diff --git a/README.md b/README.md index 98dd9a8..8dbc918 100644 --- a/README.md +++ b/README.md @@ -1,10 +1,51 @@ **University of Pennsylvania, CIS 565: GPU Programming and Architecture, Project 1 - Flocking** -* (TODO) YOUR NAME HERE -* Tested on: (TODO) Windows 22, i7-2222 @ 2.22GHz 22GB, GTX 222 222MB (Moore 2222 Lab) +* Yash Vardhan +* Tested on: Windows 10 Pro, Intel i5-4200U (4) @ 2.600GHz 4GB, NVIDIA GeForce 840M 2048MB -### (TODO: Your README) +Boids Flocking using coherent uniform grid +=========================================== -Include screenshots, analysis, etc. (Remember, this is public, so don't put -anything here that you don't want to share with the world.) +![](images/boids.gif) + +Performance Analysis +==================== + +FPS vs Number of Boids(Visualized) +---------------------------------- + +![](images/visualboid.jpg) + +FPS vs Number of Boids(Non-Visualized) +-------------------------------------- + +![](images/nonvisualboid.jpg) + +FPS vs Block Size tested on 20,000 boids +---------------------------------------- + +![](images/blocksize.jpg) + +Q/A +=== + +**Q1 - For each implementation, how does changing the number of boids affect performance? Why do you think this is?** + +A1 - Naive neighbour search has an asymptotic drop with increase in the number of boids. Updating the velocities and position of boids gets moreand more difficult as the number of boid increases. Both scattered and coherent uniform grid flocking algorithms almost have a linear decrease with the number of increasing boids, but uniform grid flocking experiences a steeper gradient inintially which could be explained due to chasing the grid cell pointers. + +Also if the number of boids is 1000 or less, the naive flocking algorithm performs better than the uniform grid algorithms, this could be explained to the sparse grid the algorithms face, costing them unneccessary sorting computations. + +**Q2 - For each implementation, how does changing the block count and block size affect performance? Why do you think this is?** + +A2 - In naive neighbour implementation, increasing the block size barely affects the performance as it resembles a straight line with no slope. On the other hand the 2 grid algorithms show a slight increase in performance over block size of 128 and then come down to resemble a straight line, which helps explain why 128 block size was chosen as the default one for this project. In my opinion this is a purely experimental observation. + + +**Q3 - For the coherent uniform grid: did you experience any performance improvements with the more coherent uniform grid? Was this the outcome you expected? Why or why not?** + +A3 - Coherent uniform grid algorithm scales better than its counterparts with large number of boids, where as other 2 algorithms just fail. This could be linked to the lesser time expenditure in keeping up with the pointers(as in uniform grid) or having to compute all those other boid velocities in naive neighbour implementation. + + +**Q4 - Did changing cell width and checking 27 vs 8 neighboring cells affect performance? Why or why not?** + +A4 - Checking 27 neighbouring cells, in my case gave a slighly higher FPS than checking 8 neighbouring cells on larger number of boids. This might be explained through the grid cell width, which is twice the radius in 8 cell check case, so greater number of boids need to be covered. \ No newline at end of file diff --git a/images/blocksize.jpg b/images/blocksize.jpg new file mode 100644 index 0000000..49a32f5 Binary files /dev/null and b/images/blocksize.jpg differ diff --git a/images/boids.gif b/images/boids.gif new file mode 100644 index 0000000..0fc9331 Binary files /dev/null and b/images/boids.gif differ diff --git a/images/nonvisualboid.jpg b/images/nonvisualboid.jpg new file mode 100644 index 0000000..62e0975 Binary files /dev/null and b/images/nonvisualboid.jpg differ diff --git a/images/visualboid.jpg b/images/visualboid.jpg new file mode 100644 index 0000000..c3dcbae Binary files /dev/null and b/images/visualboid.jpg differ diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index fdd636d..dff0113 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -10,5 +10,5 @@ set(SOURCE_FILES cuda_add_library(src ${SOURCE_FILES} - OPTIONS -arch=sm_20 + OPTIONS -arch=sm_50 ) diff --git a/src/kernel.cu b/src/kernel.cu index aaf0fbf..26deaaa 100644 --- a/src/kernel.cu +++ b/src/kernel.cu @@ -5,6 +5,7 @@ #include #include "utilityCore.hpp" #include "kernel.h" +#include // LOOK-2.1 potentially useful for doing grid-based neighbor search #ifndef imax @@ -21,14 +22,14 @@ * Check for CUDA errors; print and exit if there was a problem. */ void checkCUDAError(const char *msg, int line = -1) { - cudaError_t err = cudaGetLastError(); - if (cudaSuccess != err) { - if (line >= 0) { - fprintf(stderr, "Line %d: ", line); - } - fprintf(stderr, "Cuda error: %s: %s.