Using Software Agents in a Distributed Computing System for Procedural Planetoid Terrain Generation

Procedural terrain generation is becoming more widespread with numerous applications in different domains, like video games, movies, and simulations. As the need for content size, quantity, and fidelity rise, old methods become insufficient to perform the task. The planetary scale also imposes new limitations on data continuity and uniqueness. This paper proposes a novel cloud-based planetoid terrain generation system based on software agents. The performance is also improved through vertical parallelism using Morsels and horizontal parallelism using multithreading. The workers processing generation requests are spread per cloud computation instance, so they can be further scaled as the number of requests increases.