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 c kernel generator.
42 #include "../nb_kernel.h"
43 #include "types/simple.h"
44 #include "gromacs/math/vec.h"
48 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwLJSw_GeomW4W4_VF_c
49 * Electrostatics interaction: ReactionField
50 * VdW interaction: LennardJones
51 * Geometry: Water4-Water4
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecRFCut_VdwLJSw_GeomW4W4_VF_c
56 (t_nblist * gmx_restrict nlist,
57 rvec * gmx_restrict xx,
58 rvec * gmx_restrict ff,
59 t_forcerec * gmx_restrict fr,
60 t_mdatoms * gmx_restrict mdatoms,
61 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
62 t_nrnb * gmx_restrict nrnb)
64 int i_shift_offset,i_coord_offset,j_coord_offset;
65 int j_index_start,j_index_end;
66 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
67 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
68 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
69 real *shiftvec,*fshift,*x,*f;
71 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
73 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
75 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
77 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
79 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
81 real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
83 real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
85 real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
86 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
87 real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11;
88 real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12;
89 real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13;
90 real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21;
91 real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22;
92 real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23;
93 real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31;
94 real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32;
95 real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33;
96 real velec,felec,velecsum,facel,crf,krf,krf2;
99 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
102 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
109 jindex = nlist->jindex;
111 shiftidx = nlist->shift;
113 shiftvec = fr->shift_vec[0];
114 fshift = fr->fshift[0];
116 charge = mdatoms->chargeA;
120 nvdwtype = fr->ntype;
122 vdwtype = mdatoms->typeA;
124 /* Setup water-specific parameters */
125 inr = nlist->iinr[0];
126 iq1 = facel*charge[inr+1];
127 iq2 = facel*charge[inr+2];
128 iq3 = facel*charge[inr+3];
129 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
134 vdwjidx0 = 2*vdwtype[inr+0];
135 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
136 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
147 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
148 rcutoff = fr->rcoulomb;
149 rcutoff2 = rcutoff*rcutoff;
151 rswitch = fr->rvdw_switch;
152 /* Setup switch parameters */
154 swV3 = -10.0/(d*d*d);
155 swV4 = 15.0/(d*d*d*d);
156 swV5 = -6.0/(d*d*d*d*d);
157 swF2 = -30.0/(d*d*d);
158 swF3 = 60.0/(d*d*d*d);
159 swF4 = -30.0/(d*d*d*d*d);
164 /* Start outer loop over neighborlists */
165 for(iidx=0; iidx<nri; iidx++)
167 /* Load shift vector for this list */
168 i_shift_offset = DIM*shiftidx[iidx];
