2 * Note: this file was generated by the Gromacs c kernel generator.
4 * This source code is part of
8 * Copyright (c) 2001-2012, The GROMACS Development Team
10 * Gromacs is a library for molecular simulation and trajectory analysis,
11 * written by Erik Lindahl, David van der Spoel, Berk Hess, and others - for
12 * a full list of developers and information, check out http://www.gromacs.org
14 * This program is free software; you can redistribute it and/or modify it under
15 * the terms of the GNU Lesser General Public License as published by the Free
16 * Software Foundation; either version 2 of the License, or (at your option) any
19 * To help fund GROMACS development, we humbly ask that you cite
20 * the papers people have written on it - you can find them on the website.
28 #include "../nb_kernel.h"
29 #include "types/simple.h"
34 * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwBham_GeomW4P1_VF_c
35 * Electrostatics interaction: CubicSplineTable
36 * VdW interaction: Buckingham
37 * Geometry: Water4-Particle
38 * Calculate force/pot: PotentialAndForce
41 nb_kernel_ElecCSTab_VdwBham_GeomW4P1_VF_c
42 (t_nblist * gmx_restrict nlist,
43 rvec * gmx_restrict xx,
44 rvec * gmx_restrict ff,
45 t_forcerec * gmx_restrict fr,
46 t_mdatoms * gmx_restrict mdatoms,
47 nb_kernel_data_t * gmx_restrict kernel_data,
48 t_nrnb * gmx_restrict nrnb)
50 int i_shift_offset,i_coord_offset,j_coord_offset;
51 int j_index_start,j_index_end;
52 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
53 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
54 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
55 real *shiftvec,*fshift,*x,*f;
57 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
59 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
61 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
63 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
65 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
66 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
67 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
68 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
69 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
70 real velec,felec,velecsum,facel,crf,krf,krf2;
73 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
77 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
85 jindex = nlist->jindex;
87 shiftidx = nlist->shift;
89 shiftvec = fr->shift_vec[0];
90 fshift = fr->fshift[0];
92 charge = mdatoms->chargeA;
95 vdwtype = mdatoms->typeA;
97 vftab = kernel_data->table_elec->data;
98 vftabscale = kernel_data->table_elec->scale;
100 /* Setup water-specific parameters */
101 inr = nlist->iinr[0];
102 iq1 = facel*charge[inr+1];
103 iq2 = facel*charge[inr+2];
104 iq3 = facel*charge[inr+3];
105 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
110 /* Start outer loop over neighborlists */
111 for(iidx=0; iidx<nri; iidx++)
113 /* Load shift vector for this list */
114 i_shift_offset = DIM*shiftidx[iidx];
115 shX = shiftvec[i_shift_offset+XX];
116 shY = shiftvec[i_shift_offset+YY];
117 shZ = shiftvec[i_shift_offset+ZZ];
119 /* Load limits for loop over neighbors */
120 j_index_start = jindex[iidx];
121 j_index_end = jindex[iidx+1];
123 /* Get outer coordinate index */
125 i_coord_offset = DIM*inr;
127 /* Load i particle coords and add shift vector */
128 ix0 = shX + x[i_coord_offset+DIM*0+XX];
129 iy0 = shY + x[i_coord_offset+DIM*0+YY];
130 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
131 ix1 = shX + x[i_coord_offset+DIM*1+XX];
132 iy1 = shY + x[i_coord_offset+DIM*1+YY];
133 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
134 ix2 = shX + x[i_coord_offset+DIM*2+XX];
135 iy2 = shY + x[i_coord_offset+DIM*2+YY];
136 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
137 ix3 = shX + x[i_coord_offset+DIM*3+XX];
138 iy3 = shY + x[i_coord_offset+DIM*3+YY];
139 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
154 /* Reset potential sums */
