Reducing Power Delivery Costs for Cloud Data Centers
Bhuvan Urgaonkar, Penn State University
Abstract:
Power-related costs are significant and growing components of overall data center expenditure (both capital and operational). In this talk, I will provide a brief overview of the sources and complexities of these costs, and of our work on using batteries and several IT knobs for reducing these costs. I will then provide a more detailed description of our use of stochastic control techniques for optimizing the data center's monthly electric utility bill. Finally, I will discuss complementary related work, open problems, and future directions.
Biography:
Bhuvan Urgaonkar is an associate professor of computer science and engineering at Penn State, where he has worked since 2005. He earned the Ph.D. (2005) and the M.S. (2002) degrees in computer science at the University of Massachusetts Amherst, and the B.Tech (honors) degree in computer science and engineering at the Indian Institute of Technology Kharagpur (1999). His research involves applying ideas from distributed computing, resource management, scheduling, performance evaluation, and analytical modeling to the design and evaluation of data centers, networked systems, operating systems, virtualization techniques, and storage systems.
Urgaonkar is a recipient of the NSF CAREER Award (2010) and a research award from Cisco (2007). With his co-authors, he has received the following awards or other special recognition: "Best Paper" at IEEE IISWC 2013, "Best Paper" at IEEE MASCOTS 2012, "Best Paper Nomination" at ACM Sigmetrics 2012, "Pick of the Month" by IEEE STC (Oct. 2012) for his paper at ACM ASPLOS 2012, "Best Student Paper" at IEEE MASCOTS 2008, and "Best Student Paper" at IEEE ICAC 2005. He is a senior member of the ACM, a senior member of the IEEE, and a member of USENIX.