A Review of spiking neuromorphic hardware communication systems

Aaron R. Young; Mark Dean; James S. Plank; Garrett S. Rose

doi:10.1109/ACCESS.2019.2941772

A Review of spiking neuromorphic hardware communication systems

Aaron R. Young, Mark Dean, James S. Plank, Garrett S. Rose

Research output: Contribution to journal › Article › peer-review

83 Scopus citations

Abstract

Multiple neuromorphic systems use spiking neural networks (SNNs) to perform computation in a way that is inspired by concepts learned about the human brain. SNNs are artificial networks made up of neurons that fire a pulse, or spike, once the accumulated value of the inputs to the neuron exceeds a threshold. One of the most challenging parts of designing neuromorphic hardware is handling the vast degree of connectivity that neurons have with each other in the form of synaptic connections. This paper analyzes the neuromorphic systems Neurogrid, Braindrop, SpiNNaker, BrainScaleS, TrueNorth, Loihi, Darwin, and Dynap-SEL; and discusses the design of large scale spiking communication networks used in such systems. In particular, this paper looks at how each of these systems solved the challenges of forming packets with spiking information and how these packets are routed within the system. The routing of packets is analyzed at two scales: How the packets should be routed when traveling a short distance, and how the packets should be routed over longer global connections. Additional topics, such as the use of asynchronous circuits, robustness in communication, connection with a host machine, and network synchronization are also covered.

Original language	English
Article number	8843969
Pages (from-to)	135606-135620
Number of pages	15
Journal	IEEE Access
Volume	7
DOIs	https://doi.org/10.1109/ACCESS.2019.2941772
State	Published - 2019
Externally published	Yes

Funding

The next group of neuromorphic systems all come from the Human Brain Project (HBP) [18]. HBP is funded by the European Union with the goal of ‘‘building a research infrastructure to help advance neuroscience, medicine and computing’’ [19]. This work was supported in part by The University of Tennessee (UT) Science Alliance Joint Directed Research and Development Program, and in part by the Air Force Research Laboratory under Agreement FA8750-19-1-0025.

Funders	Funder number
Air Force Research Laboratory	FA8750-19-1-0025
University of Tennessee
European Commission

Keywords

Communication protocols
field programmable gate arrays
interconnect
network on chip
neuromorphic
spiking network communication
spiking neural network
very large scale integration

Access to Document

10.1109/ACCESS.2019.2941772

Cite this

@article{f01f6bea589d4098944c9af38e46744d,

title = "A Review of spiking neuromorphic hardware communication systems",

abstract = "Multiple neuromorphic systems use spiking neural networks (SNNs) to perform computation in a way that is inspired by concepts learned about the human brain. SNNs are artificial networks made up of neurons that fire a pulse, or spike, once the accumulated value of the inputs to the neuron exceeds a threshold. One of the most challenging parts of designing neuromorphic hardware is handling the vast degree of connectivity that neurons have with each other in the form of synaptic connections. This paper analyzes the neuromorphic systems Neurogrid, Braindrop, SpiNNaker, BrainScaleS, TrueNorth, Loihi, Darwin, and Dynap-SEL; and discusses the design of large scale spiking communication networks used in such systems. In particular, this paper looks at how each of these systems solved the challenges of forming packets with spiking information and how these packets are routed within the system. The routing of packets is analyzed at two scales: How the packets should be routed when traveling a short distance, and how the packets should be routed over longer global connections. Additional topics, such as the use of asynchronous circuits, robustness in communication, connection with a host machine, and network synchronization are also covered.",

keywords = "Communication protocols, field programmable gate arrays, interconnect, network on chip, neuromorphic, spiking network communication, spiking neural network, very large scale integration",

author = "Young, {Aaron R.} and Mark Dean and Plank, {James S.} and Rose, {Garrett S.}",

note = "Publisher Copyright: {\textcopyright} 2013 IEEE.",

year = "2019",

doi = "10.1109/ACCESS.2019.2941772",

language = "English",

volume = "7",

pages = "135606--135620",

journal = "IEEE Access",

issn = "2169-3536",

publisher = "Institute of Electrical and Electronics Engineers",

}

TY - JOUR

T1 - A Review of spiking neuromorphic hardware communication systems

AU - Young, Aaron R.

AU - Dean, Mark

AU - Plank, James S.

AU - Rose, Garrett S.

PY - 2019

Y1 - 2019

N2 - Multiple neuromorphic systems use spiking neural networks (SNNs) to perform computation in a way that is inspired by concepts learned about the human brain. SNNs are artificial networks made up of neurons that fire a pulse, or spike, once the accumulated value of the inputs to the neuron exceeds a threshold. One of the most challenging parts of designing neuromorphic hardware is handling the vast degree of connectivity that neurons have with each other in the form of synaptic connections. This paper analyzes the neuromorphic systems Neurogrid, Braindrop, SpiNNaker, BrainScaleS, TrueNorth, Loihi, Darwin, and Dynap-SEL; and discusses the design of large scale spiking communication networks used in such systems. In particular, this paper looks at how each of these systems solved the challenges of forming packets with spiking information and how these packets are routed within the system. The routing of packets is analyzed at two scales: How the packets should be routed when traveling a short distance, and how the packets should be routed over longer global connections. Additional topics, such as the use of asynchronous circuits, robustness in communication, connection with a host machine, and network synchronization are also covered.

AB - Multiple neuromorphic systems use spiking neural networks (SNNs) to perform computation in a way that is inspired by concepts learned about the human brain. SNNs are artificial networks made up of neurons that fire a pulse, or spike, once the accumulated value of the inputs to the neuron exceeds a threshold. One of the most challenging parts of designing neuromorphic hardware is handling the vast degree of connectivity that neurons have with each other in the form of synaptic connections. This paper analyzes the neuromorphic systems Neurogrid, Braindrop, SpiNNaker, BrainScaleS, TrueNorth, Loihi, Darwin, and Dynap-SEL; and discusses the design of large scale spiking communication networks used in such systems. In particular, this paper looks at how each of these systems solved the challenges of forming packets with spiking information and how these packets are routed within the system. The routing of packets is analyzed at two scales: How the packets should be routed when traveling a short distance, and how the packets should be routed over longer global connections. Additional topics, such as the use of asynchronous circuits, robustness in communication, connection with a host machine, and network synchronization are also covered.

KW - Communication protocols

KW - field programmable gate arrays

KW - interconnect

KW - network on chip

KW - neuromorphic

KW - spiking network communication

KW - spiking neural network

KW - very large scale integration

UR - http://www.scopus.com/inward/record.url?scp=85075236422&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2019.2941772

DO - 10.1109/ACCESS.2019.2941772

M3 - Article

AN - SCOPUS:85075236422

SN - 2169-3536

VL - 7

SP - 135606

EP - 135620

JO - IEEE Access

JF - IEEE Access

M1 - 8843969

ER -

A Review of spiking neuromorphic hardware communication systems

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this