Fluid Simulation Based on Material Point Method with Neural Network

A new probabilistic approach on Material Point Method (MPM) based fluid simulation is introduced in this research. This paper discusses the implementation of neural network to allow more opportunities in optimizing fluid simulation. The implication on its visual and performance will also be discussed, along with the potential of further optimization in future research. The result was tested both on CPU and GPU. The implementation on GPU was done using Compute Unified Device Architecture (CUDA). The neural network was trained using FANN library, while the inference process was implemented using FANN (on CPU) and matrix multiplication (on CPU and GPU). From this research, neural network implementation shows better performance using FANN library while still giving acceptable visualization.

Paper

Full text

PDF

Fluid Simulation Based on Material Point Method with Neural Network

Semantic Scholar · Engineering · 2019

Abstract

A new probabilistic approach on Material Point Method (MPM) based fluid simulation is introduced in this research. This paper discusses the implementation of neural network to allow more opportunities in optimizing fluid simulation. The implication on its visual and performance will also be discussed, along with the potential of further optimization in future research. The result was tested both on CPU and GPU. The implementation on GPU was done using Compute Unified Device Architecture (CUDA). The neural network was trained using FANN library, while the inference process was implemented using FANN (on CPU) and matrix multiplication (on CPU and GPU). From this research, neural network implementation shows better performance using FANN library while still giving acceptable visualization.

Similar papers

© 2026 NYSGPT2525 LLC