dx10,dy10,dz10);
445 rsq20 = gmx_fjsp_calc_rsq_v2r8(dx20,dy20,dz20);
446 rsq30 = gmx_fjsp_calc_rsq_v2r8(dx30,dy30,dz30);
448 rinv00 = gmx_fjsp_invsqrt_v2r8(rsq00);
449 rinv10 = gmx_fjsp_invsqrt_v2r8(rsq10);
450 rinv20 = gmx_fjsp_invsqrt_v2r8(rsq20);
451 rinv30 = gmx_fjsp_invsqrt_v2r8(rsq30);
453 rinvsq00 = _fjsp_mul_v2r8(rinv00,rinv00);
454 rinvsq10 = _fjsp_mul_v2r8(rinv10,rinv10);
455 rinvsq20 = _fjsp_mul_v2r8(rinv20,rinv20);
456 rinvsq30 = _fjsp_mul_v2r8(rinv30,rinv30);
458 /* Load parameters for j particles */
459 jq0 = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(),charge+jnrA+0);
460 vdwjidx0A = 2*vdwtype[jnrA+0];
462 fjx0 = _fjsp_setzero_v2r8();
463 fjy0 = _fjsp_setzero_v2r8();
464 fjz0 = _fjsp_setzero_v2r8();
466 /**************************
467 * CALCULATE INTERACTIONS *
468 **************************/
470 r00 = _fjsp_mul_v2r8(rsq00,rinv00);
472 /* Compute parameters for interactions between i and j atoms */
473 gmx_fjsp_load_1pair_swizzle_v2r8(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
475 c6grid_00 = gmx_fjsp_load_1real_swizzle_v2r8(vdwgridparam+vdwioffset0+vdwjidx0A);
477 /* Analytical LJ-PME */
478 rinvsix = _fjsp_mul_v2r8(_fjsp_mul_v2r8(rinvsq00,rinvsq00),rinvsq00);
479 ewcljrsq = _fjsp_mul_v2r8(ewclj2,rsq00);
480 ewclj6 = _fjsp_mul_v2r8(ewclj2,_fjsp_mul_v2r8(ewclj2,ewclj2));
481 exponent = gmx_simd_exp_d(-ewcljrsq);
482 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
483 poly = _fjsp_mul_v2r8(exponent,_fjsp_madd_v2r8(_fjsp_mul_v2r8(ewcljrsq,ewcljrsq),one_half,_fjsp_sub_v2r8(one,ewcljrsq)));
484 /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
485 vvdw6 = _fjsp_mul_v2r8(_fjsp_madd_v2r8(-c6grid_00,_fjsp_sub_v2r8(one,poly),c6_00),rinvsix);
486 vvdw12 = _fjsp_mul_v2r8(c12_00,_fjsp_mul_v2r8(rinvsix,rinvsix));
487 vvdw = _fjsp_msub_v2r8(vvdw12,one_twelfth,_fjsp_mul_v2r8(vvdw6,one_sixth));
488 /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
489 fvdw = _fjsp_mul_v2r8(_fjsp_add_v2r8(vvdw12,_fjsp_msub_v2r8(_fjsp_mul_v2r8(c6grid_00,one_sixth),_fjsp_mul_v2r8(exponent,ewclj6),vvdw6)),rinvsq00);
491 /* Update potential sum for this i atom from the interaction with this j atom. */
492 vvdw = _fjsp_unpacklo_v2r8(vvdw,_fjsp_setzero_v2r8());
493 vvdwsum = _fjsp_add_v2r8(vvdwsum,vvdw);
497 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
499 /* Update vectorial force */
500 fix0 = _fjsp_madd_v2r8(dx00,fscal,fix0);
501 fiy0 = _fjsp_madd_v2r8(dy00,fscal,fiy0);
502 fiz0 = _fjsp_madd_v2r8(dz00,fscal,fiz0);
504 fjx0 = _fjsp_madd_v2r8(dx00,fscal,fjx0);
505 fjy0 = _fjsp_madd_v2r8(dy00,fscal,fjy0);
506 fjz0 = _fjsp_madd_v2r8(dz00,fscal,fjz0);
508 /**************************
509 * CALCULATE INTERACTIONS *
510 **************************/
512 r10 = _fjsp_mul_v2r8(rsq10,rinv10);
514 /* Compute parameters for interactions between i and j atoms */
515 qq10 = _fjsp_mul_v2r8(iq1,jq0);
517 /* EWALD ELECTROSTATICS */
519 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
520 ewrt = _fjsp_mul_v2r8(r10,ewtabscale);
521 itab_tmp = _fjsp_dtox_v2r8(ewrt);
522 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
523 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
525 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
526 ewtabD = _fjsp_setzero_v2r8();
527 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
528 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
