Visiting Scholar Series | Highlights from Prof. Yajun Ha's Lecture

On March 25, Professor Yajun Ha from ShanghaiTech University visited Yungu Campus of Westlake University at the invitation of Chair Professor Mohamad Sawan, delivering a cutting-edge lecture in the field of microelectronics titled "Energy-Efficient Edge Computing for Intelligent Applications in the Post-Moore Era."

From robotics to autonomous driving, edge computing is quietly reshaping our lives.

Smart applications such as robot and advanced driver assistance have significantly made our life easier and better. For the sake of various constraints such as real time, many of these applications involve computing in mobile edge devices. These devices are usually powered by batteries that require energy efficient computing to enable longer interval between the charging of batteries. On the other hand, new semiconductor technology (such as non-volatile memories, cryogenic microelectronics) and new computing paradigm (such as in-memory computing and neural network computing) have been actively researched to support energy efficient computing.

Identify problems from device characteristics and seek solutions in application scenarios.

Professor Ha's lecture commenced with an insightful integration of circuit design theory and emerging memory device technologies. Drawing upon his team's pioneering work in robotics and smart vehicles, he vividly demonstrated the innovative "bottom-up meets top-down" dual-thinking methodology. His presentation detailed the team's groundbreaking energy-efficient solutions, showcasing multidimensional innovations that successfully address the dual challenges of battery endurance and real-time processing requirements.

With extensive interdisciplinary research experience, Professor Ha particularly emphasized the importance of establishing cross-disciplinary research platforms to bridge theoretical exploration and practical implementation. He also provided actionable recommendations for effective interdisciplinary collaboration.

When addressing Ph.D student Ziyang Shen's question regarding the "bottleneck in edge-side large model deployment," Professor Ha advocated for a top-down approach: "Let application demands drive innovation in edge-side model deployment." He further highlighted that the convergence of large language models with next-generation memory technologies presents fertile ground for novel ideas and represents a prime opportunity for cross-disciplinary synergy.

During this visiting, Professor Ha spoke highly of the interdisciplinary research at CenBRAIN Neurotech and engaged in discussions with our students on topics such as key technological breakthroughs and the growth of young researchers.