FP2P.hpp 93 KB
Newer Older
1 2 3 4
#ifndef FP2P_HPP
#define FP2P_HPP

#include "../../Utils/FGlobal.hpp"
5
#include "../../Utils/FMath.hpp"
6

BRAMAS Berenger's avatar
BRAMAS Berenger committed
7
#ifdef ScalFMM_USE_SSE
8
#include "../../Utils/FSse.hpp"
BRAMAS Berenger's avatar
BRAMAS Berenger committed
9 10
#endif

11 12 13 14 15 16 17 18
/**
 * @brief The FP2P class
 */
class FP2P {
public:
    /**
     *
     */
BRAMAS Berenger's avatar
BRAMAS Berenger committed
19
#ifndef ScalFMM_USE_SSE
20 21
    template <class ContainerClass>
    static void FullMutual(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
BRAMAS Berenger's avatar
BRAMAS Berenger committed
22
                           const int limiteNeighbors){
23 24 25 26 27 28 29 30 31 32 33 34

        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
BRAMAS Berenger's avatar
BRAMAS Berenger committed
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
            if( inNeighbors[idxNeighbors] ){
                const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues();
                const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];
                FReal*const sourcesForcesX = inNeighbors[idxNeighbors]->getForcesX();
                FReal*const sourcesForcesY = inNeighbors[idxNeighbors]->getForcesY();
                FReal*const sourcesForcesZ = inNeighbors[idxNeighbors]->getForcesZ();
                FReal*const sourcesPotentials = inNeighbors[idxNeighbors]->getPotentials();

                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];

                        FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                        const FReal inv_distance = FMath::Sqrt(inv_square_distance);

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        targetsForcesX[idxTarget] += dx;
                        targetsForcesY[idxTarget] += dy;
                        targetsForcesZ[idxTarget] += dz;
                        targetsPotentials[idxTarget] += inv_distance * sourcesPhysicalValues[idxSource];

                        sourcesForcesX[idxSource] -= dx;
                        sourcesForcesY[idxSource] -= dy;
                        sourcesForcesZ[idxSource] -= dz;
                        sourcesPotentials[idxSource] += inv_distance * targetsPhysicalValues[idxTarget];
                    }
                }
            }
74 75 76 77
        }

        for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
            for(int idxSource = idxTarget + 1 ; idxSource < nbParticlesTargets ; ++idxSource){
78 79 80
                FReal dx = targetsX[idxSource] - targetsX[idxTarget];
                FReal dy = targetsY[idxSource] - targetsY[idxTarget];
                FReal dz = targetsZ[idxSource] - targetsZ[idxTarget];
81

82 83
                FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                const FReal inv_distance = FMath::Sqrt(inv_square_distance);
84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

                inv_square_distance *= inv_distance;
                inv_square_distance *= targetsPhysicalValues[idxTarget] * targetsPhysicalValues[idxSource];

                dx *= inv_square_distance;
                dy *= inv_square_distance;
                dz *= inv_square_distance;

                targetsForcesX[idxTarget] += dx;
                targetsForcesY[idxTarget] += dy;
                targetsForcesZ[idxTarget] += dz;
                targetsPotentials[idxTarget] += inv_distance * targetsPhysicalValues[idxSource];

                targetsForcesX[idxSource] -= dx;
                targetsForcesY[idxSource] -= dy;
                targetsForcesZ[idxSource] -= dz;
                targetsPotentials[idxSource] += inv_distance * targetsPhysicalValues[idxTarget];
            }
        }
    }
BRAMAS Berenger's avatar
BRAMAS Berenger committed
104
#else
105 106

#ifdef ScalFMM_USE_DOUBLE_PRECISION
BRAMAS Berenger's avatar
BRAMAS Berenger committed
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
    template <class ContainerClass>
    static void FullMutual(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
                           const int limiteNeighbors){

        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        const __m128d mOne = _mm_set1_pd(1.0);

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            if( inNeighbors[idxNeighbors] ){
                const int nbParticlesSources = (inNeighbors[idxNeighbors]->getNbParticles()+1)/2;
                const __m128d*const sourcesPhysicalValues = (const __m128d*)inNeighbors[idxNeighbors]->getPhysicalValues();
                const __m128d*const sourcesX = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[0];
                const __m128d*const sourcesY = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[1];
                const __m128d*const sourcesZ = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[2];
                __m128d*const sourcesForcesX = (__m128d*)inNeighbors[idxNeighbors]->getForcesX();
                __m128d*const sourcesForcesY = (__m128d*)inNeighbors[idxNeighbors]->getForcesY();
                __m128d*const sourcesForcesZ = (__m128d*)inNeighbors[idxNeighbors]->getForcesZ();
                __m128d*const sourcesPotentials = (__m128d*)inNeighbors[idxNeighbors]->getPotentials();

                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                    const __m128d tx = _mm_load1_pd(&targetsX[idxTarget]);
                    const __m128d ty = _mm_load1_pd(&targetsY[idxTarget]);
                    const __m128d tz = _mm_load1_pd(&targetsZ[idxTarget]);
                    const __m128d tv = _mm_load1_pd(&targetsPhysicalValues[idxTarget]);
                    __m128d  tfx = _mm_setzero_pd();
                    __m128d  tfy = _mm_setzero_pd();
                    __m128d  tfz = _mm_setzero_pd();
                    __m128d  tpo = _mm_setzero_pd();

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        __m128d dx = sourcesX[idxSource] - tx;
                        __m128d dy = sourcesY[idxSource] - ty;
                        __m128d dz = sourcesZ[idxSource] - tz;

                        __m128d inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                        const __m128d inv_distance = _mm_sqrt_pd(inv_square_distance);

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        tfx += dx;
                        tfy += dy;
                        tfz += dz;
                        tpo += inv_distance * sourcesPhysicalValues[idxSource];

                        sourcesForcesX[idxSource] -= dx;
                        sourcesForcesY[idxSource] -= dy;
                        sourcesForcesZ[idxSource] -= dz;
                        sourcesPotentials[idxSource] += inv_distance * tv;
                    }

                    __attribute__((aligned(16))) double buffer[2];

                    _mm_store_pd(buffer, tfx);
                    targetsForcesX[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tfy);
                    targetsForcesY[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tfz);
                    targetsForcesZ[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tpo);
                    targetsPotentials[idxTarget] += buffer[0] + buffer[1];
                }
            }
        }