\n", msg, cudaGetErrorString(err)); - exit(EXIT_FAILURE); - } + cudaError_t err = cudaGetLastError(); + if (cudaSuccess != err) { + if (line >= 0) { + fprintf(stderr, "Line %d: ", line); + } + fprintf(stderr, "Cuda error: %s: %s.\n", msg, cudaGetErrorString(err)); + exit(EXIT_FAILURE); + } } @@ -76,7 +77,7 @@ glm::vec3 *dev_vel2; // For efficient sorting and the uniform grid. These should always be parallel. int *dev_particleArrayIndices; // What index in dev_pos and dev_velX represents this particle? int *dev_particleGridIndices; // What grid cell is this particle in? -// needed for use with thrust + // needed for use with thrust thrust::device_ptr dev_thrust_particleArrayIndices; thrust::device_ptr dev_thrust_particleGridIndices; @@ -85,6 +86,8 @@ int *dev_gridCellEndIndices; // to this cell? // TODO-2.3 - consider what additional buffers you might need to reshuffle // the position and velocity data to be coherent within cells. +glm::vec3 *coh_pos; +glm::vec3 *coh_vel; // LOOK-2.1 - Grid parameters based on simulation parameters. // These are automatically computed for you in Boids::initSimulation @@ -99,13 +102,13 @@ glm::vec3 gridMinimum; ******************/ __host__ __device__ unsigned int hash(unsigned int a) { - a = (a + 0x7ed55d16) + (a << 12); - a = (a ^ 0xc761c23c) ^ (a >> 19); - a = (a + 0x165667b1) + (a << 5); - a = (a + 0xd3a2646c) ^ (a << 9); - a = (a + 0xfd7046c5) + (a << 3); - a = (a ^ 0xb55a4f09) ^ (a >> 16); - return a; + a = (a + 0x7ed55d16) + (a << 12); + a = (a ^ 0xc761c23c) ^ (a >> 19); + a = (a + 0x165667b1) + (a << 5); + a = (a + 0xd3a2646c) ^ (a << 9); + a = (a + 0xfd7046c5) + (a << 3); + a = (a ^ 0xb55a4f09) ^ (a >> 16); + return a; } /** @@ -113,10 +116,10 @@ __host__ __device__ unsigned int hash(unsigned int a) { * Function for generating a random vec3. */ __host__ __device__ glm::vec3 generateRandomVec3(float time, int index) { - thrust::default_random_engine rng(hash((int)(index * time))); - thrust::uniform_real_distribution unitDistrib(-1, 1); + thrust::default_random_engine rng(hash((int)(index * time))); + thrust::uniform_real_distribution unitDistrib(-1, 1); - return glm::vec3((float)unitDistrib(rng), (float)unitDistrib(rng), (float)unitDistrib(rng)); + return glm::vec3((float)unitDistrib(rng), (float)unitDistrib(rng), (float)unitDistrib(rng)); } /** @@ -124,52 +127,74 @@ __host__ __device__ glm::vec3 generateRandomVec3(float time, int index) { * CUDA kernel for generating boids with a specified mass randomly around the star. */ __global__ void kernGenerateRandomPosArray(int time, int N, glm::vec3 * arr, float scale) { - int index = (blockIdx.x * blockDim.x) + threadIdx.x; - if (index < N) { - glm::vec3 rand = generateRandomVec3(time, index); - arr[index].x = scale * rand.x; - arr[index].y = scale * rand.y; - arr[index].z = scale * rand.z; - } + int index = (blockIdx.x * blockDim.x) + threadIdx.x; + if (index < N) { + glm::vec3 rand = generateRandomVec3(time, index); + arr[index].x = scale * rand.x; + arr[index].y = scale * rand.y; + arr[index].z = scale * rand.z; + } } /** * Initialize memory, update some globals */ void Boids::initSimulation(int N) { - numObjects = N; - dim3 fullBlocksPerGrid((N + blockSize - 1) / blockSize); - - // LOOK-1.2 - This is basic CUDA memory management and error checking. - // Don't forget to cudaFree in Boids::endSimulation. - cudaMalloc((void**)&dev_pos, N * sizeof(glm::vec3)); - checkCUDAErrorWithLine("cudaMalloc dev_pos failed!"); - - cudaMalloc((void**)&dev_vel1, N * sizeof(glm::vec3)); - checkCUDAErrorWithLine("cudaMalloc dev_vel1 failed!"); - - cudaMalloc((void**)&dev_vel2, N * sizeof(glm::vec3)); - checkCUDAErrorWithLine("cudaMalloc dev_vel2 failed!"); - - // LOOK-1.2 - This is a typical CUDA kernel invocation. - kernGenerateRandomPosArray<<>>(1, numObjects, - dev_pos, scene_scale); - checkCUDAErrorWithLine("kernGenerateRandomPosArray failed!"); - - // LOOK-2.1 computing grid params - gridCellWidth = 2.0f * std::max(std::max(rule1Distance, rule2Distance), rule3Distance); - int halfSideCount = (int)(scene_scale / gridCellWidth) + 1; - gridSideCount = 2 * halfSideCount; - - gridCellCount = gridSideCount * gridSideCount * gridSideCount; - gridInverseCellWidth = 1.0f / gridCellWidth; - float halfGridWidth = gridCellWidth * halfSideCount; - gridMinimum.x -= halfGridWidth; - gridMinimum.y -= halfGridWidth; - gridMinimum.z -= halfGridWidth; - - // TODO-2.1 TODO-2.3 - Allocate additional buffers here. - cudaThreadSynchronize(); + numObjects = N; + dim3 fullBlocksPerGrid((N + blockSize - 1) / blockSize); + + // LOOK-1.2 - This is basic CUDA memory management and error checking. + // Don't forget to cudaFree in Boids::endSimulation. + cudaMalloc((void**)&dev_pos, N * sizeof(glm::vec3)); + checkCUDAErrorWithLine("cudaMalloc dev_pos failed!"); + + cudaMalloc((void**)&dev_vel1, N * sizeof(glm::vec3)); + checkCUDAErrorWithLine("cudaMalloc dev_vel1 failed!"); + + cudaMalloc((void**)&dev_vel2, N * sizeof(glm::vec3)); + checkCUDAErrorWithLine("cudaMalloc dev_vel2 failed!"); + + // LOOK-1.2 - This is a typical CUDA kernel invocation. + kernGenerateRandomPosArray << > >(1, numObjects, + dev_pos, scene_scale); + checkCUDAErrorWithLine("kernGenerateRandomPosArray failed!"); + + // LOOK-2.1 computing grid params + gridCellWidth = 2.0f * std::max(std::max(rule1Distance, rule2Distance), rule3Distance); + int halfSideCount = (int)(scene_scale / gridCellWidth) + 1; + gridSideCount = 2 * halfSideCount; + + gridCellCount = gridSideCount * gridSideCount * gridSideCount; + gridInverseCellWidth = 1.0f / gridCellWidth; + float halfGridWidth = gridCellWidth * halfSideCount; + gridMinimum.x -= halfGridWidth; + gridMinimum.y -= halfGridWidth; + gridMinimum.z -= halfGridWidth; + + // TODO-2.1 TODO-2.3 - Allocate additional buffers here. + cudaMalloc((void**)&dev_particleArrayIndices, N * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_particleArrayIndices failed!"); + + cudaMalloc((void**)&dev_particleGridIndices, N * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_particleGridIndices failed!"); + + dev_thrust_particleArrayIndices = thrust::device_ptr(dev_particleArrayIndices); + + dev_thrust_particleGridIndices = thrust::device_ptr(dev_particleGridIndices); + + cudaMalloc((void**)&dev_gridCellStartIndices, gridCellCount * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_gridCellStartIndices failed!"); + + cudaMalloc((void**)&dev_gridCellEndIndices, gridCellCount * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_gridCellEndIndices failed!"); + + cudaMalloc((void**)&coh_pos, N * sizeof(glm::vec3)); + checkCUDAErrorWithLine("cudaMalloc coh_pos failed!"); + + cudaMalloc((void**)&coh_vel, N * sizeof(glm::vec3)); + checkCUDAErrorWithLine("cudaMalloc coh_vel failed!"); + + cudaThreadSynchronize(); } @@ -181,41 +206,41 @@ void Boids::initSimulation(int N) { * Copy the boid positions into the VBO so that they can be drawn by OpenGL. */ __global__ void kernCopyPositionsToVBO(int N, glm::vec3 *pos, float *vbo, float s_scale) { - int index = threadIdx.x + (blockIdx.x * blockDim.x); + int index = threadIdx.x + (blockIdx.x * blockDim.x); - float c_scale = -1.0f / s_scale; + float c_scale = -1.0f / s_scale; - if (index < N) { - vbo[4 * index + 0] = pos[index].x * c_scale; - vbo[4 * index + 1] = pos[index].y * c_scale; - vbo[4 * index + 2] = pos[index].z * c_scale; - vbo[4 * index + 3] = 1.0f; - } + if (index < N) { + vbo[4 * index + 0] = pos[index].x * c_scale; + vbo[4 * index + 1] = pos[index].y * c_scale; + vbo[4 * index + 2] = pos[index].z * c_scale; + vbo[4 * index + 3] = 1.0f; + } } __global__ void kernCopyVelocitiesToVBO(int N, glm::vec3 *vel, float *vbo, float s_scale) { - int index = threadIdx.x + (blockIdx.x * blockDim.x); - - if (index < N) { - vbo[4 * index + 0] = vel[index].x + 0.3f; - vbo[4 * index + 1] = vel[index].y + 0.3f; - vbo[4 * index + 2] = vel[index].z + 0.3f; - vbo[4 * index + 3] = 1.0f; - } + int index = threadIdx.x + (blockIdx.x * blockDim.x); + + if (index < N) { + vbo[4 * index + 0] = vel[index].x + 0.3f; + vbo[4 * index + 1] = vel[index].y + 0.3f; + vbo[4 * index + 2] = vel[index].z + 0.3f; + vbo[4 * index + 3] = 1.0f; + } } /** * Wrapper for call to the kernCopyboidsToVBO CUDA kernel. */ void Boids::copyBoidsToVBO(float *vbodptr_positions, float *vbodptr_velocities) { - dim3 fullBlocksPerGrid((numObjects + blockSize - 1) / blockSize); + dim3 fullBlocksPerGrid((numObjects + blockSize - 1) / blockSize); - kernCopyPositionsToVBO << > >(numObjects, dev_pos, vbodptr_positions, scene_scale); - kernCopyVelocitiesToVBO << > >(numObjects, dev_vel1, vbodptr_velocities, scene_scale); + kernCopyPositionsToVBO << > >(numObjects, dev_pos, vbodptr_positions, scene_scale); + kernCopyVelocitiesToVBO << > >(numObjects, dev_vel1, vbodptr_velocities, scene_scale); - checkCUDAErrorWithLine("copyBoidsToVBO failed!"); + checkCUDAErrorWithLine("copyBoidsToVBO failed!"); - cudaThreadSynchronize(); + cudaThreadSynchronize(); } @@ -230,10 +255,52 @@ void Boids::copyBoidsToVBO(float *vbodptr_positions, float *vbodptr_velocities) * in the `pos` and `vel` arrays. */ __device__ glm::vec3 computeVelocityChange(int N, int iSelf, const glm::vec3 *pos, const glm::vec3 *vel) { - // Rule 1: boids fly towards their local perceived center of mass, which excludes themselves - // Rule 2: boids try to stay a distance d away from each other - // Rule 3: boids try to match the speed of surrounding boids - return glm::vec3(0.0f, 0.0f, 0.0f); + glm::vec3 posboid = pos[iSelf]; + glm::vec3 veln = vel[iSelf]; + glm::vec3 perceived_center(0, 0, 0); + glm::vec3 c(0, 0, 0); + glm::vec3 perceived_velocity(0, 0, 0); + int n1 = 0, n3 = 0; + + // One loop to rule them all. + for (int b = 0; b < N; b++) + { + float distance = glm::distance(posboid, pos[b]); + if (b != iSelf) + { + // Rule 1: boids fly towards their local perceived center of mass, which excludes themselves + if (distance < rule1Distance) + { + perceived_center += pos[b]; + n1++; + } + // Rule 2: boids try to stay a distance d away from each other + if (distance < rule2Distance) + c -= (pos[b] - posboid); + // Rule 3: boids try to match the speed of surrounding boids + if (distance < rule3Distance) + { + perceived_velocity += vel[b]; + n3++; + } + } + } + //Adding rule velocities to initial velocity. + if (n1 > 0) + { + perceived_center /= n1; + veln += ((perceived_center - posboid) * rule1Scale); + } + + veln += (c * rule2Scale); + + if (n3 > 0) + { + perceived_velocity /= n3; + veln += (perceived_velocity * rule3Scale); + } + + return veln; } /** @@ -241,10 +308,19 @@ __device__ glm::vec3 computeVelocityChange(int N, int iSelf, const glm::vec3 *po * For each of the `N` bodies, update its position based on its current velocity. */ __global__ void kernUpdateVelocityBruteForce(int N, glm::vec3 *pos, - glm::vec3 *vel1, glm::vec3 *vel2) { - // Compute a new velocity based on pos and vel1 - // Clamp the speed - // Record the new velocity into vel2. Question: why NOT vel1? + glm::vec3 *vel1, glm::vec3 *vel2) { + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + // Compute a new velocity based on pos and vel1 + glm::vec3 veln = computeVelocityChange(N, index, pos, vel1); + // Clamp the speed + float speed = glm::length(veln); + if (speed > maxSpeed) + veln *= maxSpeed / speed; + // Record the new velocity into vel2. Question: why NOT vel1? + vel2[index] = veln; } /** @@ -252,24 +328,24 @@ __global__ void kernUpdateVelocityBruteForce(int N, glm::vec3 *pos, * For each of the `N` bodies, update its position based on its current velocity. */ __global__ void kernUpdatePos(int N, float dt, glm::vec3 *pos, glm::vec3 *vel) { - // Update position by velocity - int index = threadIdx.x + (blockIdx.x * blockDim.x); - if (index >= N) { - return; - } - glm::vec3 thisPos = pos[index]; - thisPos += vel[index] * dt; - - // Wrap the boids around so we don't lose them - thisPos.x = thisPos.x < -scene_scale ? scene_scale : thisPos.x; - thisPos.y = thisPos.y < -scene_scale ? scene_scale : thisPos.y; - thisPos.z = thisPos.z < -scene_scale ? scene_scale : thisPos.z; - - thisPos.x = thisPos.x > scene_scale ? -scene_scale : thisPos.x; - thisPos.y = thisPos.y > scene_scale ? -scene_scale : thisPos.y; - thisPos.z = thisPos.z > scene_scale ? -scene_scale : thisPos.z; - - pos[index] = thisPos; + // Update position by velocity + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + glm::vec3 thisPos = pos[index]; + thisPos += vel[index] * dt; + + // Wrap the boids around so we don't lose them + thisPos.x = thisPos.x < -scene_scale ? scene_scale : thisPos.x; + thisPos.y = thisPos.y < -scene_scale ? scene_scale : thisPos.y; + thisPos.z = thisPos.z < -scene_scale ? scene_scale : thisPos.z; + + thisPos.x = thisPos.x > scene_scale ? -scene_scale : thisPos.x; + thisPos.y = thisPos.y > scene_scale ? -scene_scale : thisPos.y; + thisPos.z = thisPos.z > scene_scale ? -scene_scale : thisPos.z; + + pos[index] = thisPos; } // LOOK-2.1 Consider this method of computing a 1D index from a 3D grid index. @@ -279,181 +355,423 @@ __global__ void kernUpdatePos(int N, float dt, glm::vec3 *pos, glm::vec3 *vel) { // for(y) // for(z)? Or some other order? __device__ int gridIndex3Dto1D(int x, int y, int z, int gridResolution) { - return x + y * gridResolution + z * gridResolution * gridResolution; + return x + y * gridResolution + z * gridResolution * gridResolution; } __global__ void kernComputeIndices(int N, int gridResolution, - glm::vec3 gridMin, float inverseCellWidth, - glm::vec3 *pos, int *indices, int *gridIndices) { - // TODO-2.1 - // - Label each boid with the index of its grid cell. - // - Set up a parallel array of integer indices as pointers to the actual - // boid data in pos and vel1/vel2 + glm::vec3 gridMin, float inverseCellWidth, + glm::vec3 *pos, int *indices, int *gridIndices) { + // TODO-2.1 + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + + // - Label each boid with the index of its grid cell. + glm::vec3 posboid = inverseCellWidth*(pos[index] - gridMin); + int x = (int)posboid.x; + int y = (int)posboid.y; + int z = (int)posboid.z; + int gindex = gridIndex3Dto1D(x, y, z, gridResolution); + gridIndices[index] = gindex; + + // - Set up a parallel array of integer indices as pointers to the actual + // boid data in pos and vel1/vel2 + indices[index] = index; } // LOOK-2.1 Consider how this could be useful for indicating that a cell // does not enclose any boids __global__ void kernResetIntBuffer(int N, int *intBuffer, int value) { - int index = (blockIdx.x * blockDim.x) + threadIdx.x; - if (index < N) { - intBuffer[index] = value; - } + int index = (blockIdx.x * blockDim.x) + threadIdx.x; + if (index < N) { + intBuffer[index] = value; + } } __global__ void kernIdentifyCellStartEnd(int N, int *particleGridIndices, - int *gridCellStartIndices, int *gridCellEndIndices) { - // TODO-2.1 - // Identify the start point of each cell in the gridIndices array. - // This is basically a parallel unrolling of a loop that goes - // "this index doesn't match the one before it, must be a new cell!" + int *gridCellStartIndices, int *gridCellEndIndices) { + // TODO-2.1 + // Identify the start point of each cell in the gridIndices array. + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + + // This is basically a parallel unrolling of a loop that goes + // "this index doesn't match the one before it, must be a new cell!" + if (index == 0) + gridCellStartIndices[particleGridIndices[index]] = 0; + else if (particleGridIndices[index] != particleGridIndices[index - 1]) + { + gridCellStartIndices[particleGridIndices[index]] = index; + gridCellEndIndices[particleGridIndices[index - 1]] = index - 1; + } + if (index == N - 1) + gridCellEndIndices[particleGridIndices[index]] = N - 1; } __global__ void kernUpdateVelNeighborSearchScattered( - int N, int gridResolution, glm::vec3 gridMin, - float inverseCellWidth, float cellWidth, - int *gridCellStartIndices, int *gridCellEndIndices, - int *particleArrayIndices, - glm::vec3 *pos, glm::vec3 *vel1, glm::vec3 *vel2) { - // TODO-2.1 - Update a boid's velocity using the uniform grid to reduce - // the number of boids that need to be checked. - // - Identify the grid cell that this particle is in - // - Identify which cells may contain neighbors. This isn't always 8. - // - For each cell, read the start/end indices in the boid pointer array. - // - Access each boid in the cell and compute velocity change from - // the boids rules, if this boid is within the neighborhood distance. - // - Clamp the speed change before putting the new speed in vel2 + int N, int gridResolution, glm::vec3 gridMin, + float inverseCellWidth, float cellWidth, + int *gridCellStartIndices, int *gridCellEndIndices, + int *particleArrayIndices, + glm::vec3 *pos, glm::vec3 *vel1, glm::vec3 *vel2) { + // TODO-2.1 - Update a boid's velocity using the uniform grid to reduce + // the number of boids that need to be checked. + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + glm::vec3 posboid = pos[index]; + glm::vec3 veln = vel1[index]; + glm::vec3 perceived_center(0, 0, 0); + glm::vec3 c(0, 0, 0); + glm::vec3 perceived_velocity(0, 0, 0); + int n1 = 0, n3 = 0; + + // - Identify the grid cell that this particle is in + glm::vec3 gridboid = inverseCellWidth*(pos[index] - gridMin); + + // - Identify which cells may contain neighbors. This isn't always 8. + // - For each cell, read the start/end indices in the boid pointer array. + for (int i = 0; i < 2; i++) + { + int x = imax((int)gridboid.x - i, 0); + for (int j = 0; j < 2; j++) + { + int y = imax((int)gridboid.y - j, 0); + for (int k = 0; k < 2; k++) + { + int z = imax((int)gridboid.z - k, 0); + int cellboid = gridIndex3Dto1D(x, y, z, gridResolution); + if (gridCellStartIndices[cellboid] >= 0) + { + int start = gridCellStartIndices[cellboid]; + int end = gridCellEndIndices[cellboid]; + for (int l = start; l <= end; l++) + { + int b = particleArrayIndices[l]; + float distance = glm::distance(posboid, pos[b]); + if (b != index) + { + // Rule 1: boids fly towards their local perceived center of mass, which excludes themselves + if (distance < rule1Distance) + { + perceived_center += pos[b]; + n1++; + } + // Rule 2: boids try to stay a distance d away from each other + if (distance < rule2Distance) + c -= (pos[b] - posboid); + // Rule 3: boids try to match the speed of surrounding boids + if (distance < rule3Distance) + { + perceived_velocity += vel1[b]; + n3++; + } + } + } + } + } + } + } + // - Access each boid in the cell and compute velocity change from + // the boids rules, if this boid is within the neighborhood distance. + if (n1 > 0) + { + perceived_center /= n1; + veln += ((perceived_center - posboid) * rule1Scale); + } + + veln += (c * rule2Scale); + + if (n3 > 0) + { + perceived_velocity /= n3; + veln += (perceived_velocity * rule3Scale); + } + // - Clamp the speed change before putting the new speed in vel2 + float speed = glm::length(veln); + if (speed > maxSpeed) + veln *= maxSpeed / speed; + vel2[index] = veln; } __global__ void kernUpdateVelNeighborSearchCoherent( - int N, int gridResolution, glm::vec3 gridMin, - float inverseCellWidth, float cellWidth, - int *gridCellStartIndices, int *gridCellEndIndices, - glm::vec3 *pos, glm::vec3 *vel1, glm::vec3 *vel2) { - // TODO-2.3 - This should be very similar to kernUpdateVelNeighborSearchScattered, - // except with one less level of indirection. - // This should expect gridCellStartIndices and gridCellEndIndices to refer - // directly to pos and vel1. - // - Identify the