Adaptive Neural Network-Based Approximation to Accelerate Eulerian Fluid Simulation

Abstract

The Eulerian fluid simulation is an important HPC application. The neural network has been applied to accelerate it. The current methods that accelerate the fluid simulation with neural networks lack flexibility and generalization. In this paper, we tackle the above limitation and aim to enhance the applicability of neural networks in the Eulerian fluid simulation. We introduce Smartfluidnet, a framework that automates model generation and application. Given an existing neural network as input, Smartfluidnet generates multiple neural networks before the simulation to meet the execution time and simulation quality requirement. During the simulation, Smartfluidnet dynamically switches the neural networks to make the best efforts to reach the user requirement on simulation quality. Evaluating with 20,480 input problems, we show that Smartfluidnet achieves 1.46x and 590x speedup comparing with a state-of-the-art neural network model and the original fluid simulation respectively on an NVIDIA Titan X Pascal GPU, while providing better simulation quality than the state-of-the-art model.

References in corpus (14)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1
• Pruning Convolutional Neural Networks for Resource Efficient Inference
• A Survey of Model Compression and Acceleration for Deep Neural Networks
• Convolutional neural networks with low-rank regularization
• Training and Inference with Integers in Deep Neural Networks
• Accelerating Eulerian Fluid Simulation With Convolutional Networks
• Data-Free Knowledge Distillation for Deep Neural Networks
• Fast machine-learning online optimization of ultra-cold-atom experiments
• Deep Model Compression: Distilling Knowledge from Noisy Teachers
• Learning Physical Intuition of Block Towers by Example
• A GPU-accelerated package for simulation of flow in nanoporous source rocks with many-body dissipative particle dynamics
• CosmoFlow: Using Deep Learning to Learn the Universe at Scale
• Ease.ml: Towards Multi-tenant Resource Sharing for Machine Learning Workloads

Cited by in corpus (6)

• CFDNet: a deep learning-based accelerator for fluid simulations
• Mosaic Flows: A Transferable Deep Learning Framework for Solving PDEs on Unseen Domains
• Smart-PGSim: Using Neural Network to Accelerate AC-OPF Power Grid Simulation
• HALO 1.0: A Hardware-agnostic Accelerator Orchestration Framework for Enabling Hardware-agnostic Programming with True Performance Portability for Heterogeneous HPC
• Adaptive Fluid Cohomology on Surfaces
• Bridging the Gap between Deep Learning and Frustrated Quantum Spin System for Extreme-scale Simulations on New Generation of Sunway Supercomputer