        {
            const int nbParticlesSources = (nbParticlesTargets+1)/2;
            const __m128d*const sourcesPhysicalValues = (const __m128d*)targetsPhysicalValues;
            const __m128d*const sourcesX = (const __m128d*)targetsX;
            const __m128d*const sourcesY = (const __m128d*)targetsY;
            const __m128d*const sourcesZ = (const __m128d*)targetsZ;
            __m128d*const sourcesForcesX = (__m128d*)targetsForcesX;
            __m128d*const sourcesForcesY = (__m128d*)targetsForcesY;
            __m128d*const sourcesForcesZ = (__m128d*)targetsForcesZ;
            __m128d*const sourcesPotentials = (__m128d*)targetsPotentials;

            for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                const __m128d tx = _mm_load1_pd(&targetsX[idxTarget]);
                const __m128d ty = _mm_load1_pd(&targetsY[idxTarget]);
                const __m128d tz = _mm_load1_pd(&targetsZ[idxTarget]);
                const __m128d tv = _mm_load1_pd(&targetsPhysicalValues[idxTarget]);
                __m128d  tfx = _mm_setzero_pd();
                __m128d  tfy = _mm_setzero_pd();
                __m128d  tfz = _mm_setzero_pd();
                __m128d  tpo = _mm_setzero_pd();

                for(int idxSource = (idxTarget+2)/2 ; idxSource < nbParticlesSources ; ++idxSource){
                    __m128d dx = sourcesX[idxSource] - tx;
                    __m128d dy = sourcesY[idxSource] - ty;
                    __m128d dz = sourcesZ[idxSource] - tz;

                    __m128d inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                    const __m128d inv_distance = _mm_sqrt_pd(inv_square_distance);

                    inv_square_distance *= inv_distance;
                    inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                    dx *= inv_square_distance;
                    dy *= inv_square_distance;
                    dz *= inv_square_distance;

                    tfx += dx;
                    tfy += dy;
                    tfz += dz;
                    tpo += inv_distance * sourcesPhysicalValues[idxSource];

                    sourcesForcesX[idxSource] -= dx;
                    sourcesForcesY[idxSource] -= dy;
                    sourcesForcesZ[idxSource] -= dz;
                    sourcesPotentials[idxSource] += inv_distance * tv;
                }

                __attribute__((aligned(16))) double buffer[2];

                _mm_store_pd(buffer, tfx);
                targetsForcesX[idxTarget] += buffer[0] + buffer[1];

                _mm_store_pd(buffer, tfy);
                targetsForcesY[idxTarget] += buffer[0] + buffer[1];

                _mm_store_pd(buffer, tfz);
                targetsForcesZ[idxTarget] += buffer[0] + buffer[1];

                _mm_store_pd(buffer, tpo);
                targetsPotentials[idxTarget] += buffer[0] + buffer[1];
            }
        }

        for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; idxTarget += 2){
            const int idxSource = idxTarget + 1;
            FReal dx = targetsX[idxSource] - targetsX[idxTarget];
            FReal dy = targetsY[idxSource] - targetsY[idxTarget];
            FReal dz = targetsZ[idxSource] - targetsZ[idxTarget];

            FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
            const FReal inv_distance = FMath::Sqrt(inv_square_distance);

            inv_square_distance *= inv_distance;
            inv_square_distance *= targetsPhysicalValues[idxTarget] * targetsPhysicalValues[idxSource];

            dx *= inv_square_distance;
            dy *= inv_square_distance;
            dz *= inv_square_distance;

            targetsForcesX[idxTarget] += dx;
            targetsForcesY[idxTarget] += dy;
            targetsForcesZ[idxTarget] += dz;
            targetsPotentials[idxTarget] += inv_distance * targetsPhysicalValues[idxSource];

            targetsForcesX[idxSource] -= dx;
            targetsForcesY[idxSource] -= dy;
            targetsForcesZ[idxSource] -= dz;
            targetsPotentials[idxSource] += inv_distance * targetsPhysicalValues[idxTarget];
        }
    }
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
#else
    template <class ContainerClass>
    static void FullMutual(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
                           const int limiteNeighbors){

        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        const __m128 mOne = _mm_set1_ps(1.0);

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            if( inNeighbors[idxNeighbors] ){
                const int nbParticlesSources = (inNeighbors[idxNeighbors]->getNbParticles()+3)/4;
                const __m128*const sourcesPhysicalValues = (const __m128*)inNeighbors[idxNeighbors]->getPhysicalValues();
                const __m128*const sourcesX = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[0];
                const __m128*const sourcesY = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[1];
                const __m128*const sourcesZ = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[2];
                __m128*const sourcesForcesX = (__m128*)inNeighbors[idxNeighbors]->getForcesX();
                __m128*const sourcesForcesY = (__m128*)inNeighbors[idxNeighbors]->getForcesY();
                __m128*const sourcesForcesZ = (__m128*)inNeighbors[idxNeighbors]->getForcesZ();
                __m128*const sourcesPotentials = (__m128*)inNeighbors[idxNeighbors]->getPotentials();

                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                    const __m128 tx = _mm_load1_ps(&targetsX[idxTarget]);
                    const __m128 ty = _mm_load1_ps(&targetsY[idxTarget]);
                    const __m128 tz = _mm_load1_ps(&targetsZ[idxTarget]);
                    const __m128 tv = _mm_load1_ps(&targetsPhysicalValues[idxTarget]);
                    __m128  tfx = _mm_setzero_ps();
                    __m128  tfy = _mm_setzero_ps();
                    __m128  tfz = _mm_setzero_ps();
                    __m128  tpo = _mm_setzero_ps();

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        __m128 dx = sourcesX[idxSource] - tx;
                        __m128 dy = sourcesY[idxSource] - ty;
                        __m128 dz = sourcesZ[idxSource] - tz;

                        __m128 inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                        const __m128 inv_distance = _mm_sqrt_ps(inv_square_distance);

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        tfx += dx;
                        tfy += dy;
                        tfz += dz;
                        tpo += inv_distance * sourcesPhysicalValues[idxSource];

                        sourcesForcesX[idxSource] -= dx;
                        sourcesForcesY[idxSource] -= dy;
                        sourcesForcesZ[idxSource] -= dz;
                        sourcesPotentials[idxSource] += inv_distance * tv;
                    }