529 ewtabFn = _fjsp_setzero_v2r8();
530 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
531 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
532 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
533 velec = _fjsp_mul_v2r8(qq10,_fjsp_sub_v2r8(rinv10,velec));
534 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq10,rinv10),_fjsp_sub_v2r8(rinvsq10,felec));
536 /* Update potential sum for this i atom from the interaction with this j atom. */
537 velec = _fjsp_unpacklo_v2r8(velec,_fjsp_setzero_v2r8());
538 velecsum = _fjsp_add_v2r8(velecsum,velec);
542 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
544 /* Update vectorial force */
545 fix1 = _fjsp_madd_v2r8(dx10,fscal,fix1);
546 fiy1 = _fjsp_madd_v2r8(dy10,fscal,fiy1);
547 fiz1 = _fjsp_madd_v2r8(dz10,fscal,fiz1);
549 fjx0 = _fjsp_madd_v2r8(dx10,fscal,fjx0);
550 fjy0 = _fjsp_madd_v2r8(dy10,fscal,fjy0);
551 fjz0 = _fjsp_madd_v2r8(dz10,fscal,fjz0);
553 /**************************
554 * CALCULATE INTERACTIONS *
555 **************************/
557 r20 = _fjsp_mul_v2r8(rsq20,rinv20);
559 /* Compute parameters for interactions between i and j atoms */
560 qq20 = _fjsp_mul_v2r8(iq2,jq0);
562 /* EWALD ELECTROSTATICS */
564 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
565 ewrt = _fjsp_mul_v2r8(r20,ewtabscale);
566 itab_tmp = _fjsp_dtox_v2r8(ewrt);
567 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
568 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
570 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
571 ewtabD = _fjsp_setzero_v2r8();
572 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
573 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
574 ewtabFn = _fjsp_setzero_v2r8();
575 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
576 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
577 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
578 velec = _fjsp_mul_v2r8(qq20,_fjsp_sub_v2r8(rinv20,velec));
579 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq20,rinv20),_fjsp_sub_v2r8(rinvsq20,felec));
581 /* Update potential sum for this i atom from the interaction with this j atom. */
582 velec = _fjsp_unpacklo_v2r8(velec,_fjsp_setzero_v2r8());
583 velecsum = _fjsp_add_v2r8(velecsum,velec);
587 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
589 /* Update vectorial force */
590 fix2 = _fjsp_madd_v2r8(dx20,fscal,fix2);
591 fiy2 = _fjsp_madd_v2r8(dy20,fscal,fiy2);
592 fiz2 = _fjsp_madd_v2r8(dz20,fscal,fiz2);
594 fjx0 = _fjsp_madd_v2r8(dx20,fscal,fjx0);
595 fjy0 = _fjsp_madd_v2r8(dy20,fscal,fjy0);
596 fjz0 = _fjsp_madd_v2r8(dz20,fscal,fjz0);
598 /**************************
599 * CALCULATE INTERACTIONS *
600 **************************/
602 r30 = _fjsp_mul_v2r8(rsq30,rinv30);
604 /* Compute parameters for interactions between i and j atoms */
605 qq30 = _fjsp_mul_v2r8(iq3,jq0);
607 /* EWALD ELECTROSTATICS */
609 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
610 ewrt = _fjsp_mul_v2r8(r30,ewtabscale);
611 itab_tmp = _fjsp_dtox_v2r8(ewrt);
612 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
613 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
615 ewtabF = _fjsp_load_v2r8( ewtab + 4*ewconv.i[0] );
616 ewtabD = _fjsp_setzero_v2r8();
617 GMX_FJSP_TRANSPOSE2_V2R8(ewtabF,ewtabD);
618 ewtabV = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(), ewtab + 4*ewconv.i[0] +2);