169 shX = shiftvec[i_shift_offset+XX];
170 shY = shiftvec[i_shift_offset+YY];
171 shZ = shiftvec[i_shift_offset+ZZ];
173 /* Load limits for loop over neighbors */
174 j_index_start = jindex[iidx];
175 j_index_end = jindex[iidx+1];
177 /* Get outer coordinate index */
179 i_coord_offset = DIM*inr;
181 /* Load i particle coords and add shift vector */
182 ix0 = shX + x[i_coord_offset+DIM*0+XX];
183 iy0 = shY + x[i_coord_offset+DIM*0+YY];
184 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
185 ix1 = shX + x[i_coord_offset+DIM*1+XX];
186 iy1 = shY + x[i_coord_offset+DIM*1+YY];
187 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
188 ix2 = shX + x[i_coord_offset+DIM*2+XX];
189 iy2 = shY + x[i_coord_offset+DIM*2+YY];
190 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
191 ix3 = shX + x[i_coord_offset+DIM*3+XX];
192 iy3 = shY + x[i_coord_offset+DIM*3+YY];
193 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
208 /* Reset potential sums */
212 /* Start inner kernel loop */
213 for(jidx=j_index_start; jidx<j_index_end; jidx++)
215 /* Get j neighbor index, and coordinate index */
217 j_coord_offset = DIM*jnr;
219 /* load j atom coordinates */
220 jx0 = x[j_coord_offset+DIM*0+XX];
221 jy0 = x[j_coord_offset+DIM*0+YY];
222 jz0 = x[j_coord_offset+DIM*0+ZZ];
223 jx1 = x[j_coord_offset+DIM*1+XX];
224 jy1 = x[j_coord_offset+DIM*1+YY];
225 jz1 = x[j_coord_offset+DIM*1+ZZ];
226 jx2 = x[j_coord_offset+DIM*2+XX];
227 jy2 = x[j_coord_offset+DIM*2+YY];
228 jz2 = x[j_coord_offset+DIM*2+ZZ];
229 jx3 = x[j_coord_offset+DIM*3+XX];
230 jy3 = x[j_coord_offset+DIM*3+YY];
231 jz3 = x[j_coord_offset+DIM*3+ZZ];
233 /* Calculate displacement vector */
265 /* Calculate squared distance and things based on it */
266 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
267 rsq11 = dx11*dx11+dy11*dy11+dz11*dz11;
268 rsq12 = dx12*dx12+dy12*dy12+dz12*dz12;
269 rsq13 = dx13*dx13+dy13*dy13+dz13*dz13;
270 rsq21 = dx21*dx21+dy21*dy21+dz21*dz21;
271 rsq22 = dx22*dx22+dy22*dy22+dz22*dz22;
272 rsq23 = dx23*dx23+dy23*dy23+dz23*dz23;
273 rsq31 = dx31*dx31+dy31*dy31+dz31*dz31;
274 rsq32 = dx32*dx32+dy32*dy32+dz32*dz32;
275 rsq33 = dx33*dx33+dy33*dy33+dz33*dz33;
277 rinv00 = gmx_invsqrt(rsq00);
278 rinv11 = gmx_invsqrt(rsq11);
279 rinv12 = gmx_invsqrt(rsq12);
280 rinv13 = gmx_invsqrt(rsq13);
281 rinv21 = gmx_invsqrt(rsq21);
282 rinv22 = gmx_invsqrt(rsq22);
283 rinv23 = gmx_invsqrt(rsq23);
284 rinv31 = gmx_invsqrt(rsq31);
285 rinv32 = gmx_invsqrt(rsq32);
286 rinv33 = gmx_invsqrt(rsq33);
288 rinvsq00 = rinv00*rinv00;
289 rinvsq11 = rinv11*rinv11;
290 rinvsq12 = rinv12*rinv12;
291 rinvsq13 = rinv13*rinv13;
292 rinvsq21 = rinv21*rinv21;
293 rinvsq22 = rinv22*rinv22;
294 rinvsq23 = rinv23*rinv23;
295 rinvsq31 = rinv31*rinv31;
296 rinvsq32 = rinv32*rinv32;
297 rinvsq33 = rinv33*rinv33;
299 /**************************
300 * CALCULATE INTERACTIONS *
301 **************************/
308 /* LENNARD-JONES DISPERSION/REPULSION */
310 rinvsix = rinvsq00*rinvsq00*rinvsq00;
311 vvdw6 = c6_00*rinvsix;
312 vvdw12 = c12_00*rinvsix*rinvsix;
313 vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
314 fvdw = (vvdw12-vvdw6)*rinvsq00;
317 d = (d>0.0) ? d : 0.0;
319 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
321 dsw = d2*(swF2+d*(swF3+d*swF4));
323 /* Evaluate switch function */