                    __attribute__((aligned(16))) float buffer[4];

                    _mm_store_ps(buffer, tfx);
                    targetsForcesX[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tfy);
                    targetsForcesY[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tfz);
                    targetsForcesZ[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tpo);
                    targetsPotentials[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];
                }
            }
        }

        {
            const int nbParticlesSources = (nbParticlesTargets+3)/4;
            const __m128*const sourcesPhysicalValues = (const __m128*)targetsPhysicalValues;
            const __m128*const sourcesX = (const __m128*)targetsX;
            const __m128*const sourcesY = (const __m128*)targetsY;
            const __m128*const sourcesZ = (const __m128*)targetsZ;
            __m128*const sourcesForcesX = (__m128*)targetsForcesX;
            __m128*const sourcesForcesY = (__m128*)targetsForcesY;
            __m128*const sourcesForcesZ = (__m128*)targetsForcesZ;
            __m128*const sourcesPotentials = (__m128*)targetsPotentials;

            for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                const __m128 tx = _mm_load1_ps(&targetsX[idxTarget]);
                const __m128 ty = _mm_load1_ps(&targetsY[idxTarget]);
                const __m128 tz = _mm_load1_ps(&targetsZ[idxTarget]);
                const __m128 tv = _mm_load1_ps(&targetsPhysicalValues[idxTarget]);
                __m128  tfx = _mm_setzero_ps();
                __m128  tfy = _mm_setzero_ps();
                __m128  tfz = _mm_setzero_ps();
                __m128  tpo = _mm_setzero_ps();

                for(int idxSource = (idxTarget+2)/2 ; idxSource < nbParticlesSources ; ++idxSource){
                    __m128 dx = sourcesX[idxSource] - tx;
                    __m128 dy = sourcesY[idxSource] - ty;
                    __m128 dz = sourcesZ[idxSource] - tz;

                    __m128 inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                    const __m128 inv_distance = _mm_sqrt_ps(inv_square_distance);

                    inv_square_distance *= inv_distance;
                    inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                    dx *= inv_square_distance;
                    dy *= inv_square_distance;
                    dz *= inv_square_distance;

                    tfx += dx;
                    tfy += dy;
                    tfz += dz;
                    tpo += inv_distance * sourcesPhysicalValues[idxSource];

                    sourcesForcesX[idxSource] -= dx;
                    sourcesForcesY[idxSource] -= dy;
                    sourcesForcesZ[idxSource] -= dz;
                    sourcesPotentials[idxSource] += inv_distance * tv;
                }

                __attribute__((aligned(16))) float buffer[4];

                _mm_store_ps(buffer, tfx);
                targetsForcesX[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                _mm_store_ps(buffer, tfy);
                targetsForcesY[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                _mm_store_ps(buffer, tfz);
                targetsForcesZ[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                _mm_store_ps(buffer, tpo);
                targetsPotentials[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];
            }
        }

        for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; idxTarget += 4){
            for(int idxClose = 1 ; idxClose < 4; ++idxClose){
                const int idxSource = idxTarget + idxClose;
                FReal dx = targetsX[idxSource] - targetsX[idxTarget];
                FReal dy = targetsY[idxSource] - targetsY[idxTarget];
                FReal dz = targetsZ[idxSource] - targetsZ[idxTarget];

                FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                const FReal inv_distance = FMath::Sqrt(inv_square_distance);

                inv_square_distance *= inv_distance;
                inv_square_distance *= targetsPhysicalValues[idxTarget] * targetsPhysicalValues[idxSource];

                dx *= inv_square_distance;
                dy *= inv_square_distance;
                dz *= inv_square_distance;

                targetsForcesX[idxTarget] += dx;
                targetsForcesY[idxTarget] += dy;
                targetsForcesZ[idxTarget] += dz;
                targetsPotentials[idxTarget] += inv_distance * targetsPhysicalValues[idxSource];

                targetsForcesX[idxSource] -= dx;
                targetsForcesY[idxSource] -= dy;
                targetsForcesZ[idxSource] -= dz;
                targetsPotentials[idxSource] += inv_distance * targetsPhysicalValues[idxTarget];
            }
        }
    }
#endif

BRAMAS Berenger's avatar
BRAMAS Berenger committed
454
#endif
455 456 457 458

    /**
     *
     */
BRAMAS Berenger's avatar
BRAMAS Berenger committed
459
#ifndef ScalFMM_USE_SSE
460 461
    template <class ContainerClass>
    static void FullRemote(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
BRAMAS Berenger's avatar
BRAMAS Berenger committed
462
                           const int limiteNeighbors){
463 464 465 466 467 468 469 470 471 472 473
        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
BRAMAS Berenger's avatar
BRAMAS Berenger committed
474 475 476 477 478 479
            if( inNeighbors[idxNeighbors] ){
                const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues();
                const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];
480

BRAMAS Berenger's avatar
BRAMAS Berenger committed
481 482
                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){

483
                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
484 485 486
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];
487

488 489
                        FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                        const FReal inv_distance = FMath::Sqrt(inv_square_distance);
490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        targetsForcesX[idxTarget] += dx;
                        targetsForcesY[idxTarget] += dy;
                        targetsForcesZ[idxTarget] += dz;
                        targetsPotentials[idxTarget] += inv_distance * sourcesPhysicalValues[idxSource];
                    }
                }
            }
        }
    }
BRAMAS Berenger's avatar
BRAMAS Berenger committed
507
#else
508 509

#ifdef ScalFMM_USE_DOUBLE_PRECISION
BRAMAS Berenger's avatar
BRAMAS Berenger committed
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
    template <class ContainerClass>
    static void FullRemote(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
                           const int limiteNeighbors){
        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        const __m128d mOne = _mm_set1_pd(1.0);

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            if( inNeighbors[idxNeighbors] ){
527
                const int nbParticlesSources = (inNeighbors[idxNeighbors]->getNbParticles()+1)/2;
BRAMAS Berenger's avatar
BRAMAS Berenger committed
528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580
                const __m128d*const sourcesPhysicalValues = (const __m128d*)inNeighbors[idxNeighbors]->getPhysicalValues();
                const __m128d*const sourcesX = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[0];
                const __m128d*const sourcesY = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[1];
                const __m128d*const sourcesZ = (const __m128d*)inNeighbors[idxNeighbors]->getPositions()[2];

