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_ElecEwSh_VdwBhamSh_GeomW4P1_VF_c
49 * Electrostatics interaction: Ewald
50 * VdW interaction: Buckingham
51 * Geometry: Water4-Particle
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecEwSh_VdwBhamSh_GeomW4P1_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;
80 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
81 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
82 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
83 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
84 real velec,felec,velecsum,facel,crf,krf,krf2;
87 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
91 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
99 jindex = nlist->jindex;
101 shiftidx = nlist->shift;
103 shiftvec = fr->shift_vec[0];
104 fshift = fr->fshift[0];
106 charge = mdatoms->chargeA;
107 nvdwtype = fr->ntype;
109 vdwtype = mdatoms->typeA;
111 sh_ewald = fr->ic->sh_ewald;
112 ewtab = fr->ic->tabq_coul_FDV0;
113 ewtabscale = fr->ic->tabq_scale;
114 ewtabhalfspace = 0.5/ewtabscale;
116 /* Setup water-specific parameters */
117 inr = nlist->iinr[0];
118 iq1 = facel*charge[inr+1];
119 iq2 = facel*charge[inr+2];
120 iq3 = facel*charge[inr+3];
121 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
123 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
124 rcutoff = fr->rcoulomb;
125 rcutoff2 = rcutoff*rcutoff;
127 sh_vdw_invrcut6 = fr->ic->sh_invrc6;
133 /* Start outer loop over neighborlists */
134 for(iidx=0; iidx<nri; iidx++)
136 /* Load shift vector for this list */
137 i_shift_offset = DIM*shiftidx[iidx];
138 shX = shiftvec[i_shift_offset+XX];
139 shY = shiftvec[i_shift_offset+YY];
140 shZ = shiftvec[i_shift_offset+ZZ];
142 /* Load limits for loop over neighbors */
143 j_index_start = jindex[iidx];
144 j_index_end = jindex[iidx+1];
146 /* Get outer coordinate index */
148 i_coord_offset = DIM*inr;
150 /* Load i particle coords and add shift vector */
151 ix0 = shX + x[i_coord_offset+DIM*0+XX];
152 iy0 = shY + x[i_coord_offset+DIM*0+YY];
153 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
154 ix1 = shX + x[i_coord_offset+DIM*1+XX];
155 iy1 = shY + x[i_coord_offset+DIM*1+YY];
156 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
157 ix2 = shX + x[i_coord_offset+DIM*2+XX];
158 iy2 = shY + x[i_coord_offset+DIM*2+YY];
159 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
160 ix3 = shX + x[i_coord_offset+DIM*3+XX];
161 iy3 = shY + x[i_coord_offset+DIM*3+YY];
162 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
177 /* Reset potential sums */
181 /* Start inner kernel loop */
182 for(jidx=j_index_start; jidx<j_index_end; jidx++)
184 /* Get j neighbor index, and coordinate index */
186 j_coord_offset = DIM*jnr;
188 /* load j atom coordinates */
189 jx0 = x[j_coord_offset+DIM*0+XX];
190 jy0 = x[j_coord_offset+DIM*0+YY];
191 jz0 = x[j_coord_offset+DIM*0+ZZ];
193 /* Calculate displacement vector */
207 /* Calculate squared distance and things based on it */
208 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
209 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
210 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
211 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
213 rinv00 = gmx_invsqrt(rsq00);
214 rinv10 = gmx_invsqrt(rsq10);
215 rinv20 = gmx_invsqrt(rsq20);
216 rinv30 = gmx_invsqrt(rsq30);
218 rinvsq00 = rinv00*rinv00;
219 rinvsq10 = rinv10*rinv10;
220 rinvsq20 = rinv20*rinv20;
221 rinvsq30 = rinv30*rinv30;
223 /* Load parameters for j particles */
225 vdwjidx0 = 3*vdwtype[jnr+0];
227 /**************************
228 * CALCULATE INTERACTIONS *
229 **************************/
236 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