619 ewtabFn = _fjsp_setzero_v2r8();
620 GMX_FJSP_TRANSPOSE2_V2R8(ewtabV,ewtabFn);
621 felec = _fjsp_madd_v2r8(eweps,ewtabD,ewtabF);
622 velec = _fjsp_nmsub_v2r8(_fjsp_mul_v2r8(ewtabhalfspace,eweps) ,_fjsp_add_v2r8(ewtabF,felec), ewtabV);
623 velec = _fjsp_mul_v2r8(qq30,_fjsp_sub_v2r8(rinv30,velec));
624 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq30,rinv30),_fjsp_sub_v2r8(rinvsq30,felec));
626 /* Update potential sum for this i atom from the interaction with this j atom. */
627 velec = _fjsp_unpacklo_v2r8(velec,_fjsp_setzero_v2r8());
628 velecsum = _fjsp_add_v2r8(velecsum,velec);
632 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
634 /* Update vectorial force */
635 fix3 = _fjsp_madd_v2r8(dx30,fscal,fix3);
636 fiy3 = _fjsp_madd_v2r8(dy30,fscal,fiy3);
637 fiz3 = _fjsp_madd_v2r8(dz30,fscal,fiz3);
639 fjx0 = _fjsp_madd_v2r8(dx30,fscal,fjx0);
640 fjy0 = _fjsp_madd_v2r8(dy30,fscal,fjy0);
641 fjz0 = _fjsp_madd_v2r8(dz30,fscal,fjz0);
643 gmx_fjsp_decrement_1rvec_1ptr_swizzle_v2r8(f+j_coord_offsetA,fjx0,fjy0,fjz0);
645 /* Inner loop uses 185 flops */
648 /* End of innermost loop */
650 gmx_fjsp_update_iforce_4atom_swizzle_v2r8(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
651 f+i_coord_offset,fshift+i_shift_offset);
654 /* Update potential energies */
655 gmx_fjsp_update_1pot_v2r8(velecsum,kernel_data->energygrp_elec+ggid);
656 gmx_fjsp_update_1pot_v2r8(vvdwsum,kernel_data->energygrp_vdw+ggid);
658 /* Increment number of inner iterations */
659 inneriter += j_index_end - j_index_start;
661 /* Outer loop uses 26 flops */
664 /* Increment number of outer iterations */
667 /* Update outer/inner flops */
669 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*185);
672 * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwLJEw_GeomW4P1_F_sparc64_hpc_ace_double
673 * Electrostatics interaction: Ewald
674 * VdW interaction: LJEwald
675 * Geometry: Water4-Particle
676 * Calculate force/pot: Force
679 nb_kernel_ElecEw_VdwLJEw_GeomW4P1_F_sparc64_hpc_ace_double
680 (t_nblist * gmx_restrict nlist,
681 rvec * gmx_restrict xx,
682 rvec * gmx_restrict ff,
683 t_forcerec * gmx_restrict fr,
684 t_mdatoms * gmx_restrict mdatoms,
685 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
686 t_nrnb * gmx_restrict nrnb)
688 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
689 * just 0 for non-waters.
690 * Suffixes A,B refer to j loop unrolling done with double precision SIMD, e.g. for the two different
691 * jnr indices corresponding to data put in the four positions in the SIMD register.
693 int i_shift_offset,i_coord_offset,outeriter,inneriter;
694 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
696 int j_coord_offsetA,j_coord_offsetB;
697 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
699 real *shiftvec,*fshift,*x,*f;
700 _fjsp_v2r8 tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
702 _fjsp_v2r8 ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
704 _fjsp_v2r8 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
706 _fjsp_v2r8 ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
708 _fjsp_v2r8 ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
709 int vdwjidx0A,vdwjidx0B;
710 _fjsp_v2r8 jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
711 _fjsp_v2r8 dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
712 _fjsp_v2r8 dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
