Instructors: Kyle Shiflett and Avinash Karanth Credit Hours: 3 EE 4900/5900 is intended to introduce students to basic deep neural networks, and provide an indepth study of computer architecture methods for efficient training and inference of deep neural networks. The recent proliferation of artificial intelligence, in particular deep neural networks (DNNs), has led to an increasing pressure on hardware systems that run these models. DNNs have established some of the leading state-of-the-art models for tasks such as image classification and speech recognition, some even achieving super-human accuracy. As the size and complexity of DNN models continue to grow, as does the need for energy-efficient and fast execution of these workloads. This course focuses on recent computer architecture trends and hardware-software co-design techniques that facilitate efficient execution of DNNs. Topics covered in this course include: * Multilayer perceptrons * Convolutional neural networks * Affine and nonlinear integer quantization * Operand dataflow and stationarity * Hardware accelerators * Compression with sparsity and pruning * Memory organization * Interconnects * Training -------------- next part -------------- An HTML attachment was scrubbed... URL: < http://listserv.ohio.edu/pipermail/eecs_mscs/attachments/20211102/c3213936/attachment.html >
(740) 593–9381 | Building 21, The Ridges
Ohio University | Athens OH 45701 | 740.593.1000 ADA Compliance | © 2018 Ohio University . All rights reserved.