Title:  Design Mehodology for 3D-stacked Imaging Systems with Integrated Deep Learning

Committee: 

Dr. Mukhopadhyay, Advisor 

Dr. Khan, Chair

Dr. Yalamanchili

Abstract:

The objective of the proposed research is to explore sensor integrated computing with Neurosensor, a 3D-stacked image sensor with integrated deep learning. Such an architecture will enable the deployment of smart sensors that perform advanced neural acceleration in-field, with 3D integration helping to avoid the various pitfalls (such as routing challenges and associated latency) of designing this type of high bandwidth systems with a large degree of connectivity. The architecture of the system will be explored and the various design implications, both from technology perspective as well as circuit perspective, will be investigated. From the technology side, the impact of ultra low power devices on these memory-heavy systems will be discussed, and the various circuit considerations will be evaluated using a 2D image sensor chip. The proposed research will conclude with a discussion on how to effectively power these sensor based systems.