Work

On the integration of geometry agnostic variational-autoencoders into large-scale SVD based models

Year: 2025

Type: article

Abstract: This manuscript introduces GAVI (Geometry Agnostic Variational-autoencoder Integration), a scalable technique for model order reduction of terabyte-level state spaces. Such methodology is agnostic to the spatial location of the data points, hence, it can handle any complex geometry regardless of the grid type in which it is represented. The dimensionality reduction of GAVI is performed on two separate steps. First, a parallel QR factorization is applied to the streamwise velocity components, then a variational autoencoder learns a latent representation of the reduced R matrix. The final step can be done at laptop level regardless of the size of the original data. The methodology is used to compress 4000 snapshots of the streamwise velocity component of the flow around a circular cylinder at $Re_D=U_{\infty}D/\nu=10^4$. The data, which is represented on a 72 million points unstructured grid and has a size of 1.1 Tb, is compressed into 3 latent vectors that recover up to the 91.28% of the energy.

Source: Computers & Fluids

DOI: 10.1016/j.compfluid.2025.106797

Authors: Benet Eiximeno, Arnau Miró, J. Nathan Kutz, Ivette Rodriguez, Oriol Lehmkuhl

Institutions: Barcelona Supercomputing Center, Universitat Politècnica de Catalunya, University of Washington

Language: English

Topic: Model Reduction and Neural Networks

Subfield: Statistical and Nonlinear Physics

Field: Physics and Astronomy

Domain: Physical Sciences

Open Access status: hybrid

Estimated APC paid: $3,700