grid cell that this particle is in - // - Identify which cells may contain neighbors. This isn't always 8. - // - For each cell, read the start/end indices in the boid pointer array. - // DIFFERENCE: For best results, consider what order the cells should be - // checked in to maximize the memory benefits of reordering the boids data. - // - Access each boid in the cell and compute velocity change from - // the boids rules, if this boid is within the neighborhood distance. - // - Clamp the speed change before putting the new speed in vel2 + int N, int gridResolution, glm::vec3 gridMin, + float inverseCellWidth, float cellWidth, + int *gridCellStartIndices, int *gridCellEndIndices, + glm::vec3 *pos, glm::vec3 *vel1, glm::vec3 *vel2) { + // TODO-2.3 - This should be very similar to kernUpdateVelNeighborSearchScattered, + // except with one less level of indirection. + int index = threadIdx.x + (blockIdx.x * blockDim.x); + if (index >= N) { + return; + } + glm::vec3 posboid = pos[index]; + glm::vec3 veln = vel1[index]; + glm::vec3 perceived_center(0, 0, 0); + glm::vec3 c(0, 0, 0); + glm::vec3 perceived_velocity(0, 0, 0); + int n1 = 0, n3 = 0; + + // This should expect gridCellStartIndices and gridCellEndIndices to refer + // directly to pos and vel1. + + // - Identify the grid cell that this particle is in + glm::vec3 gridboid = inverseCellWidth*(pos[index] - gridMin); + + // - Identify which cells may contain neighbors. This isn't always 8. + // - For each cell, read the start/end indices in the boid pointer array. + // DIFFERENCE: For best results, consider what order the cells should be + // checked in to maximize the memory benefits of reordering the boids data. + for (int i = 0; i < 2; i++) + { + int x = imax((int)gridboid.x - i, 0); + for (int j = 0; j < 2; j++) + { + int y = imax((int)gridboid.y - j, 0); + for (int k = 0; k < 2; k++) + { + int z = imax((int)gridboid.z - k, 0); + int cellboid = gridIndex3Dto1D(x, y, z, gridResolution); + if (gridCellStartIndices[cellboid] >= 0) + { + int start = gridCellStartIndices[cellboid]; + int end = gridCellEndIndices[cellboid]; + for (int b = start; b <= end; b++) + { + float distance = glm::distance(posboid, pos[b]); + if (b != index) + { + // Rule 1: boids fly towards their local perceived center of mass, which excludes themselves + if (distance < rule1Distance) + { + perceived_center += pos[b]; + n1++; + } + // Rule 2: boids try to stay a distance d away from each other + if (distance < rule2Distance) + c -= (pos[b] - posboid); + // Rule 3: boids try to match the speed of surrounding boids + if (distance < rule3Distance) + { + perceived_velocity += vel1[b]; + n3++; + } + } + } + } + } + } + } + if (n1 > 0) + { + perceived_center /= n1; + veln += ((perceived_center - posboid) * rule1Scale); + } + + veln += (c * rule2Scale); + + if (n3 > 0) + { + perceived_velocity /= n3; + veln += (perceived_velocity * rule3Scale); + } + // - Clamp the speed change before putting the new speed in vel2 + float speed = glm::length(veln); + if (speed > maxSpeed) + veln *= maxSpeed / speed; + vel2[index] = veln; } /** * Step the entire N-body simulation by `dt` seconds. */ void Boids::stepSimulationNaive(float dt) { - // TODO-1.2 - use the kernels you wrote to step the simulation forward in time. - // TODO-1.2 ping-pong the velocity buffers + dim3 fullBlocksPerGrid((numObjects + blockSize - 1) / blockSize); + // TODO-1.2 - use the kernels you wrote to step the simulation forward in time. + kernUpdateVelocityBruteForce << > >(numObjects, dev_pos, dev_vel1, dev_vel2); + kernUpdatePos << > >(numObjects, dt, dev_pos, dev_vel2); + // TODO-1.2 ping-pong the velocity buffers + glm::vec3 *tmp = dev_vel1; + dev_vel1 = dev_vel2; + dev_vel2 = tmp; } void Boids::stepSimulationScatteredGrid(float dt) { - // TODO-2.1 - // Uniform Grid Neighbor search using Thrust sort. - // In Parallel: - // - label each particle with its array index as well as its grid index. - // Use 2x width grids. - // - Unstable key sort using Thrust. A stable sort isn't necessary, but you - // are welcome to do a performance comparison. - // - Naively unroll the loop for finding the start and end indices of each - // cell's data pointers in the array of boid indices - // - Perform velocity updates using neighbor search - // - Update positions - // - Ping-pong buffers as needed + // TODO-2.1 + // Uniform Grid Neighbor search using Thrust sort. + // In Parallel: + // - label each particle with its array index as well as its grid index. + dim3 fullBlocksPerGrid((numObjects + blockSize - 1) / blockSize); + kernResetIntBuffer << > >(numObjects, dev_gridCellStartIndices, -1); + kernResetIntBuffer << > >(numObjects, dev_gridCellEndIndices, -1); + + kernComputeIndices << > >(numObjects, gridSideCount, gridMinimum, gridInverseCellWidth, dev_pos, dev_particleArrayIndices, dev_particleGridIndices); + + // Use 2x