                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                    const __m128d tx = _mm_load1_pd(&targetsX[idxTarget]);
                    const __m128d ty = _mm_load1_pd(&targetsY[idxTarget]);
                    const __m128d tz = _mm_load1_pd(&targetsZ[idxTarget]);
                    const __m128d tv = _mm_load1_pd(&targetsPhysicalValues[idxTarget]);
                    __m128d  tfx = _mm_setzero_pd();
                    __m128d  tfy = _mm_setzero_pd();
                    __m128d  tfz = _mm_setzero_pd();
                    __m128d  tpo = _mm_setzero_pd();

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        __m128d dx = sourcesX[idxSource] - tx;
                        __m128d dy = sourcesY[idxSource] - ty;
                        __m128d dz = sourcesZ[idxSource] - tz;

                        __m128d inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                        const __m128d inv_distance = _mm_sqrt_pd(inv_square_distance);

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        tfx += dx;
                        tfy += dy;
                        tfz += dz;
                        tpo += inv_distance * sourcesPhysicalValues[idxSource];
                    }

                    __attribute__((aligned(16))) double buffer[2];

                    _mm_store_pd(buffer, tfx);
                    targetsForcesX[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tfy);
                    targetsForcesY[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tfz);
                    targetsForcesZ[idxTarget] += buffer[0] + buffer[1];

                    _mm_store_pd(buffer, tpo);
                    targetsPotentials[idxTarget] += buffer[0] + buffer[1];
                }
            }
        }
    }
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654
#else
    template <class ContainerClass>
    static void FullRemote(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
                           const int limiteNeighbors){
        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        const __m128 mOne = _mm_set1_ps(1.0);

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            if( inNeighbors[idxNeighbors] ){
                const int nbParticlesSources = (inNeighbors[idxNeighbors]->getNbParticles()+3)/4;
                const __m128*const sourcesPhysicalValues = (const __m128*)inNeighbors[idxNeighbors]->getPhysicalValues();
                const __m128*const sourcesX = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[0];
                const __m128*const sourcesY = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[1];
                const __m128*const sourcesZ = (const __m128*)inNeighbors[idxNeighbors]->getPositions()[2];

                for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                    const __m128 tx = _mm_load1_ps(&targetsX[idxTarget]);
                    const __m128 ty = _mm_load1_ps(&targetsY[idxTarget]);
                    const __m128 tz = _mm_load1_ps(&targetsZ[idxTarget]);
                    const __m128 tv = _mm_load1_ps(&targetsPhysicalValues[idxTarget]);
                    __m128  tfx = _mm_setzero_ps();
                    __m128  tfy = _mm_setzero_ps();
                    __m128  tfz = _mm_setzero_ps();
                    __m128  tpo = _mm_setzero_ps();

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        __m128 dx = sourcesX[idxSource] - tx;
                        __m128 dy = sourcesY[idxSource] - ty;
                        __m128 dz = sourcesZ[idxSource] - tz;

                        __m128 inv_square_distance = mOne / (dx*dx + dy*dy + dz*dz);
                        const __m128 inv_distance = _mm_sqrt_ps(inv_square_distance);

                        inv_square_distance *= inv_distance;
                        inv_square_distance *= tv * sourcesPhysicalValues[idxSource];

                        dx *= inv_square_distance;
                        dy *= inv_square_distance;
                        dz *= inv_square_distance;

                        tfx += dx;
                        tfy += dy;
                        tfz += dz;
                        tpo += inv_distance * sourcesPhysicalValues[idxSource];
                    }

                    __attribute__((aligned(16))) float buffer[4];

                    _mm_store_ps(buffer, tfx);
                    targetsForcesX[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tfy);
                    targetsForcesY[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tfz);
                    targetsForcesZ[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];

                    _mm_store_ps(buffer, tpo);
                    targetsPotentials[idxTarget] += buffer[0] + buffer[1] + buffer[2] + buffer[3];
                }
            }
        }
    }
#endif

BRAMAS Berenger's avatar
BRAMAS Berenger committed
655
#endif
656 657 658 659 660 661 662 663 664 665 666 667 668

    /** P2P mutual interaction,
      * this function computes the interaction for 2 particles.
      *
      * Formulas are:
      * \f[
      * F = q_1 * q_2 / r^2
      * P_1 = q_2 / r ; P_2 = q_1 / r
      * \f]
      * In details :
      * \f$ F(x) = \frac{ \Delta_x * q_1 * q_2 }{ r^2 } = \Delta_x * F \f$
      */
    static void MutualParticles(const FReal sourceX,const FReal sourceY,const FReal sourceZ, const FReal sourcePhysicalValue,
BRAMAS Berenger's avatar
BRAMAS Berenger committed
669 670 671 672
                                FReal* sourceForceX, FReal* sourceForceY, FReal* sourceForceZ, FReal* sourcePotential,
                                const FReal targetX,const FReal targetY,const FReal targetZ, const FReal targetPhysicalValue,
                                FReal* targetForceX, FReal* targetForceY, FReal* targetForceZ, FReal* targetPotential
                                ){
673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709
        FReal dx = sourceX - targetX;
        FReal dy = sourceY - targetY;
        FReal dz = sourceZ - targetZ;

        FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
        FReal inv_distance = FMath::Sqrt(inv_square_distance);

        inv_square_distance *= inv_distance;
        inv_square_distance *= targetPhysicalValue * sourcePhysicalValue;

        dx *= inv_square_distance;
        dy *= inv_square_distance;
        dz *= inv_square_distance;

        *targetForceX += dx;
        *targetForceY += dy;
        *targetForceZ += dz;
        *targetPotential += ( inv_distance * sourcePhysicalValue );

        *sourceForceX -= dx;
        *sourceForceY -= dy;
        *sourceForceZ -= dz;
        *sourcePotential += ( inv_distance * targetPhysicalValue );
    }

    /** P2P mutual interaction,
      * this function computes the interaction for 2 particles.
      *
      * Formulas are:
      * \f[
      * F = q_1 * q_2 / r^2
      * P_1 = q_2 / r ; P_2 = q_1 / r
      * \f]
      * In details :
      * \f$ F(x) = \frac{ \Delta_x * q_1 * q_2 }{ r^2 } = \Delta_x * F \f$
      */
    static void NonMutualParticles(const FReal sourceX,const FReal sourceY,const FReal sourceZ, const FReal sourcePhysicalValue,
BRAMAS Berenger's avatar
BRAMAS Berenger committed
710 711
                                   const FReal targetX,const FReal targetY,const FReal targetZ, const FReal targetPhysicalValue,
                                   FReal* targetForceX, FReal* targetForceY, FReal* targetForceZ, FReal* targetPotential){
712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733
        FReal dx = sourceX - targetX;
        FReal dy = sourceY - targetY;
        FReal dz = sourceZ - targetZ;