237 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
238 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
240 /* BUCKINGHAM DISPERSION/REPULSION */
241 rinvsix = rinvsq00*rinvsq00*rinvsq00;
242 vvdw6 = c6_00*rinvsix;
244 vvdwexp = cexp1_00*exp(-br);
245 vvdw = (vvdwexp-cexp1_00*exp(-cexp2_00*rvdw)) - (vvdw6 - c6_00*sh_vdw_invrcut6)*(1.0/6.0);
246 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
248 /* Update potential sums from outer loop */
253 /* Calculate temporary vectorial force */
258 /* Update vectorial force */
262 f[j_coord_offset+DIM*0+XX] -= tx;
263 f[j_coord_offset+DIM*0+YY] -= ty;
264 f[j_coord_offset+DIM*0+ZZ] -= tz;
268 /**************************
269 * CALCULATE INTERACTIONS *
270 **************************/
279 /* EWALD ELECTROSTATICS */
281 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
282 ewrt = r10*ewtabscale;
286 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
287 velec = qq10*((rinv10-sh_ewald)-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
288 felec = qq10*rinv10*(rinvsq10-felec);
290 /* Update potential sums from outer loop */
295 /* Calculate temporary vectorial force */
300 /* Update vectorial force */
304 f[j_coord_offset+DIM*0+XX] -= tx;
305 f[j_coord_offset+DIM*0+YY] -= ty;
306 f[j_coord_offset+DIM*0+ZZ] -= tz;
310 /**************************
311 * CALCULATE INTERACTIONS *
312 **************************/
321 /* EWALD ELECTROSTATICS */
323 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
324 ewrt = r20*ewtabscale;
328 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
329 velec = qq20*((rinv20-sh_ewald)-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
330 felec = qq20*rinv20*(rinvsq20-felec);
332 /* Update potential sums from outer loop */
337 /* Calculate temporary vectorial force */
342 /* Update vectorial force */
346 f[j_coord_offset+DIM*0+XX] -= tx;
347 f[j_coord_offset+DIM*0+YY] -= ty;
348 f[j_coord_offset+DIM*0+ZZ] -= tz;
352 /**************************
353 * CALCULATE INTERACTIONS *
354 **************************/
363 /* EWALD ELECTROSTATICS */
365 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
366 ewrt = r30*ewtabscale;
370 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
371 velec = qq30*((rinv30-sh_ewald)-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
372 felec = qq30*rinv30*(rinvsq30-felec);
374 /* Update potential sums from outer loop */
379 /* Calculate temporary vectorial force */
384 /* Update vectorial force */
388 f[j_coord_offset+DIM*0+XX] -= tx;
389 f[j_coord_offset+DIM*0+YY] -= ty;
390 f[j_coord_offset+DIM*0+ZZ] -= tz;
394 /* Inner loop uses 218 flops */
396 /* End of innermost loop */
399 f[i_coord_offset+DIM*0+XX] += fix0;
400 f[i_coord_offset+DIM*0+YY] += fiy0;
401 f[i_coord_offset+DIM*0+ZZ] += fiz0;
405 f[i_coord_offset+DIM*1+XX] += fix1;
406 f[i_coord_offset+DIM*1+YY] += fiy1;
407 f[i_coord_offset+DIM*1+ZZ] += fiz1;
411 f[i_coord_offset+DIM*2+XX] += fix2;
412 f[i_coord_offset+DIM*2+YY] += fiy2;
413 f[i_coord_offset+DIM*2+ZZ] += fiz2;
417 f[i_coord_offset+DIM*3+XX] += fix3;
418 f[i_coord_offset+DIM*3+YY] += fiy3;
419 f[i_coord_offset+DIM*3+ZZ] += fiz3;
423 fshift[i_shift_offset+XX] += tx;
424 fshift[i_shift_offset+YY] += ty;
425 fshift[i_shift_offset+ZZ] += tz;
428 /* Update potential energies */
429 kernel_data->energygrp_elec[ggid] += velecsum;
430 kernel_data->energygrp_vdw[ggid] += vvdwsum;
432 /* Increment number of inner iterations */
433 inneriter += j_index_end - j_index_start;
435 /* Outer loop uses 41 flops */
438 /* Increment number of outer iterations */
441 /* Update outer/inner flops */
443 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*41 + inneriter*218);
