Massive simulation using GPU of a distributed behavioral model of a flock with obstacle avoidance

In this work, we present a massive simulation of a behavioral model using graphics hardware. In particular we took a well established model to simulate complex flocks and we focused our attention on its implementation using techniques to manage efficiently large amount of data. Using the recent programmability of GPUs and recent extensions of computer graphics programming, we implemented on the graphics hardware the model capable of managing a large aggregate motion of birds in a virtual environment as well as to avoiding both static and dynamic obstacles. We demonstrated the effectiveness of our GPU implementation when compared with CPU according to recent trends that show graphics hardware capable of also working outside of its natural application field.