A Dynamic Geometry-Based Shared Space Interaction Framework for Parallel Scientific Applications

Abstract

While large-scale parallel/distributed simulations are rapidly becoming critical research modalities in academia and industry, their efficient and scalable parallel implementations present many challenges. A key challenge is the dynamic and complex communication/coordination patterns required by these applications, which depend on states of the phenomenon being modeled and are determined by the specific numerical formulation, the domain decomposition and/or sub-domain refinement algorithms used, and are known only at runtime. In this paper, we present a dynamic geometry-based shared-space interaction framework for scientific applications. The framework provides the flexibility of shared-space coordination models while enabling scalable implementations. The design, prototype implementation and experimental evaluation using an adaptive multi-block oil reservoir simulation are presented.