324 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
325 fvdw = fvdw*sw - rinv00*vvdw*dsw;
328 /* Update potential sums from outer loop */
333 /* Calculate temporary vectorial force */
338 /* Update vectorial force */
342 f[j_coord_offset+DIM*0+XX] -= tx;
343 f[j_coord_offset+DIM*0+YY] -= ty;
344 f[j_coord_offset+DIM*0+ZZ] -= tz;
348 /**************************
349 * CALCULATE INTERACTIONS *
350 **************************/
355 /* REACTION-FIELD ELECTROSTATICS */
356 velec = qq11*(rinv11+krf*rsq11-crf);
357 felec = qq11*(rinv11*rinvsq11-krf2);
359 /* Update potential sums from outer loop */
364 /* Calculate temporary vectorial force */
369 /* Update vectorial force */
373 f[j_coord_offset+DIM*1+XX] -= tx;
374 f[j_coord_offset+DIM*1+YY] -= ty;
375 f[j_coord_offset+DIM*1+ZZ] -= tz;
379 /**************************
380 * CALCULATE INTERACTIONS *
381 **************************/
386 /* REACTION-FIELD ELECTROSTATICS */
387 velec = qq12*(rinv12+krf*rsq12-crf);
388 felec = qq12*(rinv12*rinvsq12-krf2);
390 /* Update potential sums from outer loop */
395 /* Calculate temporary vectorial force */
400 /* Update vectorial force */
404 f[j_coord_offset+DIM*2+XX] -= tx;
405 f[j_coord_offset+DIM*2+YY] -= ty;
406 f[j_coord_offset+DIM*2+ZZ] -= tz;
410 /**************************
411 * CALCULATE INTERACTIONS *
412 **************************/
417 /* REACTION-FIELD ELECTROSTATICS */
418 velec = qq13*(rinv13+krf*rsq13-crf);
419 felec = qq13*(rinv13*rinvsq13-krf2);
421 /* Update potential sums from outer loop */
426 /* Calculate temporary vectorial force */
431 /* Update vectorial force */
435 f[j_coord_offset+DIM*3+XX] -= tx;
436 f[j_coord_offset+DIM*3+YY] -= ty;
437 f[j_coord_offset+DIM*3+ZZ] -= tz;
441 /**************************
442 * CALCULATE INTERACTIONS *
443 **************************/
448 /* REACTION-FIELD ELECTROSTATICS */
449 velec = qq21*(rinv21+krf*rsq21-crf);
450 felec = qq21*(rinv21*rinvsq21-krf2);
452 /* Update potential sums from outer loop */
457 /* Calculate temporary vectorial force */
462 /* Update vectorial force */
466 f[j_coord_offset+DIM*1+XX] -= tx;
467 f[j_coord_offset+DIM*1+YY] -= ty;
468 f[j_coord_offset+DIM*1+ZZ] -= tz;
472 /**************************
473 * CALCULATE INTERACTIONS *
474 **************************/
479 /* REACTION-FIELD ELECTROSTATICS */
480 velec = qq22*(rinv22+krf*rsq22-crf);
481 felec = qq22*(rinv22*rinvsq22-krf2);
483 /* Update potential sums from outer loop */
488 /* Calculate temporary vectorial force */
493 /* Update vectorial force */
497 f[j_coord_offset+DIM*2+XX] -= tx;
498 f[j_coord_offset+DIM*2+YY] -= ty;
499 f[j_coord_offset+DIM*2+ZZ] -= tz;
503 /**************************
504 * CALCULATE INTERACTIONS *
505 **************************/
510 /* REACTION-FIELD ELECTROSTATICS */
511 velec = qq23*(rinv23+krf*rsq23-crf);
512 felec = qq23*(rinv23*rinvsq23-krf2);
514 /* Update potential sums from outer loop */
519 /* Calculate temporary vectorial force */
524 /* Update vectorial force */
528 f[j_coord_offset+DIM*3+XX] -= tx;
529 f[j_coord_offset+DIM*3+YY] -= ty;
530 f[j_coord_offset+DIM*3+ZZ] -= tz;
534 /**************************
535 * CALCULATE INTERACTIONS *
536 **************************/
541 /* REACTION-FIELD ELECTROSTATICS */
542 velec = qq31*(rinv31+krf*rsq31-crf);
543 felec = qq31*(rinv31*rinvsq31-krf2);
545 /* Update potential sums from outer loop */
