Parallel Isotropic and Anisotropic Meshing

Simulating complex multi-scale phenomena requires meshes that are not only massive but mathematically optimal. While anisotropic mesh adaptation has matured into a rigorous science governed by Riemannian metric fields and error estimates, generating these meshes at the Exascale poses a formidable system challenge.This Research Topic bridges the divide between mathematical mesh theory and High-Performance Computing (HPC). We explore how anisotropic meshing can scale to billions of elements without compromising geometric fidelity. We specifically highlight the emerging role of Artificial Intelligence (AI) as a catalyst—from using Deep Learning to accelerate metric field construction to using Reinforcement Learning for optimizing dynamic load balancing.This research topic welcomes original research articles, reviews and methods articles covering, but not restricted to, the following aspects:· Parallel Algorithms & Scalability: Distributed generation and partitioning for heterogeneous CPU/GPU architectures.· Rigorous Anisotropic Adaptation: Metric-field methods and adjoint-based error estimation.· AI-Enhanced Execution: ML models for workload prediction and minimizing synchronization latency.· Space-Time & Moving Meshes: 4D strategies for CFD and climate science.· Geometric AI: Neural representations (SDFs, GNNs) for surface recovery and metric prediction. Topic editors: Nikos P Chrisochoides, Rao Garimella Source: https://www.frontiersin.org/research-topics/78292/parallel-isotropic-and-anisotropic-meshing