Parallelized charge spreading for CPU based PME.

platforms/cuda/src/CpuPme.cpp

@@ -90,7 +90,7 @@ static int getNumProcessors() {
 #endif
 }
 
-static void spreadCharge(float* posq, float* grid, int gridx, int gridy, int gridz, int numParticles, Vec3 periodicBoxSize) {
+static void spreadCharge(int start, int end, float* posq, float* grid, int gridx, int gridy, int gridz, int numParticles, Vec3 periodicBoxSize) {
     float temp[4];
     __m128 boxSize = _mm_set_ps(0, (float) periodicBoxSize[2], (float) periodicBoxSize[1], (float) periodicBoxSize[0]);
     __m128 invBoxSize = _mm_set_ps(0, (float) (1/periodicBoxSize[2]), (float) (1/periodicBoxSize[1]), (float) (1/periodicBoxSize[0]));
@@ -100,7 +100,7 @@ static void spreadCharge(float* posq, float* grid, int gridx, int gridy, int gri
     __m128 scale = _mm_set1_ps(1.0f/(PME_ORDER-1));
     const float epsilonFactor = sqrt(ONE_4PI_EPS0);
     memset(grid, 0, sizeof(float)*gridx*gridy*gridz);
-    for (int i = 0; i < numParticles; i++) {
+    for (int i = start; i < end; i++) {
         // Find the position relative to the nearest grid point.
         
         __m128 pos = _mm_load_ps(&posq[4*i]);
@@ -303,6 +303,9 @@ static void interpolateForces(int start, int end, float* posq, float* force, flo
         int gridIndexX = _mm_extract_epi32(gridIndex, 0);
         int gridIndexY = _mm_extract_epi32(gridIndex, 1);
         int gridIndexZ = _mm_extract_epi32(gridIndex, 2);
+        __m128 zdata[PME_ORDER];
+        for (int j = 0; j < PME_ORDER; j++)
+            zdata[j] = _mm_set_ps(0, extractFloat(ddata[j], 2), extractFloat(data[j], 2), extractFloat(data[j], 2));
         __m128 f = _mm_set1_ps(0);
         for (int ix = 0; ix < PME_ORDER; ix++) {
             int xbase = gridIndexX+ix;
@@ -324,10 +327,7 @@ static void interpolateForces(int start, int end, float* posq, float* force, flo
                     int zindex = gridIndexZ+iz;
                     zindex -= (zindex >= gridz ? gridz : 0);
                     __m128 gridValue = _mm_set1_ps(grid[ybase+zindex]);
-                    float dz = extractFloat(data[iz], 2);
-                    float ddz = extractFloat(ddata[iz], 2);
-                    __m128 zdata = _mm_set_ps(0, ddz, dz, dz);
-                    f = _mm_add_ps(f, _mm_mul_ps(xydata, _mm_mul_ps(zdata, gridValue)));
+                    f = _mm_add_ps(f, _mm_mul_ps(xydata, _mm_mul_ps(zdata[iz], gridValue)));
                 }
             }
         }
@@ -340,7 +340,8 @@ class CpuPme::ThreadData {
 public:
     CpuPme& owner;
     int index;
-    ThreadData(CpuPme& owner, int index) : owner(owner), index(index) {
+    float* tempGrid;
+    ThreadData(CpuPme& owner, int index) : owner(owner), index(index), tempGrid(NULL) {
     }
 };
 
@@ -368,11 +369,12 @@ CpuPme::CpuPme(int gridx, int gridy, int gridz, int numParticles, double alpha)
         ThreadData* data = new ThreadData(*this, i);
         threadData.push_back(data);
         pthread_create(&thread[i], NULL, threadBody, data);
+        data->tempGrid = (float*) fftwf_malloc(sizeof(float)*(gridx*gridy*gridz+3));
     }
 
     // Initialize FFTW.
     