713 _fjsp_v2r8 dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
714 _fjsp_v2r8 dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
715 _fjsp_v2r8 velec,felec,velecsum,facel,crf,krf,krf2;
718 _fjsp_v2r8 rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
721 _fjsp_v2r8 one_sixth = gmx_fjsp_set1_v2r8(1.0/6.0);
722 _fjsp_v2r8 one_twelfth = gmx_fjsp_set1_v2r8(1.0/12.0);
723 _fjsp_v2r8 c6grid_00;
724 _fjsp_v2r8 c6grid_10;
725 _fjsp_v2r8 c6grid_20;
726 _fjsp_v2r8 c6grid_30;
728 _fjsp_v2r8 ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
729 _fjsp_v2r8 one_half = gmx_fjsp_set1_v2r8(0.5);
730 _fjsp_v2r8 minus_one = gmx_fjsp_set1_v2r8(-1.0);
731 _fjsp_v2r8 ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
734 _fjsp_v2r8 dummy_mask,cutoff_mask;
735 _fjsp_v2r8 one = gmx_fjsp_set1_v2r8(1.0);
736 _fjsp_v2r8 two = gmx_fjsp_set1_v2r8(2.0);
737 union { _fjsp_v2r8 simd; long long int i[2]; } vfconv,gbconv,ewconv;
744 jindex = nlist->jindex;
746 shiftidx = nlist->shift;
748 shiftvec = fr->shift_vec[0];
749 fshift = fr->fshift[0];
750 facel = gmx_fjsp_set1_v2r8(fr->epsfac);
751 charge = mdatoms->chargeA;
752 nvdwtype = fr->ntype;
754 vdwtype = mdatoms->typeA;
755 vdwgridparam = fr->ljpme_c6grid;
756 sh_lj_ewald = gmx_fjsp_set1_v2r8(fr->ic->sh_lj_ewald);
757 ewclj = gmx_fjsp_set1_v2r8(fr->ewaldcoeff_lj);
758 ewclj2 = _fjsp_mul_v2r8(minus_one,_fjsp_mul_v2r8(ewclj,ewclj));
760 sh_ewald = gmx_fjsp_set1_v2r8(fr->ic->sh_ewald);
761 ewtab = fr->ic->tabq_coul_F;
762 ewtabscale = gmx_fjsp_set1_v2r8(fr->ic->tabq_scale);
763 ewtabhalfspace = gmx_fjsp_set1_v2r8(0.5/fr->ic->tabq_scale);
765 /* Setup water-specific parameters */
766 inr = nlist->iinr[0];
767 iq1 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+1]));
768 iq2 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+2]));
769 iq3 = _fjsp_mul_v2r8(facel,gmx_fjsp_set1_v2r8(charge[inr+3]));
770 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
772 /* Avoid stupid compiler warnings */
780 /* Start outer loop over neighborlists */
781 for(iidx=0; iidx<nri; iidx++)
783 /* Load shift vector for this list */
784 i_shift_offset = DIM*shiftidx[iidx];
786 /* Load limits for loop over neighbors */
787 j_index_start = jindex[iidx];
788 j_index_end = jindex[iidx+1];
790 /* Get outer coordinate index */
792 i_coord_offset = DIM*inr;
794 /* Load i particle coords and add shift vector */
795 gmx_fjsp_load_shift_and_4rvec_broadcast_v2r8(shiftvec+i_shift_offset,x+i_coord_offset,
796 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
798 fix0 = _fjsp_setzero_v2r8();
799 fiy0 = _fjsp_setzero_v2r8();
800 fiz0 = _fjsp_setzero_v2r8();
801 fix1 = _fjsp_setzero_v2r8();
802 fiy1 = _fjsp_setzero_v2r8();
803 fiz1 = _fjsp_setzero_v2r8();
804 fix2 = _fjsp_setzero_v2r8();
805 fiy2 = _fjsp_setzero_v2r8();
806 fiz2 = _fjsp_setzero_v2r8();
807 fix3 = _fjsp_setzero_v2r8();
808 fiy3 = _fjsp_setzero_v2r8();
809 fiz3 = _fjsp_setzero_v2r8();
811 /* Start inner kernel loop */
812 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
815 /* Get j neighbor index, and coordinate index */
818 j_coord_offsetA = DIM*jnrA;
819 j_coord_offsetB = DIM*jnrB;
821 /* load j atom coordinates */
822 gmx_fjsp_load_1rvec_2ptr_swizzle_v2r8(x+j_coord_offsetA,x+j_coord_offsetB,
825 /* Calculate displacement vector */
826 dx00 = _fjsp_sub_v2r8(ix0,jx0);
