rCUDA: A Piece of Research of the Parallel Architectures Group
The Parallel Architectures Group is located at Technical University of Valencia, Spain. Its research addresses several topics, ranging from routing algorithms and congestion control for off-chip interconnects to the co-design of coherency protocols with on-chip networks. Other topics include areas as diverse as power-aware scheduling for multicore systems or remote GPU virtualization.
In this talk, after briefly introducing the research carried out in the Parallel Architectures Group, one of its most prominent technologies, rCUDA, will be presented. rCUDA is a remote GPU virtualization framework intended to make more flexible and efficient the use of GPUs within data centers.
Federico Silla received the MS and PhD degrees in Computer Engineering from the Technical University of Valencia, Spain, in 1995 and 1999, respectively. He is currently an associate professor at the Department of Computer Engineering(DISCA) at that university, and external contributor of the Advanced Computer Architecture research group at the Department of Computer Engineering at the University of Heidelberg. He is also member of the Advanced Technology Group of the HyperTransportConsortium, whose main result to date has been the development and standardization of an extension to HyperTransport (High Node Count HyperTransport Specification 1.0).
Furthermore, he worked for two years at Intel Corporation, developing on-chip networks. His research addresses high performance on-chip and off-chip interconnection networks as well as distributed memory systems and GPU virtualization. He has published numerous papers in peer-reviewed conferences and journals, as well as several book chapters. He has been member of the Program Committee in several of the most prestigious conferences in his area, including IPDPS, HiPC, SC, PACT, ISC, etc.