Systems & Networking Research Projects
Operating System Support for Large Memory Systems
This project tackles performance limitations in large server systems with extensive memory. It proposes OS-level solutions to optimize memory access for both traditional Non-Uniform Memory Access (NUMA) systems and emerging disaggregated memory architectures. The goal is to improve performance, power efficiency, and cost-effectiveness for data centers.
Machine Learning for Compilation
This project aims to enhance compiler optimization using machine learning. It utilizes graph convolutional neural networks (GCN) to predict optimization efficacy for code fragments. Key objectives include developing effective graph embeddings, selecting appropriate GCN models, and creating a compiler optimization recommender system.
Compiler-Guided High-Performance Accelerator Design
Project aims to optimize FPGA acceleration through DSL platform. It facilitates efficient, easy-to-develop designs via compiler-based generators. The goal is high-efficiency, easy deployment, and specialized FPGA designs through abstraction-level optimizations.
Towards a Testbed for Innovative Inter-Networking Research
This project aims to achieve performance-guaranteed stream processing using a serverless paradigm. By designing Stream as a Service (SaaS) abstractions and APIs, users can specify computational logic and performance goals for their stream jobs, advancing serverless technology for real-time analytics with a broader impact.
Enhancing Legal Document Services with Accessible and Private LLM Technology
This project aims to transform legal document services by developing a local, privacy-preserving Large Language Model (LLM) technology. The research tackles privacy concerns associated with external server processing, seeking to enhance efficiency, speed, and reliability while overcoming computational constraints.
Tackling Energy-Efficient Reception Challenge for Next Billion IoT Devices
The proposal addresses the energy challenge in IoT by designing energy-efficient receivers with tunnel diodes and LiFi or visible light communication, to enable sustainable, battery-free IoT devices.
Multimodal AI for IoT
This project is partially funded by Microsoft's Accelerate Foundation Models Research Program.
Research Programme in Assuring Hardware Security by Design in Systems on Chip (SOCure)
NUS-NCS Joint Laboratory
NCS Pte. Ltd. (NCS) and the National University of Singapore (NUS) have established a joint research lab that is hosted in NUS to conduct research, develop capabilities and innovative digital solutions to protect individuals, businesses and public agencies in Singapore from a wide range of cyber threats. The joint lab is governed by a Management Committee comprising members from NUS and NCS.
Continuous Monitoring of Face-to-Hand Interactions via On-Ear Wearables
- Internet of Things, Mobile Computing & Sensing
Record & Replay: Framework for Network-wide Monitoring and Debugging
- Computer Networks, Programmable Networks
In-Network Acceleration for Latency Sensitive Applications for Future Communication Systems
- Computer Networks, Wireless Networks
-
Computing 1
13 Computing Drive
Singapore 117417
Additional Links
© National University of Singapore. All Rights Reserved. • Legal • Branding guidelines