Abstract: We have long understood, from fundamental circuit operation, that analog computation can be substantially more efficient than digital computation in lower dynamic range regimes. One of the dominant trends in AI computation has been low-precision computation for neural networks, providing an attractive opening to harness the efficiency of analog computation. However, there is an important gap between fundamental circuit operation and practical circuit operation at the scales demanded by today’s AI systems. This talk looks at analog computation specifically from the lens of practical architectures for delivering scaled-up system-level efficiency inflections. While this brings to the forefront critical challenges for many analog pathways, it also illuminates analog approaches that can and are driving significant efficiency boosts, as prevailing digital trajectories are saturating. As we think about these, it is also essential to consider how such inflections can be constructively adopted within the increasingly complex and multi-faceted ecosystem and system requirements, without breaking the way systems are constructed today and in the future. Beyond circuit approaches, this requires considering architectural and software design methodologies as well as alignment with key industrial trends.
Register now: https://ieee.webex.com/weblink/register/rb74586390b25b75db1c9afbe6dbf91fd
Biography: Naveen Verma received the B.A.Sc. degree in Electrical and Computer Engineering from the UBC, Vancouver, Canada, in 2003, and the M.S. and Ph.D. degrees in Electrical Engineering from MIT in 2005 and 2009, respectively. Since July 2009, he has been at Princeton University, where he is currently the Ralph H. and Freda I. Augustine Professor of Electrical and Computer Engineering. His research focuses on advanced sensing and computing systems. This includes research on large-area flexible sensors, energy-efficient computing architectures and circuits, and machine-learning and statistical-signal-processing algorithms. Prof. Verma has been involved in a number of technology transfer activities, including founding start-up companies. Most recently, he co-founded EnCharge AI, together with industry leaders in AI computing systems, to commercialize foundational technology developed in his lab. Prof. Verma has served as a Distinguished Lecturer for the IEEE Solid-State Circuits Society and on numerous conference program committees and advisory groups. Prof. Verma is the recipient of numerous teaching and research awards, including several best paper awards with his students.