446 * Gromacs nonbonded kernel: nb_kernel_ElecEwSh_VdwBhamSh_GeomW4P1_F_c
447 * Electrostatics interaction: Ewald
448 * VdW interaction: Buckingham
449 * Geometry: Water4-Particle
450 * Calculate force/pot: Force
453 nb_kernel_ElecEwSh_VdwBhamSh_GeomW4P1_F_c
454 (t_nblist * gmx_restrict nlist,
455 rvec * gmx_restrict xx,
456 rvec * gmx_restrict ff,
457 t_forcerec * gmx_restrict fr,
458 t_mdatoms * gmx_restrict mdatoms,
459 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
460 t_nrnb * gmx_restrict nrnb)
462 int i_shift_offset,i_coord_offset,j_coord_offset;
463 int j_index_start,j_index_end;
464 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
465 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
466 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
467 real *shiftvec,*fshift,*x,*f;
469 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
471 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
473 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
475 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
477 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
478 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
479 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
480 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
481 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
482 real velec,felec,velecsum,facel,crf,krf,krf2;
485 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
489 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
497 jindex = nlist->jindex;
499 shiftidx = nlist->shift;
501 shiftvec = fr->shift_vec[0];
502 fshift = fr->fshift[0];
504 charge = mdatoms->chargeA;
505 nvdwtype = fr->ntype;
507 vdwtype = mdatoms->typeA;
509 sh_ewald = fr->ic->sh_ewald;
510 ewtab = fr->ic->tabq_coul_F;
511 ewtabscale = fr->ic->tabq_scale;
512 ewtabhalfspace = 0.5/ewtabscale;
514 /* Setup water-specific parameters */
515 inr = nlist->iinr[0];
516 iq1 = facel*charge[inr+1];
517 iq2 = facel*charge[inr+2];
518 iq3 = facel*charge[inr+3];
519 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
521 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
522 rcutoff = fr->rcoulomb;
523 rcutoff2 = rcutoff*rcutoff;
525 sh_vdw_invrcut6 = fr->ic->sh_invrc6;
531 /* Start outer loop over neighborlists */
532 for(iidx=0; iidx<nri; iidx++)
534 /* Load shift vector for this list */
535 i_shift_offset = DIM*shiftidx[iidx];
536 shX = shiftvec[i_shift_offset+XX];
537 shY = shiftvec[i_shift_offset+YY];
538 shZ = shiftvec[i_shift_offset+ZZ];
540 /* Load limits for loop over neighbors */
541 j_index_start = jindex[iidx];
542 j_index_end = jindex[iidx+1];
544 /* Get outer coordinate index */
546 i_coord_offset = DIM*inr;
548 /* Load i particle coords and add shift vector */
549 ix0 = shX + x[i_coord_offset+DIM*0+XX];
550 iy0 = shY + x[i_coord_offset+DIM*0+YY];
551 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
552 ix1 = shX + x[i_coord_offset+DIM*1+XX];
553 iy1 = shY + x[i_coord_offset+DIM*1+YY];
554 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
555 ix2 = shX + x[i_coord_offset+DIM*2+XX];
556 iy2 = shY + x[i_coord_offset+DIM*2+YY];
557 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
558 ix3 = shX + x[i_coord_offset+DIM*3+XX];
559 iy3 = shY + x[i_coord_offset+DIM*3+YY];
560 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
575 /* Start inner kernel loop */
576 for(jidx=j_index_start; jidx<j_index_end; jidx++)
578 /* Get j neighbor index, and coordinate index */
580 j_coord_offset = DIM*jnr;
582 /* load j atom coordinates */
583 jx0 = x[j_coord_offset+DIM*0+XX];
584 jy0 = x[j_coord_offset+DIM*0+YY];
585 jz0 = x[j_coord_offset+DIM*0+ZZ];
587 /* Calculate displacement vector */
601 /* Calculate squared distance and things based on it */
