Score: 2

MSPT: Efficient Large-Scale Physical Modeling via Parallelized Multi-Scale Attention

Published: December 1, 2025 | arXiv ID: 2512.01738v1

By: Pedro M. P. Curvo, Jan-Willem van de Meent, Maksim Zhdanov

Potential Business Impact:

Solves hard science problems faster on computers.

Business Areas:
Natural Language Processing Artificial Intelligence, Data and Analytics, Software

A key scalability challenge in neural solvers for industrial-scale physics simulations is efficiently capturing both fine-grained local interactions and long-range global dependencies across millions of spatial elements. We introduce the Multi-Scale Patch Transformer (MSPT), an architecture that combines local point attention within patches with global attention to coarse patch-level representations. To partition the input domain into spatially-coherent patches, we employ ball trees, which handle irregular geometries efficiently. This dual-scale design enables MSPT to scale to millions of points on a single GPU. We validate our method on standard PDE benchmarks (elasticity, plasticity, fluid dynamics, porous flow) and large-scale aerodynamic datasets (ShapeNet-Car, Ahmed-ML), achieving state-of-the-art accuracy with substantially lower memory footprint and computational cost.

Country of Origin
🇳🇱 Netherlands

Repos / Data Links

Page Count
15 pages

Category
Computer Science:
Machine Learning (CS)