Loading Events

Hardware/Software Co-optimization for Brain-inspired Learning Systems by Bipin Rajendran

Information Theory Society

Abstract: The dream of human-level artificial intelligence (AI) may well be within reach, thanks to the far-reaching advances in deep learning over the past decade. However, traditional von Neumann computing systems are ill-suited for training deep networks, as the need to continuously shuttle large amounts of data between the physically separated memory and logic engines limits the achievable performance. This has also restricted their adoption for stand-alone applications which require learning in the field such as embedded sensing, telecommunications, and robotics. Hence, a fundamental re-imagining of the algorithmic paradigms and system architecture is necessary to develop learning platforms for energy and memory constrained applications.

This talk will also describe a non-von Neumann computational architecture which is ideally suited for realising systems capable of learning in the field, as well as in an accelerated manner, possibly surpassing human performance. Recent results from our research on optimising devices, algorithms, and systems for learning and adaptation that are inspired by some of the key organisational principles of the brain will be presented.

About the speaker: Bipin Rajendran received a B.Tech degree from I.I.T. Kharagpur in 2000, and M.S. and Ph.D. degrees in Electrical Engineering from Stanford University in 2003 and 2006, respectively. He was a Master Inventor and Research Sta Member at IBM T. J. Watson Research Center in New York during 2006-’12 and a faculty member in the Electrical Engineering Department at I.I.T. Bombay during 2012-’15. His research focuses on building algorithms, devices and systems for brain-inspired computing. He has authored over 60 papers in peer-reviewed journals and conferences, and has been issued 55 U.S. patents. He is currently an Associate Professor of Electrical & Computer Engineering at New Jersey Institute of Technology.

Share This Story, Choose Your Platform!

Go to Top