Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery

Andreas G. Andreou; Tomas Figliolia; Kayode Sanni; Thomas S. Murray; Gaspar Tognetti; Daniel R. Mendat; Jamal L. Molin; Martin Villemur; Philippe O. Pouliquen; Pedro Julian; Ralph Etienne-Cummings; Isidoros Doxas

doi:10.1109/LASCAS.2016.7450995

Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery

Andreas G. Andreou, Tomas Figliolia, Kayode Sanni, Thomas S. Murray, Gaspar Tognetti, Daniel R. Mendat, Jamal L. Molin, Martin Villemur, Philippe O. Pouliquen, Pedro Julian, Ralph Etienne-Cummings, Isidoros Doxas

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

Abstract

In this paper we discuss a brain-inspired system architecture for real-time big velocity BIGDATA processing that originates in large format tiled imaging arrays used in wide area motion imagery ubiquitous surveillance. High performance and high throughput is achieved through approximate computing and fixed point arithmetic in a variable precision (6 bits to 18 bits) architecture. The architecture implements a variety of processing algorithms classes ranging from convolutional networks (Con-vNets) to linear and non-linear morphological processing, probabilistic inference using exact and approximate Bayesian methods and ConvNet based classification. The processing pipeline is implemented entirely using event based neuromorphic and stochastic computational primitives. The system is capable of processing in real-time 160 × 120 raw pixel data running on a reconfigurable computing platform (5 Xilinx Kintex-7 FPGAs). The reconfigurable computing implementation was developed to emulate the computational structures for a 3D System on Chip (3D-SOC) that will be fabricated in the 55nm CMOS technology and it has a dual goal: (i) algorithm exploration and (ii) architecture exploration.

Original language	English
Title of host publication	LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference
Publisher	IEEE Computer Society
Pages	1-6
Number of pages	6
ISBN (Print)	9781467378352
DOIs	https://doi.org/10.1109/LASCAS.2016.7450995
Publication status	Published - 11 Apr 2016
Externally published	Yes
Event	2016 IEEE 7th Latin American Symposium on Circuits & Systems (LASCAS) - Florianopolis, Brazil Duration: 28 Feb 2016 → 2 Mar 2016

Publication series

Name	LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference

Conference

Conference	2016 IEEE 7th Latin American Symposium on Circuits & Systems (LASCAS)
Period	28/02/16 → 2/03/16

Keywords

Computer architecture
Pipelines
Cameras
Real-time systems
Inference algorithms
Field programmable gate arrays
CMOS integrated circuits

Access to Document

10.1109/LASCAS.2016.7450995

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7450995

Cite this

Andreou, A. G., Figliolia, T., Sanni, K., Murray, T. S., Tognetti, G., Mendat, D. R., Molin, J. L., Villemur, M., Pouliquen, P. O., Julian, P., Etienne-Cummings, R., & Doxas, I. (2016). Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery. In LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference (pp. 1-6). Article 7450995 (LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference). IEEE Computer Society. https://doi.org/10.1109/LASCAS.2016.7450995

Andreou, Andreas G. ; Figliolia, Tomas ; Sanni, Kayode et al. / Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery. LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference. IEEE Computer Society, 2016. pp. 1-6 (LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference).

@inproceedings{6efdb5a5073c4534a5b52675b00f83e8,

title = "Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery",

abstract = "In this paper we discuss a brain-inspired system architecture for real-time big velocity BIGDATA processing that originates in large format tiled imaging arrays used in wide area motion imagery ubiquitous surveillance. High performance and high throughput is achieved through approximate computing and fixed point arithmetic in a variable precision (6 bits to 18 bits) architecture. The architecture implements a variety of processing algorithms classes ranging from convolutional networks (Con-vNets) to linear and non-linear morphological processing, probabilistic inference using exact and approximate Bayesian methods and ConvNet based classification. The processing pipeline is implemented entirely using event based neuromorphic and stochastic computational primitives. The system is capable of processing in real-time 160 × 120 raw pixel data running on a reconfigurable computing platform (5 Xilinx Kintex-7 FPGAs). The reconfigurable computing implementation was developed to emulate the computational structures for a 3D System on Chip (3D-SOC) that will be fabricated in the 55nm CMOS technology and it has a dual goal: (i) algorithm exploration and (ii) architecture exploration.",

keywords = "Computer architecture, Pipelines, Cameras, Real-time systems, Inference algorithms, Field programmable gate arrays, CMOS integrated circuits",

author = "Andreou, {Andreas G.} and Tomas Figliolia and Kayode Sanni and Murray, {Thomas S.} and Gaspar Tognetti and Mendat, {Daniel R.} and Molin, {Jamal L.} and Martin Villemur and Pouliquen, {Philippe O.} and Pedro Julian and Ralph Etienne-Cummings and Isidoros Doxas",

year = "2016",

month = apr,

day = "11",

doi = "10.1109/LASCAS.2016.7450995",

language = "English",

isbn = "9781467378352",

series = "LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference",

publisher = "IEEE Computer Society",

pages = "1--6",

booktitle = "LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference",

address = "United States",

note = "2016 IEEE 7th Latin American Symposium on Circuits & Systems (LASCAS) ; Conference date: 28-02-2016 Through 02-03-2016",

