FLocking Boids Simulation : Cuda / C++
Github Link : https://github.com/rtarun9/Cuda-Flocking-Boids
Showcase Video
[Note : The starting code / project template is adapted from : https://github.com/CIS565-Fall-2022/Project1-CUDA-Flocking]
Techniques Implemented
- (i) Brute Force approach : Each boid iterates over all other boids present in the scene.
- (ii) Uniform grid: Each boid iterates over boids only in neighboring spatial grids.
- (iii) Coherent Grid: Similar to the uniform grid, but memory access is more linear (requires fewer memory indirections, which leads to massive performance gains!).
Performance Analysis