827 dy00 = _fjsp_sub_v2r8(iy0,jy0);
828 dz00 = _fjsp_sub_v2r8(iz0,jz0);
829 dx10 = _fjsp_sub_v2r8(ix1,jx0);
830 dy10 = _fjsp_sub_v2r8(iy1,jy0);
831 dz10 = _fjsp_sub_v2r8(iz1,jz0);
832 dx20 = _fjsp_sub_v2r8(ix2,jx0);
833 dy20 = _fjsp_sub_v2r8(iy2,jy0);
834 dz20 = _fjsp_sub_v2r8(iz2,jz0);
835 dx30 = _fjsp_sub_v2r8(ix3,jx0);
836 dy30 = _fjsp_sub_v2r8(iy3,jy0);
837 dz30 = _fjsp_sub_v2r8(iz3,jz0);
839 /* Calculate squared distance and things based on it */
840 rsq00 = gmx_fjsp_calc_rsq_v2r8(dx00,dy00,dz00);
841 rsq10 = gmx_fjsp_calc_rsq_v2r8(dx10,dy10,dz10);
842 rsq20 = gmx_fjsp_calc_rsq_v2r8(dx20,dy20,dz20);
843 rsq30 = gmx_fjsp_calc_rsq_v2r8(dx30,dy30,dz30);
845 rinv00 = gmx_fjsp_invsqrt_v2r8(rsq00);
846 rinv10 = gmx_fjsp_invsqrt_v2r8(rsq10);
847 rinv20 = gmx_fjsp_invsqrt_v2r8(rsq20);
848 rinv30 = gmx_fjsp_invsqrt_v2r8(rsq30);
850 rinvsq00 = _fjsp_mul_v2r8(rinv00,rinv00);
851 rinvsq10 = _fjsp_mul_v2r8(rinv10,rinv10);
852 rinvsq20 = _fjsp_mul_v2r8(rinv20,rinv20);
853 rinvsq30 = _fjsp_mul_v2r8(rinv30,rinv30);
855 /* Load parameters for j particles */
856 jq0 = gmx_fjsp_load_2real_swizzle_v2r8(charge+jnrA+0,charge+jnrB+0);
857 vdwjidx0A = 2*vdwtype[jnrA+0];
858 vdwjidx0B = 2*vdwtype[jnrB+0];
860 fjx0 = _fjsp_setzero_v2r8();
861 fjy0 = _fjsp_setzero_v2r8();
862 fjz0 = _fjsp_setzero_v2r8();
864 /**************************
865 * CALCULATE INTERACTIONS *
866 **************************/
868 r00 = _fjsp_mul_v2r8(rsq00,rinv00);
870 /* Compute parameters for interactions between i and j atoms */
871 gmx_fjsp_load_2pair_swizzle_v2r8(vdwparam+vdwioffset0+vdwjidx0A,
872 vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
874 c6grid_00 = gmx_fjsp_load_2real_swizzle_v2r8(vdwgridparam+vdwioffset0+vdwjidx0A,
875 vdwgridparam+vdwioffset0+vdwjidx0B);
877 /* Analytical LJ-PME */
878 rinvsix = _fjsp_mul_v2r8(_fjsp_mul_v2r8(rinvsq00,rinvsq00),rinvsq00);
879 ewcljrsq = _fjsp_mul_v2r8(ewclj2,rsq00);
880 ewclj6 = _fjsp_mul_v2r8(ewclj2,_fjsp_mul_v2r8(ewclj2,ewclj2));
881 exponent = gmx_simd_exp_d(-ewcljrsq);
882 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
883 poly = _fjsp_mul_v2r8(exponent,_fjsp_madd_v2r8(_fjsp_mul_v2r8(ewcljrsq,ewcljrsq),one_half,_fjsp_sub_v2r8(one,ewcljrsq)));
884 /* f6A = 6 * C6grid * (1 - poly) */
885 f6A = _fjsp_mul_v2r8(c6grid_00,_fjsp_msub_v2r8(one,poly));
886 /* f6B = C6grid * exponent * beta^6 */
887 f6B = _fjsp_mul_v2r8(_fjsp_mul_v2r8(c6grid_00,one_sixth),_fjsp_mul_v2r8(exponent,ewclj6));
888 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
889 fvdw = _fjsp_mul_v2r8(_fjsp_madd_v2r8(_fjsp_msub_v2r8(c12_00,rinvsix,_fjsp_sub_v2r8(c6_00,f6A)),rinvsix,f6B),rinvsq00);
893 /* Update vectorial force */
894 fix0 = _fjsp_madd_v2r8(dx00,fscal,fix0);
895 fiy0 = _fjsp_madd_v2r8(dy00,fscal,fiy0);
896 fiz0 = _fjsp_madd_v2r8(dz00,fscal,fiz0);
898 fjx0 = _fjsp_madd_v2r8(dx00,fscal,fjx0);
899 fjy0 = _fjsp_madd_v2r8(dy00,fscal,fjy0);
900 fjz0 = _fjsp_madd_v2r8(dz00,fscal,fjz0);
902 /**************************
903 * CALCULATE INTERACTIONS *
904 **************************/
906 r10 = _fjsp_mul_v2r8(rsq10,rinv10);
908 /* Compute parameters for interactions between i and j atoms */
909 qq10 = _fjsp_mul_v2r8(iq1,jq0);
911 /* EWALD ELECTROSTATICS */
913 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
914 ewrt = _fjsp_mul_v2r8(r10,ewtabscale);
915 itab_tmp = _fjsp_dtox_v2r8(ewrt);