550 /* Calculate temporary vectorial force */
555 /* Update vectorial force */
559 f[j_coord_offset+DIM*1+XX] -= tx;
560 f[j_coord_offset+DIM*1+YY] -= ty;
561 f[j_coord_offset+DIM*1+ZZ] -= tz;
565 /**************************
566 * CALCULATE INTERACTIONS *
567 **************************/
572 /* REACTION-FIELD ELECTROSTATICS */
573 velec = qq32*(rinv32+krf*rsq32-crf);
574 felec = qq32*(rinv32*rinvsq32-krf2);
576 /* Update potential sums from outer loop */
581 /* Calculate temporary vectorial force */
586 /* Update vectorial force */
590 f[j_coord_offset+DIM*2+XX] -= tx;
591 f[j_coord_offset+DIM*2+YY] -= ty;
592 f[j_coord_offset+DIM*2+ZZ] -= tz;
596 /**************************
597 * CALCULATE INTERACTIONS *
598 **************************/
603 /* REACTION-FIELD ELECTROSTATICS */
604 velec = qq33*(rinv33+krf*rsq33-crf);
605 felec = qq33*(rinv33*rinvsq33-krf2);
607 /* Update potential sums from outer loop */
612 /* Calculate temporary vectorial force */
617 /* Update vectorial force */
621 f[j_coord_offset+DIM*3+XX] -= tx;
622 f[j_coord_offset+DIM*3+YY] -= ty;
623 f[j_coord_offset+DIM*3+ZZ] -= tz;
627 /* Inner loop uses 332 flops */
629 /* End of innermost loop */
632 f[i_coord_offset+DIM*0+XX] += fix0;
633 f[i_coord_offset+DIM*0+YY] += fiy0;
634 f[i_coord_offset+DIM*0+ZZ] += fiz0;
638 f[i_coord_offset+DIM*1+XX] += fix1;
639 f[i_coord_offset+DIM*1+YY] += fiy1;
640 f[i_coord_offset+DIM*1+ZZ] += fiz1;
644 f[i_coord_offset+DIM*2+XX] += fix2;
645 f[i_coord_offset+DIM*2+YY] += fiy2;
646 f[i_coord_offset+DIM*2+ZZ] += fiz2;
650 f[i_coord_offset+DIM*3+XX] += fix3;
651 f[i_coord_offset+DIM*3+YY] += fiy3;
652 f[i_coord_offset+DIM*3+ZZ] += fiz3;
656 fshift[i_shift_offset+XX] += tx;
657 fshift[i_shift_offset+YY] += ty;
658 fshift[i_shift_offset+ZZ] += tz;
661 /* Update potential energies */
662 kernel_data->energygrp_elec[ggid] += velecsum;
663 kernel_data->energygrp_vdw[ggid] += vvdwsum;
665 /* Increment number of inner iterations */
666 inneriter += j_index_end - j_index_start;
668 /* Outer loop uses 41 flops */
671 /* Increment number of outer iterations */
674 /* Update outer/inner flops */
676 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*41 + inneriter*332);
679 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwLJSw_GeomW4W4_F_c
680 * Electrostatics interaction: ReactionField
681 * VdW interaction: LennardJones
682 * Geometry: Water4-Water4
683 * Calculate force/pot: Force
686 nb_kernel_ElecRFCut_VdwLJSw_GeomW4W4_F_c
687 (t_nblist * gmx_restrict nlist,
688 rvec * gmx_restrict xx,
689 rvec * gmx_restrict ff,
690 t_forcerec * gmx_restrict fr,
691 t_mdatoms * gmx_restrict mdatoms,
692 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
693 t_nrnb * gmx_restrict nrnb)
695 int i_shift_offset,i_coord_offset,j_coord_offset;
696 int j_index_start,j_index_end;
697 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
698 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
699 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
700 real *shiftvec,*fshift,*x,*f;
702 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
704 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
706 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
708 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
710 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
