Today's graphics processing units (GPU) have tremendous resources when it comes to raw computing power. The simulation of large groups of agents in transport simulation has a huge demand of computation time. Therefore it seems reasonable to try to harvest this computing power for traffic simulation. Unfortunately simulating a network of traffic is inherently connected with random memory access. This is not a domain that the SIMD (single instruction, multiple data) architecture of GPUs is known to work well with. In this paper the authors will try to achieve a speedup by computing multi-agent traffic simulations on the graphics device using NVIDIA's CUDA framework. After presenting the related work in section 2 the authors will recall some fundamental facts about the hardware used as well as the queue traffic simulation that will be implemented in section 3. In the fourth section the various different data structures and variations of the queue simulations algorithm benchmarked will be described. The results of the benchmarking will be presented in section 5 and a conclusion will be given in the last section of this paper.
Multi-agent traffic simulation with CUDA
Multi-Agenten-Verkehrssimulation mit CUDA
2009
9 Seiten, 8 Bilder, 2 Tabellen, 24 Quellen
Conference paper
Storage medium
English
Single-lane traffic simulation with multi-agent system
IEEE | 2005
|Single-Lane Traffic Simulation with Multi-Agent System
British Library Conference Proceedings | 2005
|Multi-Agent Based Air Traffic Simulation System
AIAA | 2015
|CUDA-enabled Particle-Based 3D Fluid Haptic Simulation
IEEE | 2011
|Multi-Agent Based Air Traffic Simulation System
British Library Conference Proceedings | 2015
|