158 /* Start inner kernel loop */
159 for(jidx=j_index_start; jidx<j_index_end; jidx++)
161 /* Get j neighbor index, and coordinate index */
163 j_coord_offset = DIM*jnr;
165 /* load j atom coordinates */
166 jx0 = x[j_coord_offset+DIM*0+XX];
167 jy0 = x[j_coord_offset+DIM*0+YY];
168 jz0 = x[j_coord_offset+DIM*0+ZZ];
170 /* Calculate displacement vector */
184 /* Calculate squared distance and things based on it */
185 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
186 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
187 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
188 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
190 rinv00 = gmx_invsqrt(rsq00);
191 rinv10 = gmx_invsqrt(rsq10);
192 rinv20 = gmx_invsqrt(rsq20);
193 rinv30 = gmx_invsqrt(rsq30);
195 rinvsq00 = rinv00*rinv00;
197 /* Load parameters for j particles */
199 vdwjidx0 = 3*vdwtype[jnr+0];
201 /**************************
202 * CALCULATE INTERACTIONS *
203 **************************/
207 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
208 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
209 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
211 /* BUCKINGHAM DISPERSION/REPULSION */
212 rinvsix = rinvsq00*rinvsq00*rinvsq00;
213 vvdw6 = c6_00*rinvsix;
215 vvdwexp = cexp1_00*exp(-br);
216 vvdw = vvdwexp - vvdw6*(1.0/6.0);
217 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
219 /* Update potential sums from outer loop */
224 /* Calculate temporary vectorial force */
229 /* Update vectorial force */
233 f[j_coord_offset+DIM*0+XX] -= tx;
234 f[j_coord_offset+DIM*0+YY] -= ty;
235 f[j_coord_offset+DIM*0+ZZ] -= tz;
237 /**************************
238 * CALCULATE INTERACTIONS *
239 **************************/
245 /* Calculate table index by multiplying r with table scale and truncate to integer */
251 /* CUBIC SPLINE TABLE ELECTROSTATICS */
254 Geps = vfeps*vftab[vfitab+2];
255 Heps2 = vfeps*vfeps*vftab[vfitab+3];
259 FF = Fp+Geps+2.0*Heps2;
260 felec = -qq10*FF*vftabscale*rinv10;
262 /* Update potential sums from outer loop */
267 /* Calculate temporary vectorial force */
272 /* Update vectorial force */
276 f[j_coord_offset+DIM*0+XX] -= tx;
277 f[j_coord_offset+DIM*0+YY] -= ty;
278 f[j_coord_offset+DIM*0+ZZ] -= tz;
280 /**************************
281 * CALCULATE INTERACTIONS *
282 **************************/
288 /* Calculate table index by multiplying r with table scale and truncate to integer */
294 /* CUBIC SPLINE TABLE ELECTROSTATICS */
297 Geps = vfeps*vftab[vfitab+2];
298 Heps2 = vfeps*vfeps*vftab[vfitab+3];
302 FF = Fp+Geps+2.0*Heps2;
303 felec = -qq20*FF*vftabscale*rinv20;
305 /* Update potential sums from outer loop */
310 /* Calculate temporary vectorial force */
315 /* Update vectorial force */
319 f[j_coord_offset+DIM*0+XX] -= tx;
320 f[j_coord_offset+DIM*0+YY] -= ty;
321 f[j_coord_offset+DIM*0+ZZ] -= tz;
323 /**************************
324 * CALCULATE INTERACTIONS *
325 **************************/
331 /* Calculate table index by multiplying r with table scale and truncate to integer */
337 /* CUBIC SPLINE TABLE ELECTROSTATICS */
340 Geps = vfeps*vftab[vfitab+2];
341 Heps2 = vfeps*vfeps*vftab[vfitab+3];
345 FF = Fp+Geps+2.0*Heps2;
346 felec = -qq30*FF*vftabscale*rinv30;
348 /* Update potential sums from outer loop */
353 /* Calculate temporary vectorial force */
358 /* Update vectorial force */
362 f[j_coord_offset+DIM*0+XX] -= tx;
363 f[j_coord_offset+DIM*0+YY] -= ty;
364 f[j_coord_offset+DIM*0+ZZ] -= tz;
366 /* Inner loop uses 187 flops */
368 /* End of innermost loop */
371 f[i_coord_offset+DIM*0+XX] += fix0;
372 f[i_coord_offset+DIM*0+YY] += fiy0;
373 f[i_coord_offset+DIM*0+ZZ] += fiz0;
377 f[i_coord_offset+DIM*1+XX] += fix1;
378 f[i_coord_offset+DIM*1+YY] += fiy1;
379 f[i_coord_offset+DIM*1+ZZ] += fiz1;
383 f[i_coord_offset+DIM*2+XX] += fix2;
384 f[i_coord_offset+DIM*2+YY] += fiy2;
385 f[i_coord_offset+DIM*2+ZZ] += fiz2;
389 f[i_coord_offset+DIM*3+XX] += fix3;
390 f[i_coord_offset+DIM*3+YY] += fiy3;