712 real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
714 real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
716 real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
717 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
718 real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11;
719 real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12;
720 real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13;
721 real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21;
722 real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22;
723 real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23;
724 real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31;
725 real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32;
726 real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33;
727 real velec,felec,velecsum,facel,crf,krf,krf2;
730 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
733 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
740 jindex = nlist->jindex;
742 shiftidx = nlist->shift;
744 shiftvec = fr->shift_vec[0];
745 fshift = fr->fshift[0];
747 charge = mdatoms->chargeA;
751 nvdwtype = fr->ntype;
753 vdwtype = mdatoms->typeA;
755 /* Setup water-specific parameters */
756 inr = nlist->iinr[0];
757 iq1 = facel*charge[inr+1];
758 iq2 = facel*charge[inr+2];
759 iq3 = facel*charge[inr+3];
760 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
765 vdwjidx0 = 2*vdwtype[inr+0];
766 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
767 c12_00 = vdwparam[vdwioffset0+vdwjidx0+1];
778 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
779 rcutoff = fr->rcoulomb;
780 rcutoff2 = rcutoff*rcutoff;
782 rswitch = fr->rvdw_switch;
783 /* Setup switch parameters */
785 swV3 = -10.0/(d*d*d);
786 swV4 = 15.0/(d*d*d*d);
787 swV5 = -6.0/(d*d*d*d*d);
788 swF2 = -30.0/(d*d*d);
789 swF3 = 60.0/(d*d*d*d);
790 swF4 = -30.0/(d*d*d*d*d);
795 /* Start outer loop over neighborlists */
796 for(iidx=0; iidx<nri; iidx++)
798 /* Load shift vector for this list */
799 i_shift_offset = DIM*shiftidx[iidx];
800 shX = shiftvec[i_shift_offset+XX];
801 shY = shiftvec[i_shift_offset+YY];
802 shZ = shiftvec[i_shift_offset+ZZ];
804 /* Load limits for loop over neighbors */
805 j_index_start = jindex[iidx];
806 j_index_end = jindex[iidx+1];
808 /* Get outer coordinate index */
810 i_coord_offset = DIM*inr;
812 /* Load i particle coords and add shift vector */
813 ix0 = shX + x[i_coord_offset+DIM*0+XX];
814 iy0 = shY + x[i_coord_offset+DIM*0+YY];
815 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
816 ix1 = shX + x[i_coord_offset+DIM*1+XX];
817 iy1 = shY + x[i_coord_offset+DIM*1+YY];
818 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
819 ix2 = shX + x[i_coord_offset+DIM*2+XX];
820 iy2 = shY + x[i_coord_offset+DIM*2+YY];
821 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
822 ix3 = shX + x[i_coord_offset+DIM*3+XX];
823 iy3 = shY + x[i_coord_offset+DIM*3+YY];
824 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
839 /* Start inner kernel loop */
840 for(jidx=j_index_start; jidx<j_index_end; jidx++)
842 /* Get j neighbor index, and coordinate index */
844 j_coord_offset = DIM*jnr;
846 /* load j atom coordinates */
847 jx0 = x[j_coord_offset+DIM*0+XX];
848 jy0 = x[j_coord_offset+DIM*0+YY];
849 jz0 = x[j_coord_offset+DIM*0+ZZ];
850 jx1 = x[j_coord_offset+DIM*1+XX];
