Authors Naoki Morita, Shigeki Kaneko, Nozomi Magome, Naoto Mitsume, Kyohei Shintate
Publication date 7 July 2026
Tldr This paper proposes a novel method for distributed-memory parallel Galerkin model order reduction of large-scale problems using a metagraph to represent domain-decomposition and proper orthogonal decomposition (POD) computation subdomains, enabling decoupled parallelization of offline and online phases and supporting static load balancing through cost-based metanode weights.
This page was built for publication: Metagraph-Based Domain-Decomposed Galerkin Reduced-Order Model
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7283845)
