MQ-GNN: A Multi-Queue Pipelined Architecture for Scalable and Efficient GNN Training

Graph Neural Networks (GNNs) are powerful tools for learning graph-structured data, but their scalability is hindered by inefficient mini-batch generation, data transfer bottlenecks, and costly inter-GPU synchronization. Existing training frameworks fail to overlap these stages, leading to suboptima...

Descrición completa

Detalles Bibliográficos

Principais autores:	Irfan Ullah, Young-Koo Lee
Formato:	Artigo
Idioma:	Inglês
Publicado:	IEEE 2025-01-01
Series:	IEEE Access
Assuntos:	Graph neural network multi-GPU pipeline optimization mixed CPU-GPU training GNN training
Acceso en liña:	https://ieeexplore.ieee.org/document/10877815/
Tags:	Engadir etiqueta Sen Etiquetas, Sexa o primeiro en etiquetar este rexistro!