A GPU Parallel Scheme for Accelerating 2D and 3D Peridynamics Models

Xiaoming Wang,Qihang Wang,Boyang An,Qing He,Ping Wang, Jun Wu

Theoretical and Applied Fracture Mechanics（2022）

Southwest Jiaotong Univ | China Railway Chengdu Bur Grp Co Ltd

Cited 14|Views15

Abstract

Peridynamics (PD) is prevailing in the numerical simulations of damage evolution, but with the cost of far more required computations than traditional methods. This paper proposes a massively parallel implementation scheme for PD simulations with a single-card Graphics Processing Unit (GPU) to reduce the computational cost. The GPU parallel scheme includes two-level parallel modes of particle-mapping and bond-mapping, realizing high parallelism. By reasonably setting the data structure and using the CUDA memory model, realize the efficient utilization of GPU memory resources. Three numerical experiments involving quasi-static and transient problems, 2D and 3D problems, and impact problems are performed. The results show that the proposed parallel scheme can greatly improve the computational efficiency of various PD models while ensuring accuracy. The bond-mapping unrolls the inner loop of the particle-mapping and makes full use of registers and shared memory resources for stronger performance. To further explore the fatigue behavior of rails with hole defects using the GPU parallel scheme, which is difficult to perform using the serial and OpenMP scheme. The results show that the proposed GPU parallel scheme can explore more complex structural fracture problems by greatly reducing the computational cost.

Translated text

Key words

Peridynamics,GPU,Parallel computing,CUDA,Crack,Memory model

求助PDF

上传PDF

Bibtex

AI Read Science

AI Summary

AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.

Example

Background

Key content

Introduction

Methods

Results

Related work

Fund

Key content

Pretraining has recently greatly promoted the development of natural language processing (NLP)
We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance

Upload PDF to Generate Summary

Must-Reading Tree

Example

Generate MRT to find the research sequence of this paper