GPU environmental delegation of agent perceptions for MABS

摘要

Considering the digital simulation of complex systems, General-Purpose Computing on Graphics Processing Units (GPGPU) is a relevant approach for addressing scalability issues. However, GPU programming is a very specific approach that strongly limits both the accessibility and the re-usability of the frameworks developed using GPGPU. This paper presents our approach for the integration of GPU modules in a Multi-Agent Based Simulation (MABS) platform. Especially, this paper shows how we keep the programming accessibility of the platform while gaining advantages of the GPU power. The paper also presents how this approach could be generalized and proposes a MABS design guideline dedicated to the GPU context.