width grids. + // - Unstable key sort using Thrust. A stable sort isn't necessary, but you + // are welcome to do a performance comparison. + dev_thrust_particleArrayIndices = thrust::device_ptr(dev_particleArrayIndices); + dev_thrust_particleGridIndices = thrust::device_ptr(dev_particleGridIndices); + + thrust::sort_by_key(dev_thrust_particleGridIndices, dev_thrust_particleGridIndices + numObjects, dev_thrust_particleArrayIndices); + + // - Naively unroll the loop for finding the start and end indices of each + // cell's data pointers in the array of boid indices + kernIdentifyCellStartEnd << > >(numObjects, dev_particleGridIndices, dev_gridCellStartIndices, dev_gridCellEndIndices); + + // - Perform velocity updates using neighbor search + kernUpdateVelNeighborSearchScattered << > >(numObjects, gridSideCount, gridMinimum, + gridInverseCellWidth, gridCellWidth, + dev_gridCellStartIndices, dev_gridCellEndIndices, + dev_particleArrayIndices, dev_pos, dev_vel1, dev_vel2); + + // - Update positions + kernUpdatePos << > >(numObjects, dt, dev_pos, dev_vel2); + + // - Ping-pong buffers as needed + glm::vec3 *tmp = dev_vel1; + dev_vel1 = dev_vel2; + dev_vel2 = tmp; +} + +__global__ void kernUpdatePosVel(int N, glm::vec3 *pos, glm::vec3 *vel1, + glm::vec3 *coh_pos, glm::vec3 *coh_vel, int *particleArrayIndices) { + int index = (blockIdx.x * blockDim.x) + threadIdx.x; + if (index >= N) { + return; + } + + int index2 = particleArrayIndices[index]; + + coh_pos[index] = pos[index2]; + coh_vel[index] = vel1[index2]; } + void Boids::stepSimulationCoherentGrid(float dt) { - // TODO-2.3 - start by copying Boids::stepSimulationNaiveGrid - // Uniform Grid Neighbor search using Thrust sort on cell-coherent data. - // In Parallel: - // - Label each particle with its array index as well as its grid index. - // Use 2x width grids - // - Unstable key sort using Thrust. A stable sort isn't necessary, but you - // are welcome to do a performance comparison. - // - Naively unroll the loop for finding the start and end indices of each - // cell's data pointers in the array of boid indices - // - BIG DIFFERENCE: use the rearranged array index buffer to reshuffle all - // the particle data in the simulation array. - // CONSIDER WHAT ADDITIONAL BUFFERS YOU NEED - // - Perform velocity updates using neighbor search - // - Update positions - // - Ping-pong buffers as needed. THIS MAY BE DIFFERENT FROM BEFORE. + // TODO-2.3 - start by copying Boids::stepSimulationNaiveGrid + // Uniform Grid Neighbor search using Thrust sort on cell-coherent data. + // In Parallel: + // - Label each particle with its array index as well as its grid index. + dim3 fullBlocksPerGrid((numObjects + blockSize - 1) / blockSize); + kernResetIntBuffer << > >(numObjects, dev_gridCellStartIndices, -1); + kernResetIntBuffer << > >(numObjects, dev_gridCellEndIndices, -1); + + kernComputeIndices << > >(numObjects, gridSideCount, gridMinimum, gridInverseCellWidth, dev_pos, dev_particleArrayIndices, dev_particleGridIndices); + + // Use 2x width grids + // - Unstable key sort using Thrust. A stable sort isn't necessary, but you + // are welcome to do a performance comparison. + dev_thrust_particleArrayIndices = thrust::device_ptr(dev_particleArrayIndices); + dev_thrust_particleGridIndices = thrust::device_ptr(dev_particleGridIndices); + + thrust::sort_by_key(dev_thrust_particleGridIndices, dev_thrust_particleGridIndices + numObjects, dev_thrust_particleArrayIndices); + + // - Naively unroll the loop for finding the start and end indices of each + // cell's data pointers in the array of boid indices + kernIdentifyCellStartEnd <<>>(numObjects, dev_particleGridIndices, dev_gridCellStartIndices, dev_gridCellEndIndices); + + // - BIG DIFFERENCE: use the rearranged array index buffer to reshuffle all + // the particle data in the simulation array. + // CONSIDER WHAT ADDITIONAL BUFFERS YOU NEED + kernUpdatePosVel <<>>(numObjects, dev_pos, dev_vel1, coh_pos, coh_vel, dev_particleArrayIndices); + + // - Perform velocity updates using neighbor search + kernUpdateVelNeighborSearchCoherent <<>>(numObjects, gridSideCount, gridMinimum, gridInverseCellWidth, gridCellWidth, dev_gridCellStartIndices, dev_gridCellEndIndices, coh_pos, coh_vel, dev_vel2); + + // - Update positions + kernUpdatePos <<>>(numObjects, dt, coh_pos, dev_vel2); + + // - Ping-pong buffers as needed. THIS MAY BE DIFFERENT FROM BEFORE. + glm::vec3 *tmp = dev_vel1; + dev_vel1 = dev_vel2; + dev_vel2 = tmp; + + tmp = dev_pos; + dev_pos = coh_pos; + coh_pos = tmp; } void Boids::endSimulation() { - cudaFree(dev_vel1); - cudaFree(dev_vel2); - cudaFree(dev_pos); - - // TODO-2.1 TODO-2.3 - Free any additional buffers here. + cudaFree(dev_vel1); + cudaFree(dev_vel2); + cudaFree(dev_pos); + + // TODO-2.1 TODO-2.3 - Free any additional buffers