Feedforward neural networks on massively parallel architectures
2020 (English)In: Hardware Architectures for Deep Learning, Institution of Engineering and Technology , 2020, p. 53-76Chapter in book (Other academic)
Abstract [en]
In this chapter, we present ClosNN, a specialized NoC for NNs based on the well-known Clos topology. Clos is perhaps the most popular Multistage Interconnection Network (MIN) topology. Clos is used commonly as a base of switching infrastructures in various commercial telecommunication and network routers and switches.
Place, publisher, year, edition, pages
Institution of Engineering and Technology , 2020. p. 53-76
Keywords [en]
ClosNN, Feedforward neural nets, Feedforward neural network, Massively parallel architectures, MIN topology, Multistage interconnection network topology, Network routers, Network-on-chip, NoC, Parallel architectures, Switches, Switching infrastructures
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:mdh:diva-62548DOI: 10.1049/PBCS055E_ch3Scopus ID: 2-s2.0-85153639577ISBN: 9781785617683 (print)OAI: oai:DiVA.org:mdh-62548DiVA, id: diva2:1756355
Note
Book chapter; Export Date: 11 May 2023; Cited By: 0
2023-05-112023-05-112023-05-11Bibliographically approved