AIRE325

Course Description

This course provides an overview on the fundamental design principles employed in the holistic design of AI-driven low power microcontrollers and embedded computing systems. Energy is always a bottleneck for resource-constrained embedded devices, especially to execute computationally expensive tasks. This course explains how different hardware and software schemes allow the designers to achieve energy efficiency and optimal power consumption while maintaining the maximum AI computational performance. Furthermore, in this course, the students will learn various methods that can improve the design schematics of the neural network architectures in order to give optimal trade-off between recognition performance and computational cost when deployed on the ultra-low power resource constrained devices.

Course Details