The STAR Lab focuses on the research, development and educational endeavors in the broad area of computing systems and applications. We perform cutting-edge research on a variety of technologies to improve the performance, energy-efficiency, reliability and security of computing systems across a growing landscape, from embedded and mobile devices to supercomputers and data-centers. Some recent focuses include machine learning accelerators, post-Moore's era architecture, extreme-scale computing, applications of AI/ML in architecture designs, and Internet of Things (IoT). We are also interested in exploring novel applications in machine learning and natural language processing (e.g., large language models) that are enabled by efficient computing systems. Below are a few on-going and past projects.

Machine Learning for Natural Language Processing

Machine Learning for Computer Architecture and System

Machine Learning Accelerators

GPU Architectures and Extreme-scale computing

Harnessing Dark Silicon for Post-Moore Era Computing

Cache and Memory Bandwidth Partitioning

Deadlock-free Interconnection Networks

Application-aware Optimizations for Many-core Processors

Transactional Memory & Parallel Programming