        FReal inv_square_distance = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
        FReal inv_distance = FMath::Sqrt(inv_square_distance);

        inv_square_distance *= inv_distance;
        inv_square_distance *= targetPhysicalValue * sourcePhysicalValue;

        dx *= inv_square_distance;
        dy *= inv_square_distance;
        dz *= inv_square_distance;

        *targetForceX += dx;
        *targetForceY += dy;
        *targetForceZ += dz;
        *targetPotential += ( inv_distance * sourcePhysicalValue );
    }

    template <class ContainerClass>
    static void FullMutualLJ(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
BRAMAS Berenger's avatar
BRAMAS Berenger committed
734
                             const int limiteNeighbors){
735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829

        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                if( inNeighbors[idxNeighbors] ){
                    const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                    const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues();
                    const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                    const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                    const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];
                    FReal*const sourcesForcesX = inNeighbors[idxNeighbors]->getForcesX();
                    FReal*const sourcesForcesY = inNeighbors[idxNeighbors]->getForcesY();
                    FReal*const sourcesForcesZ = inNeighbors[idxNeighbors]->getForcesZ();
                    FReal*const sourcesPotentials = inNeighbors[idxNeighbors]->getPotentials();

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];

                        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;
                        FReal inv_distance_pow6 = inv_distance_pow3 * inv_distance_pow3;
                        FReal inv_distance_pow8 = inv_distance_pow6 * inv_distance_pow2;

                        FReal coef = ((targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource])
                                      * (FReal(12.0)*inv_distance_pow6*inv_distance_pow8 - FReal(6.0)*inv_distance_pow8));
                        FReal potentialCoef = (inv_distance_pow6*inv_distance_pow6-inv_distance_pow6);

                        dx *= coef;
                        dy *= coef;
                        dz *= coef;

                        targetsForcesX[idxTarget] += dx;
                        targetsForcesY[idxTarget] += dy;
                        targetsForcesZ[idxTarget] += dz;
                        targetsPotentials[idxTarget] += ( potentialCoef * sourcesPhysicalValues[idxSource] );

                        sourcesForcesX[idxSource] -= dx;
                        sourcesForcesY[idxSource] -= dy;
                        sourcesForcesZ[idxSource] -= dz;
                        sourcesPotentials[idxSource] += potentialCoef * targetsPhysicalValues[idxTarget];
                    }
                }
            }
        }

        for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
            for(int idxSource = idxTarget + 1 ; idxSource < nbParticlesTargets ; ++idxSource){
                FReal dx = targetsX[idxSource] - targetsX[idxTarget];
                FReal dy = targetsY[idxSource] - targetsY[idxTarget];
                FReal dz = targetsZ[idxSource] - targetsZ[idxTarget];

                FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;
                FReal inv_distance_pow6 = inv_distance_pow3 * inv_distance_pow3;
                FReal inv_distance_pow8 = inv_distance_pow6 * inv_distance_pow2;

                FReal coef = ((targetsPhysicalValues[idxTarget] * targetsPhysicalValues[idxSource])
                              * (FReal(12.0)*inv_distance_pow6*inv_distance_pow8 - FReal(6.0)*inv_distance_pow8));
                FReal potentialCoef = (inv_distance_pow6*inv_distance_pow6-inv_distance_pow6);

                dx *= coef;
                dy *= coef;
                dz *= coef;

                targetsForcesX[idxTarget] += dx;
                targetsForcesY[idxTarget] += dy;
                targetsForcesZ[idxTarget] += dz;
                targetsPotentials[idxTarget] += ( potentialCoef * targetsPhysicalValues[idxSource] );

                targetsForcesX[idxSource] -= dx;
                targetsForcesY[idxSource] -= dy;
                targetsForcesZ[idxSource] -= dz;
                targetsPotentials[idxSource] += potentialCoef * targetsPhysicalValues[idxTarget];
            }
        }
    }

    /**
     *
     */
    template <class ContainerClass>
    static void FullRemoteLJ(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
BRAMAS Berenger's avatar
BRAMAS Berenger committed
830
                             const int limiteNeighbors){
831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974
        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];
        FReal*const targetsForcesX = inTargets->getForcesX();
        FReal*const targetsForcesY = inTargets->getForcesY();
        FReal*const targetsForcesZ = inTargets->getForcesZ();
        FReal*const targetsPotentials = inTargets->getPotentials();

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                if( inNeighbors[idxNeighbors] ){
                    const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                    const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues();
                    const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                    const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                    const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        // lenard-jones potential
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];

                        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
                        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;
                        FReal inv_distance_pow6 = inv_distance_pow3 * inv_distance_pow3;
                        FReal inv_distance_pow8 = inv_distance_pow6 * inv_distance_pow2;

                        FReal coef = ((targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource])
                                      * (FReal(12.0)*inv_distance_pow6*inv_distance_pow8 - FReal(6.0)*inv_distance_pow8));

                        dx *= coef;
                        dy *= coef;
                        dz *= coef;

                        targetsForcesX[idxTarget] += dx;
                        targetsForcesY[idxTarget] += dy;
                        targetsForcesZ[idxTarget] += dz;
                        targetsPotentials[idxTarget] += ( (inv_distance_pow6*inv_distance_pow6-inv_distance_pow6) * sourcesPhysicalValues[idxSource] );
                    }
                }
            }
        }
    }


    /**
     * @brief NonMutualParticlesLJ
     * @param sourceX
     * @param sourceY
     * @param sourceZ
     * @param sourcePhysicalValue
     * @param targetX
     * @param targetY
     * @param targetZ
     * @param targetPhysicalValue
     * @param targetForceX
     * @param targetForceY
     * @param targetForceZ
     * @param targetPotential
     */
    static void NonMutualParticlesLJ(const FReal sourceX,const FReal sourceY,const FReal sourceZ, const FReal sourcePhysicalValue,
                                     const FReal targetX,const FReal targetY,const FReal targetZ, const FReal targetPhysicalValue,
                                     FReal* targetForceX, FReal* targetForceY, FReal* targetForceZ, FReal* targetPotential){
        // lenard-jones potential
        FReal dx = sourceX - targetX;
        FReal dy = sourceY - targetY;
        FReal dz = sourceZ - targetZ;

