Hybrid Neural-MPM for Interactive Fluid Simulations
in Real-Time

*Equal Contribution Correspondence
1Institute of Information Science, Beijing Jiaotong University
2School of Computing Science, Simon Fraser University
3Meta Reality Labs

     

Introduction

We propose a neural physics system for real-time, interactive fluid simulations. Traditional physics-based methods, while accurate, are computationally intensive and suffer from latency issues. Recent machine‑learning methods reduce computational costs while preserving fidelity; yet most still fail to satisfy the latency constraints for real‑time use and lack support for interactive applications. To bridge this gap, we introduce a novel hybrid method that integrates numerical simulation, neural physics, and generative control. Our neural physics jointly pursues low-latency simulation and high physical fidelity by employing a fallback safeguard to classical numerical solvers. Furthermore, we develop a diffusion-based controller that is trained using a revserve modeling strategy to generate external dynamic force fields for fluid manipulation. Our system demonstrates robust performance across diverse 2D/3D scenarios, material types, and obstacle interactions, achieving real-time simulations at high frame rates (11~29% latency) while enabling fluid control guided by user-friendly freehand sketches. We present a significant step towards practical, controllable, and physically plausible fluid simulations for real-time interactive applications.

Figure: Method Overview. To achieve real‑time simulations, we cut latency by learning neural physics at a coarse spatiotemporal resolution, while safeguarding fidelity by automatically falling back to an MPM solver when complex fluid phenomena arise. For interactive control, we train a diffusion-based generative model that infers external force fields directly from user sketches.


Fluid Control Animations

Water 2D Arrow

Prediction

Sketch

Ground Truth

Water 2D Oval

Prediction

Sketch

Ground Truth

Sand 2D Arrow

Prediction

Sketch

Ground Truth

Sand 2D Oval

Prediction

Sketch

Ground Truth

Water 3D Arrow

Prediction

Sketch

Ground Truth

Water 3D Oval

Prediction

Sketch

Ground Truth

Sand 3D Arrow

Prediction

Sketch

Ground Truth

Sand 3D Oval

Prediction

Sketch

Ground Truth



The website is under construction. We will continue updating this project!


Citation

          
@article{xu2025hybrid,
  title={Hybrid Neural-MPM for Interactive Fluid Simulations in Real-Time},
  author={Xu, Jingxuan and Huang, Hong and Zou, Chuhang and Savva, Manolis and Wei, Yunchao and Chen, Wuyang},
  journal={arXiv preprint arXiv:2505.18926},
  year={2025}
}