CS Special Seminar: Bo Zhao "Scalable and Sustainable Data-Intensive Systems"

Scalable and Sustainable Data-Intensive Systems

Bo Zhao
Imperial College London
Homepage

Abstract: Efficient data-intensive systems translate data into value for decision making. As data is collected at unprecedented rates for timely analysis, the model-centric paradigm of machine learning (ML) is shifting towards a data-centric and system-centric paradigm. Recent breakthroughs in large ML models (e.g. GPT3 and ChatGPT) and the remarkable outcomes of reinforcement learning (e.g. AlphaFold and AlphaCode) have shown that scalable data management and its optimizations are critical to obtain state-of-the-art performance. This talk aims to answer the question “how to co-design multiple layers of the software/system stack to improve scalability, performance, and energy efficiency of ML and data-intensive systems”. It addresses the challenges to build fully automated data-intensive systems that integrate the ML layer, the data management layer, and the compilation-based optimization layer. Finally, this talk will sketch and explore the vision to leverage the computational advantage of quantum computing on hybrid classic/quantum systems in the post-moore era.

Bio: Bo Zhao is a lecturer (assistant professor) in Computer Science at Queen Mary University of London. He also holds an honorary research fellow in the Department of Computing at Imperial College London. Bo Zhao conducts research of efficient data-intensive systems at the intersection of scalable reinforcement learning systems and distributed data management systems, as well as compilation-based optimization techniques. His long-term goal is to explore and understand the fundamental connections between data management and modern machine learning systems to make decision-making transparent, robust and efficient.