Change search
ReferencesLink to record
Permanent link

Direct link
CuPAN - high throughput on-chip interconnection for neural networks
Show others and affiliations
2015 (English)In: 22nd International Conference on Neural Information Processing, ICONIP 2015, Springer, 2015, 559-566 p.Conference paper (Refereed)Text
Abstract [en]

In this paper, we present a Custom Parallel Architecture for Neural networks (CuPAN). CuPAN consists of streamlined nodes that each node is able to integrate a single or a group of neurons. It relies on a high-throughput and low-cost Clos on-chip interconnection network in order to efficiently handle inter-neuron communication. We show that the similarity between the traffic pattern of neural networks (multicast-based multi-stage traffic) and topological characteristics of multi-stage interconnection networks (MINs) makes neural networks naturally suited to the MINs. The Clos network, as one of the most important classes of MINs, provide scalable low-cost interconnection fabric composed of several stages of switches to connect two groups of nodes and interestingly, can support multicast in an efficient manner. Our evaluation results show that CuPAN can manage the multicast-based traffic of neural networks better than the mesh-based topologies used in many parallel neural network implementations and gives lower average message latency, which directly translates to faster neural processing.

Place, publisher, year, edition, pages
Springer, 2015. 559-566 p.
Series
Lecture Notes in Computer Science, ISSN 0302-9743 ; 9491
Keyword [en]
Clos network, Multicast, Network-on-chip, Neural networks, Information science, MESH networking, Multicasting, Parallel architectures, Program processors, Switching networks, Throughput, Topology, VLSI circuits, Clos networks, Evaluation results, Interconnection fabrics, Multistage interconnection network, On-chip interconnection, On-chip interconnection network, Parallel neural networks, Topological characteristics, Integrated circuit interconnects
National Category
Communication Systems
Identifiers
URN: urn:nbn:se:kth:diva-181525DOI: 10.1007/978-3-319-26555-1_63ISI: 000371579800063ScopusID: 2-s2.0-84952009318ISBN: 9783319265544OAI: oai:DiVA.org:kth-181525DiVA: diva2:911315
Conference
9 November 2015 through 12 November 2015
Funder
VINNOVA
Note

QC 20160311

Available from: 2016-03-11 Created: 2016-02-02 Last updated: 2016-04-01Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full textScopus

Search in DiVA

By author/editor
Daneshtalab, Masoud
By organisation
Electronics and Embedded Systems
Communication Systems

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 12 hits
ReferencesLink to record
Permanent link

Direct link