}

Andreou, AG, Figliolia, T, Sanni, K, Murray, TS, Tognetti, G, Mendat, DR, Molin, JL, Villemur, M, Pouliquen, PO, Julian, P, Etienne-Cummings, R & Doxas, I 2016, Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery. in LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference., 7450995, LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference, IEEE Computer Society, pp. 1-6, 2016 IEEE 7th Latin American Symposium on Circuits & Systems (LASCAS), 28/02/16. https://doi.org/10.1109/LASCAS.2016.7450995

Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery. / Andreou, Andreas G.; Figliolia, Tomas; Sanni, Kayode et al.
LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference. IEEE Computer Society, 2016. p. 1-6 7450995 (LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference).

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

TY - GEN

T1 - Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery

AU - Andreou, Andreas G.

AU - Figliolia, Tomas

AU - Sanni, Kayode

AU - Murray, Thomas S.

AU - Tognetti, Gaspar

AU - Mendat, Daniel R.

AU - Molin, Jamal L.

AU - Villemur, Martin

AU - Pouliquen, Philippe O.

AU - Julian, Pedro

AU - Etienne-Cummings, Ralph

AU - Doxas, Isidoros

PY - 2016/4/11

Y1 - 2016/4/11

N2 - In this paper we discuss a brain-inspired system architecture for real-time big velocity BIGDATA processing that originates in large format tiled imaging arrays used in wide area motion imagery ubiquitous surveillance. High performance and high throughput is achieved through approximate computing and fixed point arithmetic in a variable precision (6 bits to 18 bits) architecture. The architecture implements a variety of processing algorithms classes ranging from convolutional networks (Con-vNets) to linear and non-linear morphological processing, probabilistic inference using exact and approximate Bayesian methods and ConvNet based classification. The processing pipeline is implemented entirely using event based neuromorphic and stochastic computational primitives. The system is capable of processing in real-time 160 × 120 raw pixel data running on a reconfigurable computing platform (5 Xilinx Kintex-7 FPGAs). The reconfigurable computing implementation was developed to emulate the computational structures for a 3D System on Chip (3D-SOC) that will be fabricated in the 55nm CMOS technology and it has a dual goal: (i) algorithm exploration and (ii) architecture exploration.

AB - In this paper we discuss a brain-inspired system architecture for real-time big velocity BIGDATA processing that originates in large format tiled imaging arrays used in wide area motion imagery ubiquitous surveillance. High performance and high throughput is achieved through approximate computing and fixed point arithmetic in a variable precision (6 bits to 18 bits) architecture. The architecture implements a variety of processing algorithms classes ranging from convolutional networks (Con-vNets) to linear and non-linear morphological processing, probabilistic inference using exact and approximate Bayesian methods and ConvNet based classification. The processing pipeline is implemented entirely using event based neuromorphic and stochastic computational primitives. The system is capable of processing in real-time 160 × 120 raw pixel data running on a reconfigurable computing platform (5 Xilinx Kintex-7 FPGAs). The reconfigurable computing implementation was developed to emulate the computational structures for a 3D System on Chip (3D-SOC) that will be fabricated in the 55nm CMOS technology and it has a dual goal: (i) algorithm exploration and (ii) architecture exploration.

KW - Computer architecture

KW - Pipelines

KW - Cameras

KW - Real-time systems

KW - Inference algorithms

KW - Field programmable gate arrays

KW - CMOS integrated circuits

UR - https://ieeexplore.ieee.org/document/7450995/

U2 - 10.1109/LASCAS.2016.7450995

DO - 10.1109/LASCAS.2016.7450995

M3 - Conference Paper

SN - 9781467378352

T3 - LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference

SP - 1

EP - 6

BT - LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference

PB - IEEE Computer Society

T2 - 2016 IEEE 7th Latin American Symposium on Circuits & Systems (LASCAS)

Y2 - 28 February 2016 through 2 March 2016

ER -

Andreou AG, Figliolia T, Sanni K, Murray TS, Tognetti G, Mendat DR et al. Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery. In LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference. IEEE Computer Society. 2016. p. 1-6. 7450995. (LASCAS 2016 - 7th IEEE Latin American Symposium on Circuits and Systems, R9 IEEE CASS Flagship Conference). doi: 10.1109/LASCAS.2016.7450995

Bio-inspired system architecture for energy efficient, BIGDATA computing with application to wide area motion imagery

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this