        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;
        FReal inv_distance_pow6 = inv_distance_pow3 * inv_distance_pow3;
        FReal inv_distance_pow8 = inv_distance_pow6 * inv_distance_pow2;

        FReal coef = ((targetPhysicalValue * sourcePhysicalValue) * (FReal(12.0)*inv_distance_pow6*inv_distance_pow8
                                                                     - FReal(6.0)*inv_distance_pow8));

        dx *= coef;
        dy *= coef;
        dz *= coef;

        (*targetForceX) += dx;
        (*targetForceY) += dy;
        (*targetForceZ) += dz;
        (*targetPotential) += ( (inv_distance_pow6*inv_distance_pow6-inv_distance_pow6) * sourcePhysicalValue );
    }

    /**
     * @brief MutualParticlesLJ
     * @param sourceX
     * @param sourceY
     * @param sourceZ
     * @param sourcePhysicalValue
     * @param sourceForceX
     * @param sourceForceY
     * @param sourceForceZ
     * @param sourcePotential
     * @param targetX
     * @param targetY
     * @param targetZ
     * @param targetPhysicalValue
     * @param targetForceX
     * @param targetForceY
     * @param targetForceZ
     * @param targetPotential
     */
    static void MutualParticlesLJ(const FReal sourceX,const FReal sourceY,const FReal sourceZ, const FReal sourcePhysicalValue,
                                  FReal* sourceForceX, FReal* sourceForceY, FReal* sourceForceZ, FReal* sourcePotential,
                                  const FReal targetX,const FReal targetY,const FReal targetZ, const FReal targetPhysicalValue,
                                  FReal* targetForceX, FReal* targetForceY, FReal* targetForceZ, FReal* targetPotential
                                  ){
        // lenard-jones potential
        FReal dx = sourceX - targetX;
        FReal dy = sourceY - targetY;
        FReal dz = sourceZ - targetZ;

        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz);
        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;
        FReal inv_distance_pow6 = inv_distance_pow3 * inv_distance_pow3;
        FReal inv_distance_pow8 = inv_distance_pow6 * inv_distance_pow2;

        FReal coef = ((targetPhysicalValue * sourcePhysicalValue) * (FReal(12.0)*inv_distance_pow6*inv_distance_pow8
                                                                     - FReal(6.0)*inv_distance_pow8));
        FReal potentialCoef = (inv_distance_pow6*inv_distance_pow6-inv_distance_pow6);

        dx *= coef;
        dy *= coef;
        dz *= coef;

        (*targetForceX) += dx;
        (*targetForceY) += dy;
        (*targetForceZ) += dz;
        (*targetPotential) += ( potentialCoef * sourcePhysicalValue );

        (*sourceForceX) -= dx;
        (*sourceForceY) -= dy;
        (*sourceForceZ) -= dz;
        (*sourcePotential) += ( potentialCoef * targetPhysicalValue );
    }
975 976 977 978 979

  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
980
  // R_IJ
981 982
  // PB: The following functions take the MatrixKernel as input arguments for more generic implementation 
  // Besides this MatrixKernel is already updated with any extra parameter (e.g. core width).
983 984 985 986 987 988 989 990
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////
  //////////////////////////////////////////////////////////////////////////////////////////////////////////

    /**
     * @brief FullMutualRIJ
     */
991
    template <class ContainerClass, typename MatrixKernelClass>
992
    static void FullMutualRIJ(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
993 994 995
                              const int limiteNeighbors, const MatrixKernelClass *const MatrixKernel){

      const double CoreWidth2 = MatrixKernel->getCoreWidth2(); //PB: TODO directly call evaluateBlock
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014

        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
            for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
                if( inNeighbors[idxNeighbors] ){
                    const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                    const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                    const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                    const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];

1015
                        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz + CoreWidth2);
1016 1017 1018 1019 1020 1021
                        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;

                        FReal r[3]={dx,dy,dz};

                        for(unsigned int i = 0 ; i < 3 ; ++i){
1022 1023 1024 1025 1026 1027 1028 1029 1030
                          FReal*const targetsPotentials = inTargets->getPotentials(i);
                          FReal*const targetsForcesX = inTargets->getForcesX(i);
                          FReal*const targetsForcesY = inTargets->getForcesY(i);
                          FReal*const targetsForcesZ = inTargets->getForcesZ(i);
                          FReal*const sourcesPotentials = inNeighbors[idxNeighbors]->getPotentials(i);
                          FReal*const sourcesForcesX = inNeighbors[idxNeighbors]->getForcesX(i);
                          FReal*const sourcesForcesY = inNeighbors[idxNeighbors]->getForcesY(i);
                          FReal*const sourcesForcesZ = inNeighbors[idxNeighbors]->getForcesZ(i);

1031
                          FReal ri2=r[i]*r[i];
1032

1033
                          for(unsigned int j = 0 ; j < 3 ; ++j){
1034 1035 1036
                            const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues(j);
                            const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues(j);

1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
                            // potentials
                            FReal potentialCoef;
                            if(i==j)
                              potentialCoef = inv_distance - ri2 * inv_distance_pow3;
                            else
                              potentialCoef = - r[i] * r[j] * inv_distance_pow3;

                            // forces
                            FReal rj2=r[j]*r[j];

                            FReal coef[3]; 
                            for(unsigned int k = 0 ; k < 3 ; ++k)
1049
                              coef[k]= -(targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource]);
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066

                            // Grad of RIJ kernel is RIJK kernel => use same expression as in FInterpMatrixKernel
                            for(unsigned int k = 0 ; k < 3 ; ++k){
                              if(i==j){
                                if(j==k) //i=j=k
                                  coef[k] *= FReal(3.) * ( FReal(-1.) + ri2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                                else //i=j!=k
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[k] * inv_distance_pow3;
                              }
                              else{ //(i!=j)
                                if(i==k) //i=k!=j
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[j] * inv_distance_pow3;
                                else if(j==k) //i!=k=j
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * rj2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                                else //i!=k!=j
                                  coef[k] *= FReal(3.) * r[i] * r[j] * r[k] * inv_distance_pow2 * inv_distance_pow3;
                              }
1067
                            }// k
1068

