Accelerating the Lagrangian simulation of water ages on distributed, multi-GPU platforms: The importance of dynamic load balancing

Keyphrases

• Load Balancing 100%
• Water Age 100%
• Dynamic Load Balancing 100%
• Lagrangian Simulation 100%
• Multi-GPU Platform 100%
• GPU 71%
• Parallel Approach 28%
• Load Balancing Algorithm 28%
• Processing Element 28%
• Load Imbalance 28%
• Wall Clock Time 28%
• Particle Tracking 14%
• Particle Distribution 14%
• Parallel Scaling 14%
• Flow Path 14%
• Hillslope 14%
• Domain Decomposition 14%
• Massively Parallel Computation 14%
• Balance Ability 14%
• Load Distribution 14%
• Computational Work 14%
• Computational Load 14%
• Lagrangian Approach 14%
• Fluid Mixing 14%
• Water Sources 14%
• Lagrangian Particle Tracking 14%
• Tracking Approach 14%
• OpenMP 14%
• Distributed Parallelism 14%
• Parallel Scalability 14%
• Multi-GPU 14%
• MPI Parallelization 14%
• Study Load 14%
• Aging Simulation 14%
• Particle Tracking Simulation 14%
• Physically Based Modeling 14%
• Simulation Load 14%
• Load Ratio 14%
• Source Mixing 14%
• Scale Simulation 14%
• Scale Application 14%
• Water Age Distribution 14%

Computer Science

• Load Balancing 100%
• Dynamic Load Balancing 100%
• Graphics Processing Unit 100%
• Parallel Approach 22%
• Processing Element 22%
• Load Imbalance 22%
• Parallelism 11%
• Parallelization 11%
• Domain Decomposition 11%
• Lagrangian Approach 11%
• Massively Parallel Computing 11%