Scalable and Consistent Graph Neural Networks for Distributed Mesh-based Data-driven Modeling
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912055762616320 |
|---|---|
| author | Barwey, Shivam Balin, Riccardo Lusch, Bethany Patel, Saumil Balakrishnan, Ramesh Pal, Pinaki Maulik, Romit Vishwanath, Venkatram |
| author_facet | Barwey, Shivam Balin, Riccardo Lusch, Bethany Patel, Saumil Balakrishnan, Ramesh Pal, Pinaki Maulik, Romit Vishwanath, Venkatram |
| contents | This work develops a distributed graph neural network (GNN) methodology for mesh-based modeling applications using a consistent neural message passing layer. As the name implies, the focus is on enabling scalable operations that satisfy physical consistency via halo nodes at sub-graph boundaries. Here, consistency refers to the fact that a GNN trained and evaluated on one rank (one large graph) is arithmetically equivalent to evaluations on multiple ranks (a partitioned graph). This concept is demonstrated by interfacing GNNs with NekRS, a GPU-capable exascale CFD solver developed at Argonne National Laboratory. It is shown how the NekRS mesh partitioning can be linked to the distributed GNN training and inference routines, resulting in a scalable mesh-based data-driven modeling workflow. We study the impact of consistency on the scalability of mesh-based GNNs, demonstrating efficient scaling in consistent GNNs for up to O(1B) graph nodes on the Frontier exascale supercomputer. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_01657 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Scalable and Consistent Graph Neural Networks for Distributed Mesh-based Data-driven Modeling Barwey, Shivam Balin, Riccardo Lusch, Bethany Patel, Saumil Balakrishnan, Ramesh Pal, Pinaki Maulik, Romit Vishwanath, Venkatram Distributed, Parallel, and Cluster Computing Machine Learning Computational Physics This work develops a distributed graph neural network (GNN) methodology for mesh-based modeling applications using a consistent neural message passing layer. As the name implies, the focus is on enabling scalable operations that satisfy physical consistency via halo nodes at sub-graph boundaries. Here, consistency refers to the fact that a GNN trained and evaluated on one rank (one large graph) is arithmetically equivalent to evaluations on multiple ranks (a partitioned graph). This concept is demonstrated by interfacing GNNs with NekRS, a GPU-capable exascale CFD solver developed at Argonne National Laboratory. It is shown how the NekRS mesh partitioning can be linked to the distributed GNN training and inference routines, resulting in a scalable mesh-based data-driven modeling workflow. We study the impact of consistency on the scalability of mesh-based GNNs, demonstrating efficient scaling in consistent GNNs for up to O(1B) graph nodes on the Frontier exascale supercomputer. |
| title | Scalable and Consistent Graph Neural Networks for Distributed Mesh-based Data-driven Modeling |
| topic | Distributed, Parallel, and Cluster Computing Machine Learning Computational Physics |
| url | https://arxiv.org/abs/2410.01657 |