1069 1070 1071 1072
                            targetsForcesX[idxTarget] += coef[0];
                            targetsForcesY[idxTarget] += coef[1];
                            targetsForcesZ[idxTarget] += coef[2];
                            targetsPotentials[idxTarget] += ( potentialCoef * sourcesPhysicalValues[idxSource] );
1073

1074 1075 1076 1077
                            sourcesForcesX[idxSource] -= coef[0];
                            sourcesForcesY[idxSource] -= coef[1];
                            sourcesForcesZ[idxSource] -= coef[2];
                            sourcesPotentials[idxSource] += potentialCoef * targetsPhysicalValues[idxTarget];
1078 1079

                             
1080 1081
                          }// j
                        }// i
1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
                    }
                }
            }
        }

        for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
            for(int idxSource = idxTarget + 1 ; idxSource < nbParticlesTargets ; ++idxSource){
                FReal dx = targetsX[idxSource] - targetsX[idxTarget];
                FReal dy = targetsY[idxSource] - targetsY[idxTarget];
                FReal dz = targetsZ[idxSource] - targetsZ[idxTarget];

1093
                FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz + CoreWidth2);
1094 1095 1096 1097 1098 1099
                FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;

                FReal r[3]={dx,dy,dz};

                for(unsigned int i = 0 ; i < 3 ; ++i){
1100 1101 1102 1103
                  FReal*const targetsPotentials = inTargets->getPotentials(i);
                  FReal*const targetsForcesX = inTargets->getForcesX(i);
                  FReal*const targetsForcesY = inTargets->getForcesY(i);
                  FReal*const targetsForcesZ = inTargets->getForcesZ(i);
1104
                  FReal ri2=r[i]*r[i];
1105

1106
                  for(unsigned int j = 0 ; j < 3 ; ++j){
1107
                    const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues(j);
1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120

                    // potentials
                    FReal potentialCoef;
                    if(i==j)
                      potentialCoef = inv_distance - ri2 * inv_distance_pow3;
                    else
                      potentialCoef = - r[i] * r[j] * inv_distance_pow3;

                    // forces
                    FReal rj2=r[j]*r[j];

                    FReal coef[3]; 
                    for(unsigned int k = 0 ; k < 3 ; ++k)
1121
                      coef[k]= -(targetsPhysicalValues[idxTarget] * targetsPhysicalValues[idxSource]);
1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138

                    // Grad of RIJ kernel is RIJK kernel => use same expression as in FInterpMatrixKernel
                    for(unsigned int k = 0 ; k < 3 ; ++k){
                      if(i==j){
                        if(j==k) //i=j=k
                          coef[k] *= FReal(3.) * ( FReal(-1.) + ri2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                        else //i=j!=k
                          coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[k] * inv_distance_pow3;
                      }
                      else{ //(i!=j)
                        if(i==k) //i=k!=j
                          coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[j] * inv_distance_pow3;
                        else if(j==k) //i!=k=j
                          coef[k] *= ( FReal(-1.) + FReal(3.) * rj2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                        else //i!=k!=j
                          coef[k] *= FReal(3.) * r[i] * r[j] * r[k] * inv_distance_pow2 * inv_distance_pow3;
                      }
1139
                    }// k
1140 1141


1142 1143 1144 1145
                    targetsForcesX[idxTarget] += coef[0];
                    targetsForcesY[idxTarget] += coef[1];
                    targetsForcesZ[idxTarget] += coef[2];
                    targetsPotentials[idxTarget] += ( potentialCoef * targetsPhysicalValues[idxSource] );
1146

1147 1148 1149 1150 1151 1152
                    targetsForcesX[idxSource] -= coef[0];
                    targetsForcesY[idxSource] -= coef[1];
                    targetsForcesZ[idxSource] -= coef[2];
                    targetsPotentials[idxSource] += potentialCoef * targetsPhysicalValues[idxTarget];
                  }// j
                }// i
1153 1154 1155 1156 1157 1158 1159 1160

            }
        }
    }

    /**
     * @brief FullRemoteRIJ
     */
1161
    template <class ContainerClass, typename MatrixKernelClass>
1162
    static void FullRemoteRIJ(ContainerClass* const FRestrict inTargets, ContainerClass* const inNeighbors[],
1163 1164 1165
                             const int limiteNeighbors, const MatrixKernelClass *const MatrixKernel){

      const double CoreWidth2 = MatrixKernel->getCoreWidth2(); //PB: TODO directly call evaluateBlock
1166

1167 1168 1169 1170 1171 1172
        const int nbParticlesTargets = inTargets->getNbParticles();
        const FReal*const targetsX = inTargets->getPositions()[0];
        const FReal*const targetsY = inTargets->getPositions()[1];
        const FReal*const targetsZ = inTargets->getPositions()[2];

        for(int idxNeighbors = 0 ; idxNeighbors < limiteNeighbors ; ++idxNeighbors){
1173
          for(int idxTarget = 0 ; idxTarget < nbParticlesTargets ; ++idxTarget){
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
                if( inNeighbors[idxNeighbors] ){
                    const int nbParticlesSources = inNeighbors[idxNeighbors]->getNbParticles();
                    const FReal*const sourcesX = inNeighbors[idxNeighbors]->getPositions()[0];
                    const FReal*const sourcesY = inNeighbors[idxNeighbors]->getPositions()[1];
                    const FReal*const sourcesZ = inNeighbors[idxNeighbors]->getPositions()[2];

                    for(int idxSource = 0 ; idxSource < nbParticlesSources ; ++idxSource){
                        // potential
                        FReal dx = sourcesX[idxSource] - targetsX[idxTarget];
                        FReal dy = sourcesY[idxSource] - targetsY[idxTarget];
                        FReal dz = sourcesZ[idxSource] - targetsZ[idxTarget];

1186
                        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz + CoreWidth2);
1187 1188 1189 1190 1191 1192
                        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
                        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;

                        FReal r[3]={dx,dy,dz};

                        for(unsigned int i = 0 ; i < 3 ; ++i){
1193 1194 1195 1196
                          FReal*const targetsPotentials = inTargets->getPotentials(i);
                          FReal*const targetsForcesX = inTargets->getForcesX(i);
                          FReal*const targetsForcesY = inTargets->getForcesY(i);
                          FReal*const targetsForcesZ = inTargets->getForcesZ(i);
1197
                          FReal ri2=r[i]*r[i];
1198