391 f[i_coord_offset+DIM*3+ZZ] += fiz3;
395 fshift[i_shift_offset+XX] += tx;
396 fshift[i_shift_offset+YY] += ty;
397 fshift[i_shift_offset+ZZ] += tz;
400 /* Update potential energies */
401 kernel_data->energygrp_elec[ggid] += velecsum;
402 kernel_data->energygrp_vdw[ggid] += vvdwsum;
404 /* Increment number of inner iterations */
405 inneriter += j_index_end - j_index_start;
407 /* Outer loop uses 41 flops */
410 /* Increment number of outer iterations */
413 /* Update outer/inner flops */
415 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*41 + inneriter*187);
418 * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwBham_GeomW4P1_F_c
419 * Electrostatics interaction: CubicSplineTable
420 * VdW interaction: Buckingham
421 * Geometry: Water4-Particle
422 * Calculate force/pot: Force
425 nb_kernel_ElecCSTab_VdwBham_GeomW4P1_F_c
426 (t_nblist * gmx_restrict nlist,
427 rvec * gmx_restrict xx,
428 rvec * gmx_restrict ff,
429 t_forcerec * gmx_restrict fr,
430 t_mdatoms * gmx_restrict mdatoms,
431 nb_kernel_data_t * gmx_restrict kernel_data,
432 t_nrnb * gmx_restrict nrnb)
434 int i_shift_offset,i_coord_offset,j_coord_offset;
435 int j_index_start,j_index_end;
436 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
437 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
438 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
439 real *shiftvec,*fshift,*x,*f;
441 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
443 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
445 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
447 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
449 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
450 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
451 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
452 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
453 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
454 real velec,felec,velecsum,facel,crf,krf,krf2;
457 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
461 real rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
469 jindex = nlist->jindex;
471 shiftidx = nlist->shift;
473 shiftvec = fr->shift_vec[0];
474 fshift = fr->fshift[0];
476 charge = mdatoms->chargeA;
477 nvdwtype = fr->ntype;
479 vdwtype = mdatoms->typeA;
481 vftab = kernel_data->table_elec->data;
482 vftabscale = kernel_data->table_elec->scale;
484 /* Setup water-specific parameters */
485 inr = nlist->iinr[0];
486 iq1 = facel*charge[inr+1];
487 iq2 = facel*charge[inr+2];
488 iq3 = facel*charge[inr+3];
489 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
494 /* Start outer loop over neighborlists */
495 for(iidx=0; iidx<nri; iidx++)
497 /* Load shift vector for this list */
498 i_shift_offset = DIM*shiftidx[iidx];
499 shX = shiftvec[i_shift_offset+XX];
500 shY = shiftvec[i_shift_offset+YY];
501 shZ = shiftvec[i_shift_offset+ZZ];
503 /* Load limits for loop over neighbors */
504 j_index_start = jindex[iidx];
505 j_index_end = jindex[iidx+1];
507 /* Get outer coordinate index */
509 i_coord_offset = DIM*inr;
511 /* Load i particle coords and add shift vector */
512 ix0 = shX + x[i_coord_offset+DIM*0+XX];
513 iy0 = shY + x[i_coord_offset+DIM*0+YY];
514 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
515 ix1 = shX + x[i_coord_offset+DIM*1+XX];
516 iy1 = shY + x[i_coord_offset+DIM*1+YY];
517 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
518 ix2 = shX + x[i_coord_offset+DIM*2+XX];
519 iy2 = shY + x[i_coord_offset+DIM*2+YY];
520 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
521 ix3 = shX + x[i_coord_offset+DIM*3+XX];
522 iy3 = shY + x[i_coord_offset+DIM*3+YY];
523 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
538 /* Start inner kernel loop */
539 for(jidx=j_index_start; jidx<j_index_end; jidx++)
541 /* Get j neighbor index, and coordinate index */
543 j_coord_offset = DIM*jnr;
545 /* load j atom coordinates */
546 jx0 = x[j_coord_offset+DIM*0+XX];