-    realGrid = (float*) fftwf_malloc(sizeof(float)*gridx*gridy*gridz);
+    realGrid = threadData[0]->tempGrid;
     complexGrid = (fftwf_complex*) fftwf_malloc(sizeof(fftwf_complex)*gridx*gridy*(gridz/2+1));
     fftwf_plan_with_nthreads(numThreads);
     forwardFFT = fftwf_plan_dft_r2c_3d(gridx, gridy, gridz, realGrid, complexGrid, FFTW_MEASURE);
@@ -445,11 +447,16 @@ CpuPme::~CpuPme() {
     isFinished = true;
     pthread_cond_broadcast(&startCondition);
     pthread_mutex_unlock(&lock);
-    for (int i = 0; i < thread.size(); i++)
+    for (int i = 0; i < (int) thread.size(); i++)
         pthread_join(thread[i], NULL);
     pthread_mutex_destroy(&lock);
     pthread_cond_destroy(&startCondition);
     pthread_cond_destroy(&endCondition);
+    for (int i = 0; i < (int) threadData.size(); i++) {
+        if (threadData[i]->tempGrid != NULL)
+            fftwf_free(threadData[i]->tempGrid);
+        delete threadData[i];
+    }
 }
 
 #include <sys/time.h>
@@ -460,20 +467,33 @@ double diff(struct timeval t1, struct timeval t2) {
 void CpuPme::runThread(int index) {
     int particleStart = (index*numParticles)/numThreads;
     int particleEnd = ((index+1)*numParticles)/numThreads;
-    int gridStart = (index*gridx)/numThreads;
-    int gridEnd = ((index+1)*gridx)/numThreads;
+    int gridxStart = (index*gridx)/numThreads;
+    int gridxEnd = ((index+1)*gridx)/numThreads;
+    int gridSize = (gridx*gridy*gridz+3)/4;
+    int gridStart = 4*((index*gridSize)/numThreads);
+    int gridEnd = 4*(((index+1)*gridSize)/numThreads);
     while (!isFinished) {
+        threadWait();
+        spreadCharge(particleStart, particleEnd, posq, threadData[index]->tempGrid, gridx, gridy, gridz, numParticles, periodicBoxSize);
+        threadWait();
+        int numGrids = threadData.size();
+        for (int i = gridStart; i < gridEnd; i += 4) {
+            __m128 sum = _mm_load_ps(&realGrid[i]);
+            for (int j = 1; j < numGrids; j++)
+                sum = _mm_add_ps(sum, _mm_load_ps(&threadData[j]->tempGrid[i]));
+            _mm_store_ps(&realGrid[i], sum);
+        }
         threadWait();
         if (isFinished)
             break;
         if (includeEnergy) {
-            double threadEnergy = reciprocalEnergy(gridStart, gridEnd, &complexGrid[0], gridx, gridy, gridz, alpha, bsplineModuli, periodicBoxSize);
+            double threadEnergy = reciprocalEnergy(gridxStart, gridxEnd, complexGrid, gridx, gridy, gridz, alpha, bsplineModuli, periodicBoxSize);
             pthread_mutex_lock(&lock);
             energy += threadEnergy;
             pthread_mutex_unlock(&lock);
             threadWait();
         }
-        reciprocalConvolution(gridStart, gridEnd, complexGrid, gridx, gridy, gridz, alpha, bsplineModuli, periodicBoxSize);
+        reciprocalConvolution(gridxStart, gridxEnd, complexGrid, gridx, gridy, gridz, alpha, bsplineModuli, periodicBoxSize);
         threadWait();
         interpolateForces(particleStart, particleEnd, posq, force, realGrid, gridx, gridy, gridz, numParticles, periodicBoxSize);
     }
@@ -505,7 +525,8 @@ double CpuPme::computeForceAndEnergy(float* posq, float* force, Vec3 periodicBox
     energy = 0.0;
     struct timeval t1, t2, t3, t4, t5, t6, t7;
     gettimeofday(&t1, NULL);
-    spreadCharge(posq, realGrid, gridx, gridy, gridz, numParticles, periodicBoxSize);
+    advanceThreads(); // Signal threads to perform charge spreading.
+    advanceThreads(); // Signal threads to sum the charge grids.
     gettimeofday(&t2, NULL);
     fftwf_execute_dft_r2c(forwardFFT, realGrid, complexGrid);
     gettimeofday(&t3, NULL);