1199
                          for(unsigned int j = 0 ; j < 3 ; ++j){
1200 1201
                            const FReal*const targetsPhysicalValues = inTargets->getPhysicalValues(j);
                            const FReal*const sourcesPhysicalValues = inNeighbors[idxNeighbors]->getPhysicalValues(j);
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214

                            // potentials
                            FReal potentialCoef;
                            if(i==j)
                              potentialCoef = inv_distance - ri2 * inv_distance_pow3;
                            else
                              potentialCoef = - r[i] * r[j] * inv_distance_pow3;

                            // forces
                            FReal rj2=r[j]*r[j];

                            FReal coef[3]; 
                            for(unsigned int k = 0 ; k < 3 ; ++k)
1215
                              coef[k]= -(targetsPhysicalValues[idxTarget] * sourcesPhysicalValues[idxSource]);
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232

                            // Grad of RIJ kernel is RIJK kernel => use same expression as in FInterpMatrixKernel
                            for(unsigned int k = 0 ; k < 3 ; ++k){
                              if(i==j){
                                if(j==k) //i=j=k
                                  coef[k] *= FReal(3.) * ( FReal(-1.) + ri2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                                else //i=j!=k
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[k] * inv_distance_pow3;
                              }
                              else{ //(i!=j)
                                if(i==k) //i=k!=j
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[j] * inv_distance_pow3;
                                else if(j==k) //i!=k=j
                                  coef[k] *= ( FReal(-1.) + FReal(3.) * rj2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                                else //i!=k!=j
                                  coef[k] *= FReal(3.) * r[i] * r[j] * r[k] * inv_distance_pow2 * inv_distance_pow3;
                              }
1233
                            }// k
1234

1235 1236 1237 1238
                            targetsForcesX[idxTarget] += coef[0];
                            targetsForcesY[idxTarget] += coef[1];
                            targetsForcesZ[idxTarget] += coef[2];
                            targetsPotentials[idxTarget] += ( potentialCoef * sourcesPhysicalValues[idxSource] );
1239

1240 1241
                          }// j
                        }// i
1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268

                    }
                }
            }
        }
    }


    /**
     * @brief MutualParticlesRIJ
     * @param sourceX
     * @param sourceY
     * @param sourceZ
     * @param sourcePhysicalValue
     * @param sourceForceX
     * @param sourceForceY
     * @param sourceForceZ
     * @param sourcePotential
     * @param targetX
     * @param targetY
     * @param targetZ
     * @param targetPhysicalValue
     * @param targetForceX
     * @param targetForceY
     * @param targetForceZ
     * @param targetPotential
     */
1269
  template<typename MatrixKernelClass>
1270
    static void MutualParticlesRIJ(const FReal sourceX,const FReal sourceY,const FReal sourceZ, const FReal* sourcePhysicalValue,
1271
                                  FReal* sourceForceX, FReal* sourceForceY, FReal* sourceForceZ, FReal* sourcePotential,
1272
                                  const FReal targetX,const FReal targetY,const FReal targetZ, const FReal* targetPhysicalValue,
1273 1274 1275 1276 1277
                                  FReal* targetForceX, FReal* targetForceY, FReal* targetForceZ, FReal* targetPotential,
                                  const MatrixKernelClass *const MatrixKernel){

      const double CoreWidth2 = MatrixKernel->getCoreWidth2(); //PB: TODO directly call evaluateBlock

1278 1279 1280 1281 1282
        // GradGradR potential
        FReal dx = sourceX - targetX;
        FReal dy = sourceY - targetY;
        FReal dz = sourceZ - targetZ;

1283
        FReal inv_distance_pow2 = FReal(1.0) / (dx*dx + dy*dy + dz*dz + CoreWidth2);
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304
        FReal inv_distance = FMath::Sqrt(inv_distance_pow2);
        FReal inv_distance_pow3 = inv_distance_pow2 * inv_distance;

        FReal r[3]={dx,dy,dz};

        for(unsigned int i = 0 ; i < 3 ; ++i){
          FReal ri2=r[i]*r[i];
          for(unsigned int j = 0 ; j < 3 ; ++j){

            // potentials
            FReal potentialCoef;
            if(i==j)
              potentialCoef = inv_distance - ri2 * inv_distance_pow3;
            else
              potentialCoef = - r[i] * r[j] * inv_distance_pow3;

            // forces
            FReal rj2=r[j]*r[j];

            FReal coef[3]; 
            for(unsigned int k = 0 ; k < 3 ; ++k)
1305
              coef[k]= -(targetPhysicalValue[j] * sourcePhysicalValue[j]);
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322

            // Grad of RIJ kernel is RIJK kernel => use same expression as in FInterpMatrixKernel
            for(unsigned int k = 0 ; k < 3 ; ++k){
              if(i==j){
                if(j==k) //i=j=k
                  coef[k] *= FReal(3.) * ( FReal(-1.) + ri2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                else //i=j!=k
                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[k] * inv_distance_pow3;
              }
              else{ //(i!=j)
                if(i==k) //i=k!=j
                  coef[k] *= ( FReal(-1.) + FReal(3.) * ri2 * inv_distance_pow2 ) * r[j] * inv_distance_pow3;
                else if(j==k) //i!=k=j
                  coef[k] *= ( FReal(-1.) + FReal(3.) * rj2 * inv_distance_pow2 ) * r[i] * inv_distance_pow3;
                else //i!=k!=j
                  coef[k] *= FReal(3.) * r[i] * r[j] * r[k] * inv_distance_pow2 * inv_distance_pow3;
              }
1323
            }// k
1324

1325 1326 1327 1328
            targetForceX[i] += coef[0];
            targetForceY[i] += coef[1];
            targetForceZ[i] += coef[2];
            targetPotential[i] += ( potentialCoef * sourcePhysicalValue[j] );
1329

1330 1331 1332 1333
            sourceForceX[i] -= coef[0];
            sourceForceY[i] -= coef[1];
            sourceForceZ[i] -= coef[2];
            sourcePotential[i] += ( potentialCoef * targetPhysicalValue[j] );
1334

1335 1336
          }// j
        }// i
1337 1338 1339

    }

1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489