here. + cudaFree(dev_particleArrayIndices); + cudaFree(dev_particleGridIndices); + cudaFree(dev_gridCellStartIndices); + cudaFree(dev_gridCellEndIndices); + cudaFree(coh_pos); + cudaFree(coh_vel); } void Boids::unitTest() { - // LOOK-1.2 Feel free to write additional tests here. - - // test unstable sort - int *dev_intKeys; - int *dev_intValues; - int N = 10; - - int *intKeys = new int[N]; - int *intValues = new int[N]; - - intKeys[0] = 0; intValues[0] = 0; - intKeys[1] = 1; intValues[1] = 1; - intKeys[2] = 0; intValues[2] = 2; - intKeys[3] = 3; intValues[3] = 3; - intKeys[4] = 0; intValues[4] = 4; - intKeys[5] = 2; intValues[5] = 5; - intKeys[6] = 2; intValues[6] = 6; - intKeys[7] = 0; intValues[7] = 7; - intKeys[8] = 5; intValues[8] = 8; - intKeys[9] = 6; intValues[9] = 9; - - cudaMalloc((void**)&dev_intKeys, N * sizeof(int)); - checkCUDAErrorWithLine("cudaMalloc dev_intKeys failed!"); - - cudaMalloc((void**)&dev_intValues, N * sizeof(int)); - checkCUDAErrorWithLine("cudaMalloc dev_intValues failed!"); - - dim3 fullBlocksPerGrid((N + blockSize - 1) / blockSize); - - std::cout << "before unstable sort: " << std::endl; - for (int i = 0; i < N; i++) { - std::cout << " key: " << intKeys[i]; - std::cout << " value: " << intValues[i] << std::endl; - } - - // How to copy data to the GPU - cudaMemcpy(dev_intKeys, intKeys, sizeof(int) * N, cudaMemcpyHostToDevice); - cudaMemcpy(dev_intValues, intValues, sizeof(int) * N, cudaMemcpyHostToDevice); - - // Wrap device vectors in thrust iterators for use with thrust. - thrust::device_ptr dev_thrust_keys(dev_intKeys); - thrust::device_ptr dev_thrust_values(dev_intValues); - // LOOK-2.1 Example for using thrust::sort_by_key - thrust::sort_by_key(dev_thrust_keys, dev_thrust_keys + N, dev_thrust_values); - - // How to copy data back to the CPU side from the GPU - cudaMemcpy(intKeys, dev_intKeys, sizeof(int) * N, cudaMemcpyDeviceToHost); - cudaMemcpy(intValues, dev_intValues, sizeof(int) * N, cudaMemcpyDeviceToHost); - checkCUDAErrorWithLine("memcpy back failed!"); - - std::cout << "after unstable sort: " << std::endl; - for (int i = 0; i < N; i++) { - std::cout << " key: " << intKeys[i]; - std::cout << " value: " << intValues[i] << std::endl; - } - - // cleanup - delete[] intKeys; - delete[] intValues; - cudaFree(dev_intKeys); - cudaFree(dev_intValues); - checkCUDAErrorWithLine("cudaFree failed!"); - return; + // LOOK-1.2 Feel free to write additional tests here. + + // test unstable sort + int *dev_intKeys; + int *dev_intValues; + int N = 10; + + int *intKeys = new int[N]; + int *intValues = new int[N]; + + intKeys[0] = 0; intValues[0] = 0; + intKeys[1] = 1; intValues[1] = 1; + intKeys[2] = 0; intValues[2] = 2; + intKeys[3] = 3; intValues[3] = 3; + intKeys[4] = 0; intValues[4] = 4; + intKeys[5] = 2; intValues[5] = 5; + intKeys[6] = 2; intValues[6] = 6; + intKeys[7] = 0; intValues[7] = 7; + intKeys[8] = 5; intValues[8] = 8; + intKeys[9] = 6; intValues[9] = 9; + + cudaMalloc((void**)&dev_intKeys, N * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_intKeys failed!"); + + cudaMalloc((void**)&dev_intValues, N * sizeof(int)); + checkCUDAErrorWithLine("cudaMalloc dev_intValues failed!"); + + dim3 fullBlocksPerGrid((N + blockSize - 1) / blockSize); + + std::cout << "before unstable sort: " << std::endl; + for (int i = 0; i < N; i++) { + std::cout << " key: " << intKeys[i]; + std::cout << " value: " << intValues[i] << std::endl; + } + + // How to copy data to the GPU + cudaMemcpy(dev_intKeys, intKeys, sizeof(int) * N, cudaMemcpyHostToDevice); + cudaMemcpy(dev_intValues, intValues, sizeof(int) * N, cudaMemcpyHostToDevice); + + // Wrap device vectors in thrust iterators for use with thrust. + thrust::device_ptr dev_thrust_keys(dev_intKeys); + thrust::device_ptr dev_thrust_values(dev_intValues); + // LOOK-2.1 Example for using thrust::sort_by_key + thrust::sort_by_key(dev_thrust_keys, dev_thrust_keys + N, dev_thrust_values); + + // How to copy data back to the CPU side from the GPU + cudaMemcpy(intKeys, dev_intKeys, sizeof(int) * N, cudaMemcpyDeviceToHost); + cudaMemcpy(intValues, dev_intValues, sizeof(int) * N, cudaMemcpyDeviceToHost); + checkCUDAErrorWithLine("memcpy back failed!"); + + std::cout << "after unstable sort: " << std::endl; + for (int i = 0; i < N; i++) { + std::cout << " key: " << intKeys[i]; + std::cout << " value: " << intValues[i] << std::endl; + } + + // cleanup + delete[] intKeys; + delete[] intValues; + cudaFree(dev_intKeys); + cudaFree(dev_intValues); + checkCUDAErrorWithLine("cudaFree failed!"); + return; } diff --git a/src/main.cpp b/src/main.cpp index a29471d..a7c2c99 100644 --- a/src/main.cpp +++ b/src/main.cpp @@ -14,8 +14,8 @@ // LOOK-2.1 LOOK-2.3 - toggles for UNIFORM_GRID and COHERENT_GRID #define VISUALIZE 1 -#define UNIFORM_GRID 0 -#define COHERENT_GRID 0 +#define UNIFORM_GRID 1 +#define COHERENT_GRID 1 // LOOK-1.2 - change this to adjust particle count in the simulation const int N_FOR_VIS = 5000;