Transolver-3: Scaling Up Transformer Solvers to Industrial-Scale Geometries

• URL: http://arxiv.org/abs/2602.04940v1
• Date: Wed, 04 Feb 2026 16:52:44 GMT
• Title: Transolver-3: Scaling Up Transformer Solvers to Industrial-Scale Geometries
• Authors: Hang Zhou, Haixu Wu, Haonan Shangguan, Yuezhou Ma, Huikun Weng, Jianmin Wang, Mingsheng Long,
• Abstract summary: Transolver-3 is a new member of the Transolver family designed for high-fidelity physics simulations.<n>We show that Transolver-3 is capable of handling meshes with over 160 million cells, achieving impressive performance across three challenging simulation benchmarks.
• Score: 51.028432812178266
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Deep learning has emerged as a transformative tool for the neural surrogate modeling of partial differential equations (PDEs), known as neural PDE solvers. However, scaling these solvers to industrial-scale geometries with over $10^8$ cells remains a fundamental challenge due to the prohibitive memory complexity of processing high-resolution meshes. We present Transolver-3, a new member of the Transolver family as a highly scalable framework designed for high-fidelity physics simulations. To bridge the gap between limited GPU capacity and the resolution requirements of complex engineering tasks, we introduce two key architectural optimizations: faster slice and deslice by exploiting matrix multiplication associative property and geometry slice tiling to partition the computation of physical states. Combined with an amortized training strategy by learning on random subsets of original high-resolution meshes and a physical state caching technique during inference, Transolver-3 enables high-fidelity field prediction on industrial-scale meshes. Extensive experiments demonstrate that Transolver-3 is capable of handling meshes with over 160 million cells, achieving impressive performance across three challenging simulation benchmarks, including aircraft and automotive design tasks.

Related papers

• Tail-Aware Post-Training Quantization for 3D Geometry Models [58.79500829118265]
  Post-Training Quantization (PTQ) enables efficient inference without retraining.<n>PTQ fails to transfer effectively to 3D models due to intricate feature distributions and prohibitive calibration overhead.<n>We propose TAPTQ, a Tail-Aware Post-Training Quantization pipeline for 3D geometric learning.
  arXiv  Detail & Related papers  (2026-02-02T07:21:15Z)
• Efficient Generative Transformer Operators For Million-Point PDEs [12.324265832276538]
  ECHO is a transformer-operator framework for generating million-point PDE trajectories.<n>We demonstrate state-of-the-art performance on million-point simulations featuring complex, high-frequency dynamics, and long-term horizons.
  arXiv  Detail & Related papers  (2025-12-04T16:46:48Z)
• Multi-patch isogeometric neural solver for partial differential equations on computer-aided design domains [0.0]
  This work develops a computational framework that combines physics-informed neural networks with multi-patch isogeometric analysis.<n>The method utilizes patch-local neural networks that operate on the reference domain of isogeometric analysis.<n>The effectiveness of the suggested method is demonstrated on two non-trivial and practically relevant use-cases.
  arXiv  Detail & Related papers  (2025-09-29T19:57:54Z)
• AB-UPT: Scaling Neural CFD Surrogates for High-Fidelity Automotive Aerodynamics Simulations via Anchored-Branched Universal Physics Transformers [20.363139473435048]
  We introduce AB-UPT as a novel modeling scheme for building neural surrogates for CFD simulations.<n>We show that AB-UPT yields state-of-the-art predictive accuracy of surface and volume fields on automotive CFD simulations ranging from 33 thousand up to 150 million mesh cells.<n> Notably, the proposed models can be trained on a single GPU in less than a day and predict industry-standard surface and volume fields within seconds.
  arXiv  Detail & Related papers  (2025-02-13T17:58:07Z)
• Transolver++: An Accurate Neural Solver for PDEs on Million-Scale Geometries [67.63077028746191]
  Transolver++ is a highly parallel and efficient neural solver that can solve PDEs on million-scale geometries.<n>Transolver++ increases the single- GPU input capacity to million-scale points for the first time.<n>It achieves over 20% performance gain in million-scale high-fidelity industrial simulations.
  arXiv  Detail & Related papers  (2025-02-04T15:33:50Z)
• Transolver: A Fast Transformer Solver for PDEs on General Geometries [66.82060415622871]
  We present Transolver, which learns intrinsic physical states hidden behind discretized geometries. By calculating attention to physics-aware tokens encoded from slices, Transovler can effectively capture intricate physical correlations. Transolver achieves consistent state-of-the-art with 22% relative gain across six standard benchmarks and also excels in large-scale industrial simulations.
  arXiv  Detail & Related papers  (2024-02-04T06:37:38Z)
• Learning Controllable Adaptive Simulation for Multi-resolution Physics [86.8993558124143]
  We introduce Learning controllable Adaptive simulation for Multi-resolution Physics (LAMP) as the first full deep learning-based surrogate model. LAMP consists of a Graph Neural Network (GNN) for learning the forward evolution, and a GNN-based actor-critic for learning the policy of spatial refinement and coarsening. We demonstrate that our LAMP outperforms state-of-the-art deep learning surrogate models, and can adaptively trade-off computation to improve long-term prediction error.
  arXiv  Detail & Related papers  (2023-05-01T23:20:27Z)
• NeuralStagger: Accelerating Physics-constrained Neural PDE Solver with Spatial-temporal Decomposition [67.46012350241969]
  This paper proposes a general acceleration methodology called NeuralStagger. It decomposing the original learning tasks into several coarser-resolution subtasks. We demonstrate the successful application of NeuralStagger on 2D and 3D fluid dynamics simulations.
  arXiv  Detail & Related papers  (2023-02-20T19:36:52Z)
• PhySRNet: Physics informed super-resolution network for application in computational solid mechanics [0.0]
  This work aims at developing a physics-informed deep learning based super-resolution framework (PhySRNet) It enables reconstruction of high-resolution deformation fields from their low-resolution counterparts without requiring high-resolution labeled data.
  arXiv  Detail & Related papers  (2022-06-30T17:51:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.