A framework for accelerating neuromorphic-vision algorithms on FPGAs

M. Debole, A. Al Maashri, M. Cotter, C. L. Yu, C. Chakrabarti, V. Narayanan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

Implementations of neuromorphic algorithms are traditionally implemented on platforms which consume significant power, falling short of their biologically underpinnings. Recent improvements in FPGA technology have led to FPGAs becoming a platform in which these rapidly evolving algorithms can be implemented. Unfortunately, implementing designs on FPGAs still prove challenging for nonexperts, limiting their use in the neuroscience domain. In this paper, a FPGA framework is presented which enables neuroscientists to compose multi-FPGA systems for a cortical object classification model. This is demonstrated by mapping this algorithm onto two distinct platforms providing speedups of up to 28X over a reference CPU implementation.

Original languageEnglish
Title of host publication2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011
Pages810-813
Number of pages4
DOIs
Publication statusPublished - 2011
Event2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011 - San Jose, CA, United States
Duration: Nov 7 2011Nov 10 2011

Other

Other2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011
CountryUnited States
CitySan Jose, CA
Period11/7/1111/10/11

Fingerprint

Field programmable gate arrays (FPGA)
Program processors

Keywords

  • FPGA application mapping
  • FPGA programming
  • Multi-FPGA partitioning
  • Neuromorphic vision algorithms

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Cite this

Debole, M., Maashri, A. A., Cotter, M., Yu, C. L., Chakrabarti, C., & Narayanan, V. (2011). A framework for accelerating neuromorphic-vision algorithms on FPGAs. In 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011 (pp. 810-813). [6105351] https://doi.org/10.1109/ICCAD.2011.6105351

A framework for accelerating neuromorphic-vision algorithms on FPGAs. / Debole, M.; Maashri, A. Al; Cotter, M.; Yu, C. L.; Chakrabarti, C.; Narayanan, V.

2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. p. 810-813 6105351.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Debole, M, Maashri, AA, Cotter, M, Yu, CL, Chakrabarti, C & Narayanan, V 2011, A framework for accelerating neuromorphic-vision algorithms on FPGAs. in 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011., 6105351, pp. 810-813, 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011, San Jose, CA, United States, 11/7/11. https://doi.org/10.1109/ICCAD.2011.6105351
Debole M, Maashri AA, Cotter M, Yu CL, Chakrabarti C, Narayanan V. A framework for accelerating neuromorphic-vision algorithms on FPGAs. In 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. p. 810-813. 6105351 https://doi.org/10.1109/ICCAD.2011.6105351
Debole, M. ; Maashri, A. Al ; Cotter, M. ; Yu, C. L. ; Chakrabarti, C. ; Narayanan, V. / A framework for accelerating neuromorphic-vision algorithms on FPGAs. 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. pp. 810-813
@inproceedings{cb8506f5d52d4bd4ad1ad363555ea6ee,
title = "A framework for accelerating neuromorphic-vision algorithms on FPGAs",
abstract = "Implementations of neuromorphic algorithms are traditionally implemented on platforms which consume significant power, falling short of their biologically underpinnings. Recent improvements in FPGA technology have led to FPGAs becoming a platform in which these rapidly evolving algorithms can be implemented. Unfortunately, implementing designs on FPGAs still prove challenging for nonexperts, limiting their use in the neuroscience domain. In this paper, a FPGA framework is presented which enables neuroscientists to compose multi-FPGA systems for a cortical object classification model. This is demonstrated by mapping this algorithm onto two distinct platforms providing speedups of up to 28X over a reference CPU implementation.",
keywords = "FPGA application mapping, FPGA programming, Multi-FPGA partitioning, Neuromorphic vision algorithms",
author = "M. Debole and Maashri, {A. Al} and M. Cotter and Yu, {C. L.} and C. Chakrabarti and V. Narayanan",
year = "2011",
doi = "10.1109/ICCAD.2011.6105351",
language = "English",
isbn = "9781457713989",
pages = "810--813",
booktitle = "2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011",

}

TY - GEN

T1 - A framework for accelerating neuromorphic-vision algorithms on FPGAs

AU - Debole, M.

AU - Maashri, A. Al

AU - Cotter, M.

AU - Yu, C. L.

AU - Chakrabarti, C.

AU - Narayanan, V.

PY - 2011

Y1 - 2011

N2 - Implementations of neuromorphic algorithms are traditionally implemented on platforms which consume significant power, falling short of their biologically underpinnings. Recent improvements in FPGA technology have led to FPGAs becoming a platform in which these rapidly evolving algorithms can be implemented. Unfortunately, implementing designs on FPGAs still prove challenging for nonexperts, limiting their use in the neuroscience domain. In this paper, a FPGA framework is presented which enables neuroscientists to compose multi-FPGA systems for a cortical object classification model. This is demonstrated by mapping this algorithm onto two distinct platforms providing speedups of up to 28X over a reference CPU implementation.

AB - Implementations of neuromorphic algorithms are traditionally implemented on platforms which consume significant power, falling short of their biologically underpinnings. Recent improvements in FPGA technology have led to FPGAs becoming a platform in which these rapidly evolving algorithms can be implemented. Unfortunately, implementing designs on FPGAs still prove challenging for nonexperts, limiting their use in the neuroscience domain. In this paper, a FPGA framework is presented which enables neuroscientists to compose multi-FPGA systems for a cortical object classification model. This is demonstrated by mapping this algorithm onto two distinct platforms providing speedups of up to 28X over a reference CPU implementation.

KW - FPGA application mapping

KW - FPGA programming

KW - Multi-FPGA partitioning

KW - Neuromorphic vision algorithms

UR - http://www.scopus.com/inward/record.url?scp=84855778958&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84855778958&partnerID=8YFLogxK

U2 - 10.1109/ICCAD.2011.6105351

DO - 10.1109/ICCAD.2011.6105351

M3 - Conference contribution

SN - 9781457713989

SP - 810

EP - 813

BT - 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011

ER -