106/10/11(三) Rethinking Memory System Design for Emerging Non-Volatile Memory Technologies
國立清華大學資訊工程學系
Department of Computer Science
National Tsing Hua University
專題演講
SEMINAR
主講人：Hsiang-Yun Cheng
SPEAKER (Assistant Research Fellow CITI at Academia Sinica)
題 目：Rethinking Memory System Design for
TOPIC Emerging Non-Volatile Memory Technologies
時 間：106年10月11日(三) 1:30PM - 3:00PM
DATE
地 點： 資電館地下演講廳
PLACE
聯絡人：何宗易教授
Abstract:
The memory system, including processor caches, main memory, and storage,
has become a performance and energy bottleneck in almost all computer systems,
as commodity memory technologies with limited scalability are not sufficient to meet
the increasing bandwidth and energy efficiency demands of modern applications.
Recently, emerging byte-addressable non-volatile memories (NVMs) have been
explored as potential alternatives to replace conventional SRAM, DRAM, and HDD.
These NVM technologies, such as Spin Torque Transfer RAM (STT-RAM),
Phase-Change RAM (PCM), and Resistive RAM (ReRAM), have promising features:
they have high density, consume low leakage, and can retain data without power
supply. However, these NVMs also have various shortcomings that need to be
overcome. As a result, substantial advancements in redesigning the existing memory
organization are needed to realize the full potential of NVMs.
In this talk, we will first introduce the background of emerging NVM
technologies, with an emphasis on their promising features, drawbacks, and how to
leverage NVMs to replace commodity memories in the memory hierarchy. We will
then examine some promising research and design directions to tackle challenges
posed by NVMs. In particular, we will introduce an energy-efficient inclusion policy
designed for NVM-based caches and discuss the design issues of NVM-based main
memory and storage. We will also explore new opportunities that NVMs may bring
forward in future computer systems.
Bio:
Hsiang-Yun Cheng is currently an assistant research fellow of the Research
Center for Information Technology Innovation (CITI) at Academia Sinica. She
received her B.S. and M.S. degree in Computer Science and Information
Engineering from National Taiwan University. She later received her Ph.D. degree
from Pennsylvania State University in 2016. Her research falls primarily in the field of
computer architecture, with an emphasis on memory system design. She is
especially interested in exploiting emerging technologies (e.g., non-volatile
memories, 3D integrated circuits) and the characteristics of modern applications
(e.g., big data processing, parallel computing, machine learning) to design
energy-efficient computer systems.
瀏覽數: