Weiwei Jia Receives NSF CRII Award

University of Rhode Island (URI) computer engineering assistant professor Weiwei Jia was selected for a National Science Foundation (NSF) Computer and Information Science and Engineering Research Initiation Initiative (CRII) Award for his groundbreaking proposal, “A Novel Address Translation Architecture for Virtualized Clouds.”

This award supports early-career scientists in computer and information science engineering, enabling them to undertake exploratory research and develop collaborations and new approaches.

Cloud computing is leading to amazing productivity, but using systems that establish trust can cut processor cycles and energy use, Jia explains.

“It is an honor to receive this award. It is very competitive and makes me proud that NSF recognizes my research, and that of the faculty members at our university,” said Jia.

Jia’s research aims to drive the evolution of cloud computing, facilitating modern clouds and data centers to seamlessly operate new data-intensive applications like AI with exceptional performance, scalability, and minimal overhead. This would be achieved by changing the way address translation is done in cloud environments, specifically beneficial to persistent memory (PM) and other emerging cloud memory/storage devices. The idea is to bypass and short-circuit the lengthy translation process of traditional techniques by leveraging unique observations through extensive experiments.

This project introduces XLANE, a general, effective, and scalable address translation architecture for virtualized clouds with high efficiency. XLANE includes two techniques: direct memory translation (DMT) and direct file translation (DFT).

“Imagine it like a navigational app, not only finding the fastest route to your destination, but creating one specifically for directness, eliminating roadblocks,” said Jia.

DMT allows hardware subsystems to translate main system memory independently of the central processing unit and can reduce the number of sequential memory accesses from 24 to two in each address translation for memory virtualization. DFT is the first direct file mapping design for virtualizing ultra-low latency storage devices like persistent memory with efficient direct file translation. XLANE is a general approach that benefits not only PM systems, but also compute express link (CXL)-based memory systems, tiered memory systems, and many others.

This project will create a new outreach program based on URI Capstone Design Program and the Established Program to Stimulate Competitive Research (EPSCoR) to involve at least two students every year, one undergraduate student and one graduate student, especially underrepresented students.

“I am working to develop next-generation cloud-computing and this grant helps to pave the way,” said Jia.

Based on the results of this CRII proposal, Jia plans to study how to efficiently virtualize emerging memory/storage devices to significantly improve the performance of new applications like AI in modern clouds and data centers. The resultant systems are likely to be adopted by major cloud platforms and the research results will be incorporated into URI’s computer engineering curriculum.

The total award is $174,999 for a two-year period after which Jia intends to submit an NSF CAREER proposal for continued research.

Story by Krysta Murray

“It is an honor to receive this award. It is very competitive and makes me proud that NSF recognizes my research, and that of the faculty members at our university.”