•   When: Tuesday, May 21, 2019 from 01:30 PM to 02:30 PM
•   Speakers: Tevfik Kosar, University at Buffalo (SUNY)
•   Location: Nguyen Engineering 4201
•   Export to iCal

Abstract:
With the emergence of complex scientific applications, social media, video over IP, and more recently the trend for Internet of Things (IoT), the global data movement requirements have already exceeded the exabyte scale. It is estimated that, in 2019, more IP traffic will traverse global networks than all prior "Internet years" combined. To support this massive amount of data movement across the globe, the networking infrastructure as well as the source and destination nodes at the edges consume immense amount of electric power, with an estimated cost measured in billions of dollars. Although considerable amount of research has been done on power management techniques for the networking infrastructure, there has not been much prior work focusing on power-aware data transfer algorithms for minimizing the power consumed at the end-systems. In this work, we analyze various factors that affect the power consumption in end-to-end (cloud-to-edge and edge-to-edge) data transfers, such as the level of parallelism, concurrency, pipelining, CPU frequency scaling, and multi-core scheduling at the end-systems. We introduce novel data transfer algorithms which aim to achieve high data transfer throughput while keeping the energy consumption during the transfers at the minimal levels. Our novel solution combines historical data analysis with
real-time online learning and achieves significant amount of energy savings (60-80%) at the end-systems during data transfers without sacrificing from the performance. In many cases, we can achieve performance increase and save energy at the same time. 

Biography:
Tevfik Kosar is an Associate Professor in the Department of Computer Science and Engineering at the State University of New York at Buffalo (UB). He has received his Ph.D. in Computer Science from the University of Wisconsin-Madison in 2005. His main research interests include data-intensive distributed computing, big-data analytics, datacenter networking and I/O optimization, performance and energy efficiency in IoT and edge computing systems. Some of the awards received by Dr. Kosar include NSF CAREER Award, IBM Research Award, Google Research Award, LSU Rainmaker Award, LSU Flagship Faculty Award, Baton Rouge Business Report’s Top 40 Under 40 Award, 1012 Corridor’s Young Scientist Award, UB Senior Faculty Teaching Award, and UB Senior Faculty Research Award. Dr. Kosar has published more than 100 scholarly articles in top journals and conferences in his area. He consistently serves on the PC of related conferences such as IEEE/ACM SC, ACM HPDC, IEEE CLUSTER, IEEE ICDCS, IEEE CLOUD and IEEE BigData. He is currently serving in the US Environmental Protection Agency’s e-Manifest Advisory Board, and in the Editorial Boards of the IEEE Transactions on Parallel and Distributed Systems, and the Cluster Computing journal.