Grantees of the 2018 TechAccelR Grants
Efficient Convolutional-Neural-Network (CNN) Processing System
Professor Zeljko Zilic (Electrical and Computer Engineering-ECE) and Pavel Sinha, PhD candidate (ECE)
Executive Summary
To further develop an efficient computing platform technology in silicon that provides industry best in class performance in throughput, low-power and achievable depth of the deep-learning network in real-time. We have re-defined the computation platform by introducing a highly scalable architecture to accommodate any order deep-learning architecture as desired by the developer.