A new Dune grid for scalable dynamic adaptivity based on the p4est software library

2025-07-15

AI Summary

Existing grid modules (e.g., ALUGrid) in the Dune framework suffer from load imbalance and structural redundancy under extreme-scale MPI parallelism, limiting scalability in large-scale dynamic adaptive simulations. Method: This work introduces a novel, p4est-based grid interface for Dune, deeply integrating p4est’s lightweight forest-of-trees data structure with the Dune-Grid abstraction. We enhance p4est’s 2:1 balance constraint to improve load distribution and extend native support to 2D/3D multi-block adaptive meshes. Contribution/Results: The new interface achieves high parallel scalability and low memory overhead. On up to ten million CPU cores, it delivers 35–50% higher parallel efficiency than Dune-ALUGrid. It demonstrates superior strong and weak scaling—particularly in dynamic adaptive scenarios—while maintaining full compatibility with Dune’s grid interface specifications.

Technology Category

• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems
• Machine Learning: Hardware-aware ML
• Search and Optimization: Distributed Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

In this work we extend the Dune solver library with another grid interface to the open-source p4est software. While Dune already supports about a dozen different mesh implementations through its mesh interface Dune-Grid, we undertake this new coupling effort in order to inherit p4est's practically unlimited MPI scalability as well as its relatively thin data structures, and its native support for multi-block (forest) mesh topologies in both 2D and 3D. The presented implementation is compared to an existing implementation based on Dune-ALUGrid for a variety of challenging test examples in a parallel environment. The numerical experiments show that the implementation presented here is outperforming Dune-ALUGrid in terms of scalability. In addition, an alternative balancing strategy is presented to ensure 2:1 balancing across element faces showing improved performance compared to the existing p4est balance strategy in the numerical examples considered in this work.

Problem

• Extend Dune solver library with p4est for MPI scalability
• Compare new p4est-based implementation to Dune-ALUGrid
• Propose improved balancing strategy for element faces

Innovation

• Coupling Dune with p4est for scalability
• Thin data structures and multi-block support
• Improved balancing strategy for element faces