Towards Unlocking the Full Potential of Persistent Memory: A Hardware-Driven Approach

(CSE Colloquium Lecture Series)

Jishen Zhao

Speaker: Jishen Zhao, Assistant Professor, Computer Engineering, UC Santa Cruz
Date: Wednesday, April 12, 2017
Time: 11am
Location: Room 1202, CSE Building
Host: CSE Prof. Steve Swanson (

Abstract:  Persistent memory is a new tier in the memory/storage stack. It unifies memory and storage systems, offering fast access of memory and data recoverability of storage in a single device. Seeing the great value, many computer hardware and software suppliers recently started to adopt persistent memory in their next-generation systems. Though promising, persistent memory fundamentally changes the assumptions of current memory and storage systems. Unlocking its full potential is challenging.

In this talk, I will present my research on addressing a couple of such challenges of persistent memory systems. While most previous persistent memory designs manipulate data persistence by software, our recent study on hardware-driven undo+redo logging shows that abundant opportunities exist at much higher performance by leveraging existing memory hierarchy hardware. I will also discuss the implications of such logging scheme to persistent memory system software design. Furthermore, I will introduce the resource contention issue at the memory interface imposed by supporting data persistence in memory. We address this issue by developing a persistence-aware memory scheduling policy and a set of software support to facilitate such scheduling.

Bio:  Jishen Zhao is an Assistant Professor at the Computer Engineering Department of University of California, Santa Cruz. Before joining UCSC, she was a Research Scientist at HP Labs. Her research interests span the fields of computer architecture and system software, with an emphasis on memory and storage systems, domain-specific acceleration, and high performance computing. Her work received NSF CAREER award in 2017 and best paper honorable mention award at MICRO 2013.

Related Research Publications
Relaxing Persistent Memory Constraints with Hardware-Driven Undo+Redo Logging
FIRM: Fair and High-Performance Memory Control for Persistent Memory Systems