Efficient Resource Management

by Christos Kozyrakis, EPFL & Stanford University

June 6th 2016 @ 11:00, in Room Levico - Povo 2

Abstract:

An increasing amount of information technology services and data are now hosted in the cloud, primarily due to the cost and scalability benefits for both the end-users and the operators of the warehouse-scale datacenters (DCs) that host cloud services. Hence, it is vital to continuously improve the capabilities and efficiency of these large-scale systems. Over the past ten years, capability has improved by increasing the number of servers in a DC and the bandwidth of the network that connects them. Cost and energy efficiency have improved by eliminating the high overheads of the power delivery and cooling infrastructure. However, most of these scaling techniques are now reaching the points of diminishing returns.

To achieve further improvements, we must now examine how well we are utilizing the servers themselves, which are the primary determinant for DC performance, cost, and energy efficiency. The primary sources of low server utilization are the poor understanding of the relationship between resource usage and performance, the interference that makes it difficult to share resources in particular for latency critical workloads, and the hardware heterogeneity in modern datacenters. This talk presents our initial results on improving resource efficiency in data centers, including work on reducing and managing interference at the hardware, operating system, and cluster management level.

Speaker:

Christos Kozyrakis is a Professor of Computer and Communication Sciences at EPFL and an Associate Professor of Electrical Engineering & Computer Science at Stanford University.

He received a BS degree from the University of Crete (Greece) and a PhD degree from the University of California at Berkeley (USA), both in Computer Science. He is a fellow of the IEEE and a senior member of the ACM. He has received the ACM SIGARCH Maurice Wilkes award, the NSF Career Award, an IBM Faculty Award, the Okawa Foundation Research Grant, the Noyce Family Faculty Scholarship, and the Willard R. and Inez Kerr Bell Faculty Scholarship.

Dr. Kozyrakis works on architectures, runtime environments, and programming models for parallel computer systems. At Berkeley, he developed the IRAM architecture, a novel media-processor system that combined vector processing with embedded DRAM technology. At Stanford, he led the Transactional Coherence and Consistency (TCC) project at Stanford that developed hardware and software mechanisms for programming with transactional memory. He also led the Raksha project, that developed practical hardware support and security policies to deter high-level and low-level security attacks against deployed software. Dr. Kozyrakis is currently working on resource efficient computing for warehouse-scale datacenters.

Contact Info: Yannis Velegrakis (velgias@disi.unitn.eu)