研究方向：
Predictable Performance in the Cloud (云计算性能保证): Performance guarantee for cloud workloads; Performance measurements, modeling, and optimization in clouds; Bandwidth allocation of virtual machines; Live migration of virtual machines and containers.
Distributed Machine Learning Systems (分布式机器学习系统): Performance measurements, modeling, and optimization of distributed machine learning (e.g., Tensorflow, MXNet, PyTorch).
Performance Optimization of Big Data Processing Frameworks (大数据处理框架性能优化): Performance modeling and optimization of DAG-style data analytics jobs (i.e., MapReduce, Spark); Task/Stage/Job scheduling of big data processing frameworks.