916 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
917 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
919 gmx_fjsp_load_2pair_swizzle_v2r8(ewtab+ewconv.i[0],ewtab+ewconv.i[1],
921 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
922 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq10,rinv10),_fjsp_sub_v2r8(rinvsq10,felec));
926 /* Update vectorial force */
927 fix1 = _fjsp_madd_v2r8(dx10,fscal,fix1);
928 fiy1 = _fjsp_madd_v2r8(dy10,fscal,fiy1);
929 fiz1 = _fjsp_madd_v2r8(dz10,fscal,fiz1);
931 fjx0 = _fjsp_madd_v2r8(dx10,fscal,fjx0);
932 fjy0 = _fjsp_madd_v2r8(dy10,fscal,fjy0);
933 fjz0 = _fjsp_madd_v2r8(dz10,fscal,fjz0);
935 /**************************
936 * CALCULATE INTERACTIONS *
937 **************************/
939 r20 = _fjsp_mul_v2r8(rsq20,rinv20);
941 /* Compute parameters for interactions between i and j atoms */
942 qq20 = _fjsp_mul_v2r8(iq2,jq0);
944 /* EWALD ELECTROSTATICS */
946 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
947 ewrt = _fjsp_mul_v2r8(r20,ewtabscale);
948 itab_tmp = _fjsp_dtox_v2r8(ewrt);
949 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
950 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
952 gmx_fjsp_load_2pair_swizzle_v2r8(ewtab+ewconv.i[0],ewtab+ewconv.i[1],
954 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
955 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq20,rinv20),_fjsp_sub_v2r8(rinvsq20,felec));
959 /* Update vectorial force */
960 fix2 = _fjsp_madd_v2r8(dx20,fscal,fix2);
961 fiy2 = _fjsp_madd_v2r8(dy20,fscal,fiy2);
962 fiz2 = _fjsp_madd_v2r8(dz20,fscal,fiz2);
964 fjx0 = _fjsp_madd_v2r8(dx20,fscal,fjx0);
965 fjy0 = _fjsp_madd_v2r8(dy20,fscal,fjy0);
966 fjz0 = _fjsp_madd_v2r8(dz20,fscal,fjz0);
968 /**************************
969 * CALCULATE INTERACTIONS *
970 **************************/
972 r30 = _fjsp_mul_v2r8(rsq30,rinv30);
974 /* Compute parameters for interactions between i and j atoms */
975 qq30 = _fjsp_mul_v2r8(iq3,jq0);
977 /* EWALD ELECTROSTATICS */
979 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
980 ewrt = _fjsp_mul_v2r8(r30,ewtabscale);
981 itab_tmp = _fjsp_dtox_v2r8(ewrt);
982 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
983 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
985 gmx_fjsp_load_2pair_swizzle_v2r8(ewtab+ewconv.i[0],ewtab+ewconv.i[1],
987 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
988 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq30,rinv30),_fjsp_sub_v2r8(rinvsq30,felec));
992 /* Update vectorial force */
993 fix3 = _fjsp_madd_v2r8(dx30,fscal,fix3);
994 fiy3 = _fjsp_madd_v2r8(dy30,fscal,fiy3);
995 fiz3 = _fjsp_madd_v2r8(dz30,fscal,fiz3);
997 fjx0 = _fjsp_madd_v2r8(dx30,fscal,fjx0);
998 fjy0 = _fjsp_madd_v2r8(dy30,fscal,fjy0);
999 fjz0 = _fjsp_madd_v2r8(dz30,fscal,fjz0);
1001 gmx_fjsp_decrement_1rvec_2ptr_swizzle_v2r8(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
1003 /* Inner loop uses 168 flops */
1006 if(jidx<j_index_end)
1010 j_coord_offsetA = DIM*jnrA;
1012 /* load j atom coordinates */
1013 gmx_fjsp_load_1rvec_1ptr_swizzle_v2r8(x+j_coord_offsetA,
1016 /* Calculate displacement vector */
1017 dx00 = _fjsp_sub_v2r8(ix0,jx0);
1018 dy00 = _fjsp_sub_v2r8(iy0,jy0);
1019 dz00 = _fjsp_sub_v2r8(iz0,jz0);
1020 dx10 = _fjsp_sub_v2r8(ix1,jx0);
1021 dy10 = _fjsp_sub_v2r8(iy1,jy0);
1022 dz10 = _fjsp_sub_v2r8(iz1,jz0);
1023 dx20 = _fjsp_sub_v2r8(ix2,jx0);
