The field of memory management has focused primarily on automated, low overhead techniques that minimize the memory capacity used by applications. However, as the memory system is becoming a major determinant of the performance, cost, and energy consumption for computers of any size, there are other memory system issues that merit automated management. This talk will review three such issues, discuss the hardware and system trends behind them, and highlight potential directions for research.
First, we will discuss the use of memory for fast, distributed storage for data center applications that require low latency guarantees. This issue will likely require the management of a heterogeneous memory system in each server that mixes fast and energy efficient memory devices. It also requires memory management across thousands of servers in order to achieve system wide performance goals. Second, we will discuss the issue of locality in many-core systems. Even within a single chip, scheduling parallel tasks is now an issue of locality management as frequent remote accesses can cause large performance and energy waste. This issue motivates scheduling techniques that use domain or application specific information about memory usage. Finally, we will discuss opportunities to introduce new functionality in the memory system in order to reduce communication, assist with synchronization, or enforce quality of service requirements. To justify the hardware changes needed for such functionality, we need automated management techniques that can build upon its potential.
Bio: Christos Kozyrakis is an Associate Professor of Electrical Engineering & Computer Science at Stanford University and a Visiting Scientist at Google Inc. He works on architectures, runtime environments, and programming models for parallel computing 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 hardware and software techniques for next-generation data centers. He is also a member of the Pervasive Parallelism Lab at Stanford, a multi-faculty effort to make parallel computing practical for the masses.
Christos 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 the Willard R. and Inez Kerr Bell faculty scholar at Stanford and a senior member of the ACM and the IEEE. Christos has received the NSF Career Award, an IBM Faculty Award, the Okawa Fundantion Research Grant, and a Noyce Family Faculty Scholarship.