Event-driven simulation of arbitrary spiking neural networks on SpiNNaker

Thomas Sharp; Luis A. Plana; Francesco Galluppi; Steve Furber

doi:10.1007/978-3-642-24965-5_48

Event-driven simulation of arbitrary spiking neural networks on SpiNNaker

Thomas Sharp, Luis A. Plana, Francesco Galluppi, Steve Furber

Advanced Processor Technology

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

Abstract

Programming supercomputers correctly and optimally is non-trivial, which presents a problem for scientists simulating large areas of the brain. Researchers face the challenges of learning how to fully exploit hardware whilst avoiding the numerous pitfalls of parallel programming such as race conditions, deadlock and poor scaling. The SpiNNaker architecture is designed to exploit up to a million processors in modelling as many as one billion neurons in real-time. We present a programming interface for the architecture to allow modelling of arbitrary neuron and synapse dynamics using standard sequential C code, without concern for parallel-programming techniques or interprocessor communication mechanisms. An example is presented in which SpiNNaker is programmed to model multiple synaptic dynamics that are exchanged on the fly and the results of the different synaptic efficacies are shown. © 2011 Springer-Verlag.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)\|Lect. Notes Comput. Sci.
Place of Publication	Berlin / Heidelberg
Publisher	Springer Nature
Pages	424-430
Number of pages	6
Volume	7064
ISBN (Print)	9783642249648
DOIs	https://doi.org/10.1007/978-3-642-24965-5_48
Publication status	Published - 2011
Event	18th International Conference on Neural Information Processing, ICONIP 2011 - Shanghai Duration: 1 Jul 2011 → …

Conference

Conference	18th International Conference on Neural Information Processing, ICONIP 2011
City	Shanghai
Period	1/07/11 → …

Keywords

C
callbacks
event driven
kernel
neural simulation
parallel programming
SpiNNaker
tasks

Access to Document

10.1007/978-3-642-24965-5_48

Cite this

Sharp, T., Plana, L. A., Galluppi, F., & Furber, S. (2011). Event-driven simulation of arbitrary spiking neural networks on SpiNNaker. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci. (Vol. 7064, pp. 424-430). Springer Nature. https://doi.org/10.1007/978-3-642-24965-5_48

@inproceedings{3bfe6e1661b14a0eb3005daa132f912c,

title = "Event-driven simulation of arbitrary spiking neural networks on SpiNNaker",

abstract = "Programming supercomputers correctly and optimally is non-trivial, which presents a problem for scientists simulating large areas of the brain. Researchers face the challenges of learning how to fully exploit hardware whilst avoiding the numerous pitfalls of parallel programming such as race conditions, deadlock and poor scaling. The SpiNNaker architecture is designed to exploit up to a million processors in modelling as many as one billion neurons in real-time. We present a programming interface for the architecture to allow modelling of arbitrary neuron and synapse dynamics using standard sequential C code, without concern for parallel-programming techniques or interprocessor communication mechanisms. An example is presented in which SpiNNaker is programmed to model multiple synaptic dynamics that are exchanged on the fly and the results of the different synaptic efficacies are shown. {\textcopyright} 2011 Springer-Verlag.",

keywords = "C, callbacks, event driven, kernel, neural simulation, parallel programming, SpiNNaker, tasks",

author = "Thomas Sharp and Plana, {Luis A.} and Francesco Galluppi and Steve Furber",

year = "2011",

doi = "10.1007/978-3-642-24965-5_48",

language = "English",

isbn = "9783642249648",

volume = "7064",

pages = "424--430",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.",

publisher = "Springer Nature",

address = "United States",

note = "18th International Conference on Neural Information Processing, ICONIP 2011 ; Conference date: 01-07-2011",

}

Sharp, T, Plana, LA, Galluppi, F & Furber, S 2011, Event-driven simulation of arbitrary spiking neural networks on SpiNNaker. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. vol. 7064, Springer Nature, Berlin / Heidelberg, pp. 424-430, 18th International Conference on Neural Information Processing, ICONIP 2011, Shanghai, 1/07/11. https://doi.org/10.1007/978-3-642-24965-5_48

Event-driven simulation of arbitrary spiking neural networks on SpiNNaker. / Sharp, Thomas; Plana, Luis A.; Galluppi, Francesco et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. Vol. 7064 Berlin / Heidelberg: Springer Nature, 2011. p. 424-430.

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Event-driven simulation of arbitrary spiking neural networks on SpiNNaker

AU - Sharp, Thomas

AU - Plana, Luis A.

AU - Galluppi, Francesco

AU - Furber, Steve

PY - 2011

Y1 - 2011

N2 - Programming supercomputers correctly and optimally is non-trivial, which presents a problem for scientists simulating large areas of the brain. Researchers face the challenges of learning how to fully exploit hardware whilst avoiding the numerous pitfalls of parallel programming such as race conditions, deadlock and poor scaling. The SpiNNaker architecture is designed to exploit up to a million processors in modelling as many as one billion neurons in real-time. We present a programming interface for the architecture to allow modelling of arbitrary neuron and synapse dynamics using standard sequential C code, without concern for parallel-programming techniques or interprocessor communication mechanisms. An example is presented in which SpiNNaker is programmed to model multiple synaptic dynamics that are exchanged on the fly and the results of the different synaptic efficacies are shown. © 2011 Springer-Verlag.

AB - Programming supercomputers correctly and optimally is non-trivial, which presents a problem for scientists simulating large areas of the brain. Researchers face the challenges of learning how to fully exploit hardware whilst avoiding the numerous pitfalls of parallel programming such as race conditions, deadlock and poor scaling. The SpiNNaker architecture is designed to exploit up to a million processors in modelling as many as one billion neurons in real-time. We present a programming interface for the architecture to allow modelling of arbitrary neuron and synapse dynamics using standard sequential C code, without concern for parallel-programming techniques or interprocessor communication mechanisms. An example is presented in which SpiNNaker is programmed to model multiple synaptic dynamics that are exchanged on the fly and the results of the different synaptic efficacies are shown. © 2011 Springer-Verlag.

KW - C

KW - callbacks

KW - event driven

KW - kernel

KW - neural simulation

KW - parallel programming

KW - SpiNNaker

KW - tasks

U2 - 10.1007/978-3-642-24965-5_48

DO - 10.1007/978-3-642-24965-5_48

M3 - Conference contribution

SN - 9783642249648

VL - 7064

SP - 424

EP - 430

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.

PB - Springer Nature

CY - Berlin / Heidelberg

T2 - 18th International Conference on Neural Information Processing, ICONIP 2011

Y2 - 1 July 2011

ER -

Sharp T, Plana LA, Galluppi F, Furber S. Event-driven simulation of arbitrary spiking neural networks on SpiNNaker. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. Vol. 7064. Berlin / Heidelberg: Springer Nature. 2011. p. 424-430 doi: 10.1007/978-3-642-24965-5_48

Event-driven simulation of arbitrary spiking neural networks on SpiNNaker

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this