851 jy1 = x[j_coord_offset+DIM*1+YY];
852 jz1 = x[j_coord_offset+DIM*1+ZZ];
853 jx2 = x[j_coord_offset+DIM*2+XX];
854 jy2 = x[j_coord_offset+DIM*2+YY];
855 jz2 = x[j_coord_offset+DIM*2+ZZ];
856 jx3 = x[j_coord_offset+DIM*3+XX];
857 jy3 = x[j_coord_offset+DIM*3+YY];
858 jz3 = x[j_coord_offset+DIM*3+ZZ];
860 /* Calculate displacement vector */
892 /* Calculate squared distance and things based on it */
893 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
894 rsq11 = dx11*dx11+dy11*dy11+dz11*dz11;
895 rsq12 = dx12*dx12+dy12*dy12+dz12*dz12;
896 rsq13 = dx13*dx13+dy13*dy13+dz13*dz13;
897 rsq21 = dx21*dx21+dy21*dy21+dz21*dz21;
898 rsq22 = dx22*dx22+dy22*dy22+dz22*dz22;
899 rsq23 = dx23*dx23+dy23*dy23+dz23*dz23;
900 rsq31 = dx31*dx31+dy31*dy31+dz31*dz31;
901 rsq32 = dx32*dx32+dy32*dy32+dz32*dz32;
902 rsq33 = dx33*dx33+dy33*dy33+dz33*dz33;
904 rinv00 = gmx_invsqrt(rsq00);
905 rinv11 = gmx_invsqrt(rsq11);
906 rinv12 = gmx_invsqrt(rsq12);
907 rinv13 = gmx_invsqrt(rsq13);
908 rinv21 = gmx_invsqrt(rsq21);
909 rinv22 = gmx_invsqrt(rsq22);
910 rinv23 = gmx_invsqrt(rsq23);
911 rinv31 = gmx_invsqrt(rsq31);
912 rinv32 = gmx_invsqrt(rsq32);
913 rinv33 = gmx_invsqrt(rsq33);
915 rinvsq00 = rinv00*rinv00;
916 rinvsq11 = rinv11*rinv11;
917 rinvsq12 = rinv12*rinv12;
918 rinvsq13 = rinv13*rinv13;
919 rinvsq21 = rinv21*rinv21;
920 rinvsq22 = rinv22*rinv22;
921 rinvsq23 = rinv23*rinv23;
922 rinvsq31 = rinv31*rinv31;
923 rinvsq32 = rinv32*rinv32;
924 rinvsq33 = rinv33*rinv33;
926 /**************************
927 * CALCULATE INTERACTIONS *
928 **************************/
935 /* LENNARD-JONES DISPERSION/REPULSION */
937 rinvsix = rinvsq00*rinvsq00*rinvsq00;
938 vvdw6 = c6_00*rinvsix;
939 vvdw12 = c12_00*rinvsix*rinvsix;
940 vvdw = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
941 fvdw = (vvdw12-vvdw6)*rinvsq00;
944 d = (d>0.0) ? d : 0.0;
946 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
948 dsw = d2*(swF2+d*(swF3+d*swF4));
950 /* Evaluate switch function */
951 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
952 fvdw = fvdw*sw - rinv00*vvdw*dsw;
956 /* Calculate temporary vectorial force */
961 /* Update vectorial force */
965 f[j_coord_offset+DIM*0+XX] -= tx;
966 f[j_coord_offset+DIM*0+YY] -= ty;
967 f[j_coord_offset+DIM*0+ZZ] -= tz;
971 /**************************
972 * CALCULATE INTERACTIONS *
973 **************************/
978 /* REACTION-FIELD ELECTROSTATICS */
979 felec = qq11*(rinv11*rinvsq11-krf2);
983 /* Calculate temporary vectorial force */
988 /* Update vectorial force */
992 f[j_coord_offset+DIM*1+XX] -= tx;
993 f[j_coord_offset+DIM*1+YY] -= ty;
994 f[j_coord_offset+DIM*1+ZZ] -= tz;
998 /**************************
999 * CALCULATE INTERACTIONS *
1000 **************************/
1005 /* REACTION-FIELD ELECTROSTATICS */
1006 felec = qq12*(rinv12*rinvsq12-krf2);
1010 /* Calculate temporary vectorial force */
1015 /* Update vectorial force */
1019 f[j_coord_offset+DIM*2+XX] -= tx;
1020 f[j_coord_offset+DIM*2+YY] -= ty;
1021 f[j_coord_offset+DIM*2+ZZ] -= tz;
1025 /**************************
1026 * CALCULATE INTERACTIONS *
1027 **************************/
1032 /* REACTION-FIELD ELECTROSTATICS */
1033 felec = qq13*(rinv13*rinvsq13-krf2);
1037 /* Calculate temporary vectorial force */
1042 /* Update vectorial force */
1046 f[j_coord_offset+DIM*3+XX] -= tx;
