Dorylus: Affordable, Scalable, and Accurate GNN Training with Distributed CPU Servers and Serverless Threads

Authors: 

John Thorpe, Yifan Qiao, Jonathan Eyolfson, and Shen Teng, UCLA; Guanzhou Hu, University of Wisconsin, Madison; Zhihao Jia, CMU and Facebook; Jinliang Wei, Google; Keval Vora, Simon Fraser University; Ravi Netravali, UCLA/Princeton University; Miryung Kim and Guoqing Harry Xu, UCLA

BibTeX
@inproceedings {273743,
title = {Dorylus: Affordable, Scalable, and Accurate {GNN} Training with Distributed {CPU} Servers and Serverless Threads},
booktitle = {15th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 21)},
year = {2021},
url = {https://www.usenix.org/conference/osdi21/presentation/thorpe},
publisher = {{USENIX} Association},
month = jul,
}