547 jy0 = x[j_coord_offset+DIM*0+YY];
548 jz0 = x[j_coord_offset+DIM*0+ZZ];
550 /* Calculate displacement vector */
564 /* Calculate squared distance and things based on it */
565 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
566 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
567 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
568 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
570 rinv00 = gmx_invsqrt(rsq00);
571 rinv10 = gmx_invsqrt(rsq10);
572 rinv20 = gmx_invsqrt(rsq20);
573 rinv30 = gmx_invsqrt(rsq30);
575 rinvsq00 = rinv00*rinv00;
577 /* Load parameters for j particles */
579 vdwjidx0 = 3*vdwtype[jnr+0];
581 /**************************
582 * CALCULATE INTERACTIONS *
583 **************************/
587 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
588 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
589 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
591 /* BUCKINGHAM DISPERSION/REPULSION */
592 rinvsix = rinvsq00*rinvsq00*rinvsq00;
593 vvdw6 = c6_00*rinvsix;
595 vvdwexp = cexp1_00*exp(-br);
596 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
600 /* Calculate temporary vectorial force */
605 /* Update vectorial force */
609 f[j_coord_offset+DIM*0+XX] -= tx;
610 f[j_coord_offset+DIM*0+YY] -= ty;
611 f[j_coord_offset+DIM*0+ZZ] -= tz;
613 /**************************
614 * CALCULATE INTERACTIONS *
615 **************************/
621 /* Calculate table index by multiplying r with table scale and truncate to integer */
627 /* CUBIC SPLINE TABLE ELECTROSTATICS */
629 Geps = vfeps*vftab[vfitab+2];
630 Heps2 = vfeps*vfeps*vftab[vfitab+3];
632 FF = Fp+Geps+2.0*Heps2;
633 felec = -qq10*FF*vftabscale*rinv10;
637 /* Calculate temporary vectorial force */
642 /* Update vectorial force */
646 f[j_coord_offset+DIM*0+XX] -= tx;
647 f[j_coord_offset+DIM*0+YY] -= ty;
648 f[j_coord_offset+DIM*0+ZZ] -= tz;
650 /**************************
651 * CALCULATE INTERACTIONS *
652 **************************/
658 /* Calculate table index by multiplying r with table scale and truncate to integer */
664 /* CUBIC SPLINE TABLE ELECTROSTATICS */
666 Geps = vfeps*vftab[vfitab+2];
667 Heps2 = vfeps*vfeps*vftab[vfitab+3];
669 FF = Fp+Geps+2.0*Heps2;
670 felec = -qq20*FF*vftabscale*rinv20;
674 /* Calculate temporary vectorial force */
679 /* Update vectorial force */
683 f[j_coord_offset+DIM*0+XX] -= tx;
684 f[j_coord_offset+DIM*0+YY] -= ty;
685 f[j_coord_offset+DIM*0+ZZ] -= tz;
687 /**************************
688 * CALCULATE INTERACTIONS *
689 **************************/
695 /* Calculate table index by multiplying r with table scale and truncate to integer */
701 /* CUBIC SPLINE TABLE ELECTROSTATICS */
703 Geps = vfeps*vftab[vfitab+2];
704 Heps2 = vfeps*vfeps*vftab[vfitab+3];
706 FF = Fp+Geps+2.0*Heps2;
707 felec = -qq30*FF*vftabscale*rinv30;
711 /* Calculate temporary vectorial force */
716 /* Update vectorial force */
720 f[j_coord_offset+DIM*0+XX] -= tx;
721 f[j_coord_offset+DIM*0+YY] -= ty;
722 f[j_coord_offset+DIM*0+ZZ] -= tz;
724 /* Inner loop uses 172 flops */
726 /* End of innermost loop */
729 f[i_coord_offset+DIM*0+XX] += fix0;
730 f[i_coord_offset+DIM*0+YY] += fiy0;
731 f[i_coord_offset+DIM*0+ZZ] += fiz0;
735 f[i_coord_offset+DIM*1+XX] += fix1;
736 f[i_coord_offset+DIM*1+YY] += fiy1;
737 f[i_coord_offset+DIM*1+ZZ] += fiz1;
741 f[i_coord_offset+DIM*2+XX] += fix2;
742 f[i_coord_offset+DIM*2+YY] += fiy2;
743 f[i_coord_offset+DIM*2+ZZ] += fiz2;
747 f[i_coord_offset+DIM*3+XX] += fix3;
748 f[i_coord_offset+DIM*3+YY] += fiy3;
749 f[i_coord_offset+DIM*3+ZZ] += fiz3;
753 fshift[i_shift_offset+XX] += tx;
754 fshift[i_shift_offset+YY] += ty;
755 fshift[i_shift_offset+ZZ] += tz;
757 /* Increment number of inner iterations */
758 inneriter += j_index_end - j_index_start;
760 /* Outer loop uses 39 flops */
763 /* Increment number of outer iterations */
766 /* Update outer/inner flops */
768 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*39 + inneriter*172);