602 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
603 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
604 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
605 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
607 rinv00 = gmx_invsqrt(rsq00);
608 rinv10 = gmx_invsqrt(rsq10);
609 rinv20 = gmx_invsqrt(rsq20);
610 rinv30 = gmx_invsqrt(rsq30);
612 rinvsq00 = rinv00*rinv00;
613 rinvsq10 = rinv10*rinv10;
614 rinvsq20 = rinv20*rinv20;
615 rinvsq30 = rinv30*rinv30;
617 /* Load parameters for j particles */
619 vdwjidx0 = 3*vdwtype[jnr+0];
621 /**************************
622 * CALCULATE INTERACTIONS *
623 **************************/
630 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
631 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
632 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
634 /* BUCKINGHAM DISPERSION/REPULSION */
635 rinvsix = rinvsq00*rinvsq00*rinvsq00;
636 vvdw6 = c6_00*rinvsix;
638 vvdwexp = cexp1_00*exp(-br);
639 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
643 /* Calculate temporary vectorial force */
648 /* Update vectorial force */
652 f[j_coord_offset+DIM*0+XX] -= tx;
653 f[j_coord_offset+DIM*0+YY] -= ty;
654 f[j_coord_offset+DIM*0+ZZ] -= tz;
658 /**************************
659 * CALCULATE INTERACTIONS *
660 **************************/
669 /* EWALD ELECTROSTATICS */
671 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
672 ewrt = r10*ewtabscale;
675 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
676 felec = qq10*rinv10*(rinvsq10-felec);
680 /* Calculate temporary vectorial force */
685 /* Update vectorial force */
689 f[j_coord_offset+DIM*0+XX] -= tx;
690 f[j_coord_offset+DIM*0+YY] -= ty;
691 f[j_coord_offset+DIM*0+ZZ] -= tz;
695 /**************************
696 * CALCULATE INTERACTIONS *
697 **************************/
706 /* EWALD ELECTROSTATICS */
708 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
709 ewrt = r20*ewtabscale;
712 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
713 felec = qq20*rinv20*(rinvsq20-felec);
717 /* Calculate temporary vectorial force */
722 /* Update vectorial force */
726 f[j_coord_offset+DIM*0+XX] -= tx;
727 f[j_coord_offset+DIM*0+YY] -= ty;
728 f[j_coord_offset+DIM*0+ZZ] -= tz;
732 /**************************
733 * CALCULATE INTERACTIONS *
734 **************************/
743 /* EWALD ELECTROSTATICS */
745 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
746 ewrt = r30*ewtabscale;
749 felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
750 felec = qq30*rinv30*(rinvsq30-felec);
754 /* Calculate temporary vectorial force */
759 /* Update vectorial force */
763 f[j_coord_offset+DIM*0+XX] -= tx;
764 f[j_coord_offset+DIM*0+YY] -= ty;
765 f[j_coord_offset+DIM*0+ZZ] -= tz;
769 /* Inner loop uses 160 flops */
771 /* End of innermost loop */
774 f[i_coord_offset+DIM*0+XX] += fix0;
775 f[i_coord_offset+DIM*0+YY] += fiy0;
776 f[i_coord_offset+DIM*0+ZZ] += fiz0;
780 f[i_coord_offset+DIM*1+XX] += fix1;
781 f[i_coord_offset+DIM*1+YY] += fiy1;
782 f[i_coord_offset+DIM*1+ZZ] += fiz1;
786 f[i_coord_offset+DIM*2+XX] += fix2;
787 f[i_coord_offset+DIM*2+YY] += fiy2;
788 f[i_coord_offset+DIM*2+ZZ] += fiz2;
792 f[i_coord_offset+DIM*3+XX] += fix3;
793 f[i_coord_offset+DIM*3+YY] += fiy3;
794 f[i_coord_offset+DIM*3+ZZ] += fiz3;
798 fshift[i_shift_offset+XX] += tx;
799 fshift[i_shift_offset+YY] += ty;
800 fshift[i_shift_offset+ZZ] += tz;
802 /* Increment number of inner iterations */
803 inneriter += j_index_end - j_index_start;
805 /* Outer loop uses 39 flops */
808 /* Increment number of outer iterations */
811 /* Update outer/inner flops */
813 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*39 + inneriter*160);