报告人：香港科技大学 Kai Chen教授
报告题目：Towards a Scalable and Efficient RDMA Networking for Datacenters
个人简介: Kai Chen is a Professor at HKUST and directs the HKUST iSING Lab. He received his BS and MS from University of Science and Technology of China (USTC) in 2004 and 2007, and PhD from Northwestern University in 2012, respectively. His research interests include Data Center Networking, High-performance Networking, Machine Learning Systems, and Hardware Acceleration. His work has been published in various top venues such as SIGCOMM, NSDI, OSDI and TON, etc., including a SIGCOMM best paper candidate. He is the Steering Committee Chair of APNet, serves on the Program Committees of SIGCOMM, NSDI, CoNEXT, etc., and Editorial Boards of IEEE/ACM Transactions on Networking, Big Data, and Cloud Computing.
摘要：As the datacenter networking migrates from 10Gb/s to 40/100Gb/s or beyond, the traditional software-based TCP/IP stack cannot keep up with the increasing network speed. Consequently, hardware-based RDMA (Remote Direct Memory Access), originally invented in HPC, is now experiencing a renaissance in Ethernet-based datacenters. However, RDMA relies on lossless network to take effect, which poses significant challenges before it can be well utilized in lossy datacenter networking. In this lecture, I will overview the research works from the community in the past 10 years to address these challenges, as well as our efforts in the HKUST iSING Lab towards a scalable and efficient RDMA networking for datacenters.