TY - GEN
T1 - Exploring the Impact of Attacks on Ring AllReduce
AU - Wang, Jiayu
AU - Liu, Peng
AU - Guo, Zehua
AU - Liu, Sen
AU - Yao, Chao
N1 - Publisher Copyright:
© 2021 ACM.
PY - 2021/6/24
Y1 - 2021/6/24
N2 - Distributed Machine Learning (DML) is widely used to accelerate the training of the deep learning model. In DML, Parameter-Server (PS) and Ring AllReduce are two typical architectures. Recently, observing that many works address the security problem in PS, whose performance can be greatly degraded by malicious participation during the training process. However, the robustness of Ring AllReduce, which can solve the communication bandwidth problem in PS, to the malicious participant is still unknown. In this paper, we design a series of experiments to explore the security problem in Ring AllReduce, and reveal it can also suffer from the malicious participant.
AB - Distributed Machine Learning (DML) is widely used to accelerate the training of the deep learning model. In DML, Parameter-Server (PS) and Ring AllReduce are two typical architectures. Recently, observing that many works address the security problem in PS, whose performance can be greatly degraded by malicious participation during the training process. However, the robustness of Ring AllReduce, which can solve the communication bandwidth problem in PS, to the malicious participant is still unknown. In this paper, we design a series of experiments to explore the security problem in Ring AllReduce, and reveal it can also suffer from the malicious participant.
UR - http://www.scopus.com/inward/record.url?scp=85124481787&partnerID=8YFLogxK
U2 - 10.1145/3469393.3469676
DO - 10.1145/3469393.3469676
M3 - Conference contribution
AN - SCOPUS:85124481787
T3 - ACM International Conference Proceeding Series
SP - 12
EP - 13
BT - Proceeding of the 5th Asia-Pacific Workshop on Networking, APNet 2021
PB - Association for Computing Machinery
T2 - 5th Asia-Pacific Workshop on Networking, APNet 2021
Y2 - 24 June 2021 through 25 June 2021
ER -