1047 f[j_coord_offset+DIM*3+YY] -= ty;
1048 f[j_coord_offset+DIM*3+ZZ] -= tz;
1052 /**************************
1053 * CALCULATE INTERACTIONS *
1054 **************************/
1059 /* REACTION-FIELD ELECTROSTATICS */
1060 felec = qq21*(rinv21*rinvsq21-krf2);
1064 /* Calculate temporary vectorial force */
1069 /* Update vectorial force */
1073 f[j_coord_offset+DIM*1+XX] -= tx;
1074 f[j_coord_offset+DIM*1+YY] -= ty;
1075 f[j_coord_offset+DIM*1+ZZ] -= tz;
1079 /**************************
1080 * CALCULATE INTERACTIONS *
1081 **************************/
1086 /* REACTION-FIELD ELECTROSTATICS */
1087 felec = qq22*(rinv22*rinvsq22-krf2);
1091 /* Calculate temporary vectorial force */
1096 /* Update vectorial force */
1100 f[j_coord_offset+DIM*2+XX] -= tx;
1101 f[j_coord_offset+DIM*2+YY] -= ty;
1102 f[j_coord_offset+DIM*2+ZZ] -= tz;
1106 /**************************
1107 * CALCULATE INTERACTIONS *
1108 **************************/
1113 /* REACTION-FIELD ELECTROSTATICS */
1114 felec = qq23*(rinv23*rinvsq23-krf2);
1118 /* Calculate temporary vectorial force */
1123 /* Update vectorial force */
1127 f[j_coord_offset+DIM*3+XX] -= tx;
1128 f[j_coord_offset+DIM*3+YY] -= ty;
1129 f[j_coord_offset+DIM*3+ZZ] -= tz;
1133 /**************************
1134 * CALCULATE INTERACTIONS *
1135 **************************/
1140 /* REACTION-FIELD ELECTROSTATICS */
1141 felec = qq31*(rinv31*rinvsq31-krf2);
1145 /* Calculate temporary vectorial force */
1150 /* Update vectorial force */
1154 f[j_coord_offset+DIM*1+XX] -= tx;
1155 f[j_coord_offset+DIM*1+YY] -= ty;
1156 f[j_coord_offset+DIM*1+ZZ] -= tz;
1160 /**************************
1161 * CALCULATE INTERACTIONS *
1162 **************************/
1167 /* REACTION-FIELD ELECTROSTATICS */
1168 felec = qq32*(rinv32*rinvsq32-krf2);
1172 /* Calculate temporary vectorial force */
1177 /* Update vectorial force */
1181 f[j_coord_offset+DIM*2+XX] -= tx;
1182 f[j_coord_offset+DIM*2+YY] -= ty;
1183 f[j_coord_offset+DIM*2+ZZ] -= tz;
1187 /**************************
1188 * CALCULATE INTERACTIONS *
1189 **************************/
1194 /* REACTION-FIELD ELECTROSTATICS */
1195 felec = qq33*(rinv33*rinvsq33-krf2);
1199 /* Calculate temporary vectorial force */
1204 /* Update vectorial force */
1208 f[j_coord_offset+DIM*3+XX] -= tx;
1209 f[j_coord_offset+DIM*3+YY] -= ty;
1210 f[j_coord_offset+DIM*3+ZZ] -= tz;
1214 /* Inner loop uses 285 flops */
1216 /* End of innermost loop */
1219 f[i_coord_offset+DIM*0+XX] += fix0;
1220 f[i_coord_offset+DIM*0+YY] += fiy0;
1221 f[i_coord_offset+DIM*0+ZZ] += fiz0;
1225 f[i_coord_offset+DIM*1+XX] += fix1;
1226 f[i_coord_offset+DIM*1+YY] += fiy1;
1227 f[i_coord_offset+DIM*1+ZZ] += fiz1;
1231 f[i_coord_offset+DIM*2+XX] += fix2;
1232 f[i_coord_offset+DIM*2+YY] += fiy2;
1233 f[i_coord_offset+DIM*2+ZZ] += fiz2;
1237 f[i_coord_offset+DIM*3+XX] += fix3;
1238 f[i_coord_offset+DIM*3+YY] += fiy3;
1239 f[i_coord_offset+DIM*3+ZZ] += fiz3;
1243 fshift[i_shift_offset+XX] += tx;
1244 fshift[i_shift_offset+YY] += ty;
1245 fshift[i_shift_offset+ZZ] += tz;
1247 /* Increment number of inner iterations */
1248 inneriter += j_index_end - j_index_start;
1250 /* Outer loop uses 39 flops */
1253 /* Increment number of outer iterations */
1256 /* Update outer/inner flops */
1258 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*39 + inneriter*285);