Networking and SystemsNetworking and Systems focuses on research and development of major networking topics, including but not limited to wireless networks, mobile location estimation, ubiquitous computing, Internet systems, communication services, peer-to-peer networks, optical networks and high-performance computing etc.
Funded Research and Consultancy Projects in the Past Few Years:
|Energy-efficient Training of Multiple Deep Learning Models on GPU Clusters
||Prof. CHU, Xiaowen |
- To build up an open data set of GPU performance and power with DVFS;
- To develop performance and power models for the training jobs of deep learning models;
- To design online algorithms for resource allocation and task scheduling problem on GPU clusters;
- To implement an open-source job management and scheduling system, and to evaluate the performance of the proposed solutions using real-world experiments.