Lecture
This lecture explores the concept of in-memory computing (IMC) as a means to enhance energy efficiency in machine learning tasks. It delves into the transformation of memory accesses into analog/mixed-signal computations, leading to significant energy reduction. The talk covers IMC design principles, current trends, and future opportunities in deploying IMCs at scale. The instructor, Naresh Shanbhag, a prominent figure in electrical and computer engineering, has been actively involved in research on energy-efficient systems for machine learning and signal processing.