Speaker: Associate Professor John Kim (KAIST)
Date & Time: 2018년 2월 6일 (화) 오전 10:30
Where: 삼성전자 서울대학교 공동연구소 (관악 연구공원 내) A동 1층 대강당
Memory-centric networks had been proposed to interconnect multiple memory modules together. While it has been proposed as a new paradigm, it is not clear if this approach is simply an old, same problem or if there are new opportunities or challenges. In this talk, I will present some of our work on memory network, how it can be used to provide a more secure system, and how it can be used to accelerate deep learning workloads.
John Kim is an associate professor in the School of Computing at KAIST and currently spending his sabbatical at HP Labs. He received his Ph.D. from Stanford University and his B.S. and M.Eng from Cornell University. Prior to graduate school, John has worked on the design of several processors at Motorola and Intel. His research interest includes computer architecture, interconnection networks, and mobile systems. His work on the Dragonfly topology was selected as an IEEE Micro Top Pick and has received a Google Faculty Research Award.