Deep Neural Network Hardware

Prof. Dr.-Ing. habil. Dr. sc. nat. Christian Mayr,

Dr.-Ing. Johannes Partzsch

"Deep Neural Network Hardware"

(EuI-IST-E-DNNH, EuI-NES-E-DNNH)

in the summer semester
(to the timetable)

Compulsory elective module for

• the Master's degree program Nanoelectronic
  Systems

Content of the module is the design of hardware accelerators for artificial neural networks - Deep Neural Networks, DNN-, from architectures to arithmetic building blocks.

Qualification goals:

After completing the module, students will have solid knowledge of the key design decisions in DNN accelerators. They will be able to select or design an accelerator for a given application. They know and understand the necessary steps for executing DNNs on hardware accelerators, as well as common optimization methods for DNN accelerators.