G-ML-Octree: An Update-Efficient Index Structure for Simulating 3D Moving Objects Across GPUs

摘要: In real simulation applications, simulations often involve large volumes of three-dimensinal (3D) moving objects. With the rapid growth scale simulation-problem domains, it has become a key requirement to efficiently manage massive 3D Conventional indexing approaches for managing objects during generally suffer from excessive update costs. Aiming this problem, paper first proposes an update-efficient structure by fusing loose Octree and one update-memo structure, namely ML-Octree. ML-Octree significantly reduces costs involving Towards providing more efficient approach, explored feasibility paralleling employing Graphic Processing Unit (GPU). A load-balancing scheme is used further improve performance GPU-aided Finally, distributed proposed large-scale simulations. The experimental results indicate that (1) can acquire update-performance gain order magnitude similar Octree, (2) accelerate 5.07 $\times$ faster than parallel with 8 CPU threads on average, (3) load-balance 2.3 (4) support