GDR-HGNN：一种具有图解耦与重耦的异构图神经网络加速器前端
GDR-HGNN: A Heterogeneous Graph Neural Networks Accelerator Frontend with Graph Decoupling and Recoupling
AI总结：
针对HGNN加速器因不规则内存访问导致的缓冲区颠簸问题，提出GDR-HGNN硬件前端，通过运行时图解耦与重耦增强数据局部性，最高带来14.6倍平均加速。
摘要：
异构图神经网络（HGNN）已将图表示学习的适用性扩展到异构图。然而，HGNN的不规则内存访问模式会导致HGNN加速器中的缓冲区颠簸问题。在本工作中，我们通过分析异构图的拓扑结构，发现了解决HGNN加速中缓冲区颠簸问题的机会。为利用这一机会，我们提出了一种图重构方法，并将其映射到名为GDR-HGNN的硬件前端中。GDR-HGNN会在运行时动态重构图，以增强HGNN加速器的数据局部性。实验结果表明，在GDR-HGNN的辅助下，一款领先的HGNN加速器与运行在A100 GPU上的最先进软件框架以及其自身相比，分别实现了14.6倍和1.78倍的平均加速比。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