1024 dy20 = _fjsp_sub_v2r8(iy2,jy0);
1025 dz20 = _fjsp_sub_v2r8(iz2,jz0);
1026 dx30 = _fjsp_sub_v2r8(ix3,jx0);
1027 dy30 = _fjsp_sub_v2r8(iy3,jy0);
1028 dz30 = _fjsp_sub_v2r8(iz3,jz0);
1030 /* Calculate squared distance and things based on it */
1031 rsq00 = gmx_fjsp_calc_rsq_v2r8(dx00,dy00,dz00);
1032 rsq10 = gmx_fjsp_calc_rsq_v2r8(dx10,dy10,dz10);
1033 rsq20 = gmx_fjsp_calc_rsq_v2r8(dx20,dy20,dz20);
1034 rsq30 = gmx_fjsp_calc_rsq_v2r8(dx30,dy30,dz30);
1036 rinv00 = gmx_fjsp_invsqrt_v2r8(rsq00);
1037 rinv10 = gmx_fjsp_invsqrt_v2r8(rsq10);
1038 rinv20 = gmx_fjsp_invsqrt_v2r8(rsq20);
1039 rinv30 = gmx_fjsp_invsqrt_v2r8(rsq30);
1041 rinvsq00 = _fjsp_mul_v2r8(rinv00,rinv00);
1042 rinvsq10 = _fjsp_mul_v2r8(rinv10,rinv10);
1043 rinvsq20 = _fjsp_mul_v2r8(rinv20,rinv20);
1044 rinvsq30 = _fjsp_mul_v2r8(rinv30,rinv30);
1046 /* Load parameters for j particles */
1047 jq0 = _fjsp_loadl_v2r8(_fjsp_setzero_v2r8(),charge+jnrA+0);
1048 vdwjidx0A = 2*vdwtype[jnrA+0];
1050 fjx0 = _fjsp_setzero_v2r8();
1051 fjy0 = _fjsp_setzero_v2r8();
1052 fjz0 = _fjsp_setzero_v2r8();
1054 /**************************
1055 * CALCULATE INTERACTIONS *
1056 **************************/
1058 r00 = _fjsp_mul_v2r8(rsq00,rinv00);
1060 /* Compute parameters for interactions between i and j atoms */
1061 gmx_fjsp_load_1pair_swizzle_v2r8(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
1063 c6grid_00 = gmx_fjsp_load_1real_swizzle_v2r8(vdwgridparam+vdwioffset0+vdwjidx0A);
1065 /* Analytical LJ-PME */
1066 rinvsix = _fjsp_mul_v2r8(_fjsp_mul_v2r8(rinvsq00,rinvsq00),rinvsq00);
1067 ewcljrsq = _fjsp_mul_v2r8(ewclj2,rsq00);
1068 ewclj6 = _fjsp_mul_v2r8(ewclj2,_fjsp_mul_v2r8(ewclj2,ewclj2));
1069 exponent = gmx_simd_exp_d(-ewcljrsq);
1070 /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
1071 poly = _fjsp_mul_v2r8(exponent,_fjsp_madd_v2r8(_fjsp_mul_v2r8(ewcljrsq,ewcljrsq),one_half,_fjsp_sub_v2r8(one,ewcljrsq)));
1072 /* f6A = 6 * C6grid * (1 - poly) */
1073 f6A = _fjsp_mul_v2r8(c6grid_00,_fjsp_msub_v2r8(one,poly));
1074 /* f6B = C6grid * exponent * beta^6 */
1075 f6B = _fjsp_mul_v2r8(_fjsp_mul_v2r8(c6grid_00,one_sixth),_fjsp_mul_v2r8(exponent,ewclj6));
1076 /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
1077 fvdw = _fjsp_mul_v2r8(_fjsp_madd_v2r8(_fjsp_msub_v2r8(c12_00,rinvsix,_fjsp_sub_v2r8(c6_00,f6A)),rinvsix,f6B),rinvsq00);
1081 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
1083 /* Update vectorial force */
1084 fix0 = _fjsp_madd_v2r8(dx00,fscal,fix0);
1085 fiy0 = _fjsp_madd_v2r8(dy00,fscal,fiy0);
1086 fiz0 = _fjsp_madd_v2r8(dz00,fscal,fiz0);
1088 fjx0 = _fjsp_madd_v2r8(dx00,fscal,fjx0);
1089 fjy0 = _fjsp_madd_v2r8(dy00,fscal,fjy0);
1090 fjz0 = _fjsp_madd_v2r8(dz00,fscal,fjz0);
1092 /**************************
1093 * CALCULATE INTERACTIONS *
1094 **************************/
1096 r10 = _fjsp_mul_v2r8(rsq10,rinv10);
1098 /* Compute parameters for interactions between i and j atoms */
1099 qq10 = _fjsp_mul_v2r8(iq1,jq0);
1101 /* EWALD ELECTROSTATICS */
1103 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1104 ewrt = _fjsp_mul_v2r8(r10,ewtabscale);
1105 itab_tmp = _fjsp_dtox_v2r8(ewrt);
1106 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
1107 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
1109 gmx_fjsp_load_1pair_swizzle_v2r8(ewtab+ewconv.i[0],&ewtabF,&ewtabFn);
1110 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
