GDR-HGNN: A Heterogeneous Graph Neural Networks Accelerator Frontend with Graph Decoupling and Recoupling

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xue, Runzhen, Yan, Mingyu, Han, Dengke, Teng, Yihan, Tang, Zhimin, Ye, Xiaochun, Fan, Dongrui
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913304662769664
author Xue, Runzhen
Yan, Mingyu
Han, Dengke
Teng, Yihan
Tang, Zhimin
Ye, Xiaochun
Fan, Dongrui
author_facet Xue, Runzhen
Yan, Mingyu
Han, Dengke
Teng, Yihan
Tang, Zhimin
Ye, Xiaochun
Fan, Dongrui
contents Heterogeneous Graph Neural Networks (HGNNs) have broadened the applicability of graph representation learning to heterogeneous graphs. However, the irregular memory access pattern of HGNNs leads to the buffer thrashing issue in HGNN accelerators. In this work, we identify an opportunity to address buffer thrashing in HGNN acceleration through an analysis of the topology of heterogeneous graphs. To harvest this opportunity, we propose a graph restructuring method and map it into a hardware frontend named GDR-HGNN. GDR-HGNN dynamically restructures the graph on the fly to enhance data locality for HGNN accelerators. Experimental results demonstrate that, with the assistance of GDR-HGNN, a leading HGNN accelerator achieves an average speedup of 14.6 times and 1.78 times compared to the state-of-the-art software framework running on A100 GPU and itself, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2404_04792
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GDR-HGNN: A Heterogeneous Graph Neural Networks Accelerator Frontend with Graph Decoupling and Recoupling
Xue, Runzhen
Yan, Mingyu
Han, Dengke
Teng, Yihan
Tang, Zhimin
Ye, Xiaochun
Fan, Dongrui
Hardware Architecture
Heterogeneous Graph Neural Networks (HGNNs) have broadened the applicability of graph representation learning to heterogeneous graphs. However, the irregular memory access pattern of HGNNs leads to the buffer thrashing issue in HGNN accelerators. In this work, we identify an opportunity to address buffer thrashing in HGNN acceleration through an analysis of the topology of heterogeneous graphs. To harvest this opportunity, we propose a graph restructuring method and map it into a hardware frontend named GDR-HGNN. GDR-HGNN dynamically restructures the graph on the fly to enhance data locality for HGNN accelerators. Experimental results demonstrate that, with the assistance of GDR-HGNN, a leading HGNN accelerator achieves an average speedup of 14.6 times and 1.78 times compared to the state-of-the-art software framework running on A100 GPU and itself, respectively.
title GDR-HGNN: A Heterogeneous Graph Neural Networks Accelerator Frontend with Graph Decoupling and Recoupling
topic Hardware Architecture
url https://arxiv.org/abs/2404.04792