1111 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq10,rinv10),_fjsp_sub_v2r8(rinvsq10,felec));
1115 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
1117 /* Update vectorial force */
1118 fix1 = _fjsp_madd_v2r8(dx10,fscal,fix1);
1119 fiy1 = _fjsp_madd_v2r8(dy10,fscal,fiy1);
1120 fiz1 = _fjsp_madd_v2r8(dz10,fscal,fiz1);
1122 fjx0 = _fjsp_madd_v2r8(dx10,fscal,fjx0);
1123 fjy0 = _fjsp_madd_v2r8(dy10,fscal,fjy0);
1124 fjz0 = _fjsp_madd_v2r8(dz10,fscal,fjz0);
1126 /**************************
1127 * CALCULATE INTERACTIONS *
1128 **************************/
1130 r20 = _fjsp_mul_v2r8(rsq20,rinv20);
1132 /* Compute parameters for interactions between i and j atoms */
1133 qq20 = _fjsp_mul_v2r8(iq2,jq0);
1135 /* EWALD ELECTROSTATICS */
1137 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1138 ewrt = _fjsp_mul_v2r8(r20,ewtabscale);
1139 itab_tmp = _fjsp_dtox_v2r8(ewrt);
1140 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
1141 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
1143 gmx_fjsp_load_1pair_swizzle_v2r8(ewtab+ewconv.i[0],&ewtabF,&ewtabFn);
1144 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
1145 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq20,rinv20),_fjsp_sub_v2r8(rinvsq20,felec));
1149 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
1151 /* Update vectorial force */
1152 fix2 = _fjsp_madd_v2r8(dx20,fscal,fix2);
1153 fiy2 = _fjsp_madd_v2r8(dy20,fscal,fiy2);
1154 fiz2 = _fjsp_madd_v2r8(dz20,fscal,fiz2);
1156 fjx0 = _fjsp_madd_v2r8(dx20,fscal,fjx0);
1157 fjy0 = _fjsp_madd_v2r8(dy20,fscal,fjy0);
1158 fjz0 = _fjsp_madd_v2r8(dz20,fscal,fjz0);
1160 /**************************
1161 * CALCULATE INTERACTIONS *
1162 **************************/
1164 r30 = _fjsp_mul_v2r8(rsq30,rinv30);
1166 /* Compute parameters for interactions between i and j atoms */
1167 qq30 = _fjsp_mul_v2r8(iq3,jq0);
1169 /* EWALD ELECTROSTATICS */
1171 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1172 ewrt = _fjsp_mul_v2r8(r30,ewtabscale);
1173 itab_tmp = _fjsp_dtox_v2r8(ewrt);
1174 eweps = _fjsp_sub_v2r8(ewrt,_fjsp_xtod_v2r8(itab_tmp));
1175 _fjsp_store_v2r8(&ewconv.simd,itab_tmp);
1177 gmx_fjsp_load_1pair_swizzle_v2r8(ewtab+ewconv.i[0],&ewtabF,&ewtabFn);
1178 felec = _fjsp_madd_v2r8(eweps,ewtabFn,_fjsp_nmsub_v2r8(eweps,ewtabF,ewtabF));
1179 felec = _fjsp_mul_v2r8(_fjsp_mul_v2r8(qq30,rinv30),_fjsp_sub_v2r8(rinvsq30,felec));
1183 fscal = _fjsp_unpacklo_v2r8(fscal,_fjsp_setzero_v2r8());
1185 /* Update vectorial force */
1186 fix3 = _fjsp_madd_v2r8(dx30,fscal,fix3);
1187 fiy3 = _fjsp_madd_v2r8(dy30,fscal,fiy3);
1188 fiz3 = _fjsp_madd_v2r8(dz30,fscal,fiz3);
1190 fjx0 = _fjsp_madd_v2r8(dx30,fscal,fjx0);
1191 fjy0 = _fjsp_madd_v2r8(dy30,fscal,fjy0);
1192 fjz0 = _fjsp_madd_v2r8(dz30,fscal,fjz0);
1194 gmx_fjsp_decrement_1rvec_1ptr_swizzle_v2r8(f+j_coord_offsetA,fjx0,fjy0,fjz0);
1196 /* Inner loop uses 168 flops */
1199 /* End of innermost loop */
1201 gmx_fjsp_update_iforce_4atom_swizzle_v2r8(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1202 f+i_coord_offset,fshift+i_shift_offset);
1204 /* Increment number of inner iterations */
1205 inneriter += j_index_end - j_index_start;
1207 /* Outer loop uses 24 flops */
1210 /* Increment number of outer iterations */
1213 /* Update outer/inner flops */
1215 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*168);