CNN model on stream processing platform

László Füredi, P. Szolgay

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

2 Citations (Scopus)

Abstract

Stream processor, Inc. (SPI) has introduced the Storm-1 stream processing System-on-a-Chip (SoC) which contains 16 data-parallel processing arrays with five Very Long Instruction Word (VLIW[1]) Arithmetic Logic Units (ALUs) in each lane (altogether 80 processing units). In the paper the implementation of emulated-digital Cellular Neural Networks Universal Machine (CNN-UM) simulation kernel on the Storm-1 Stream Processing Platform (SPP) is described and optimized according to the special requirements of SPP. The area/speed/power tradeoffs of our solution and other emulated digital CNN implementations are also compared.

Original languageEnglish
Title of host publicationECCTD 2009 - European Conference on Circuit Theory and Design Conference Program
Pages843-846
Number of pages4
DOIs
Publication statusPublished - 2009
EventECCTD 2009 - European Conference on Circuit Theory and Design Conference Program - Antalya, Turkey
Duration: Aug 23 2009Aug 27 2009

Other

OtherECCTD 2009 - European Conference on Circuit Theory and Design Conference Program
CountryTurkey
CityAntalya
Period8/23/098/27/09

Fingerprint

Processing
Array processing
Cellular neural networks

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

Füredi, L., & Szolgay, P. (2009). CNN model on stream processing platform. In ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program (pp. 843-846). [5275115] https://doi.org/10.1109/ECCTD.2009.5275115

CNN model on stream processing platform. / Füredi, László; Szolgay, P.

ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program. 2009. p. 843-846 5275115.

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

Füredi, L & Szolgay, P 2009, CNN model on stream processing platform. in ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program., 5275115, pp. 843-846, ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program, Antalya, Turkey, 8/23/09. https://doi.org/10.1109/ECCTD.2009.5275115
Füredi L, Szolgay P. CNN model on stream processing platform. In ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program. 2009. p. 843-846. 5275115 https://doi.org/10.1109/ECCTD.2009.5275115
Füredi, László ; Szolgay, P. / CNN model on stream processing platform. ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program. 2009. pp. 843-846
@inproceedings{4f372b1a00564349afa9cf0539212415,
title = "CNN model on stream processing platform",
abstract = "Stream processor, Inc. (SPI) has introduced the Storm-1 stream processing System-on-a-Chip (SoC) which contains 16 data-parallel processing arrays with five Very Long Instruction Word (VLIW[1]) Arithmetic Logic Units (ALUs) in each lane (altogether 80 processing units). In the paper the implementation of emulated-digital Cellular Neural Networks Universal Machine (CNN-UM) simulation kernel on the Storm-1 Stream Processing Platform (SPP) is described and optimized according to the special requirements of SPP. The area/speed/power tradeoffs of our solution and other emulated digital CNN implementations are also compared.",
author = "L{\'a}szl{\'o} F{\"u}redi and P. Szolgay",
year = "2009",
doi = "10.1109/ECCTD.2009.5275115",
language = "English",
isbn = "9781424438969",
pages = "843--846",
booktitle = "ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program",

}

TY - GEN

T1 - CNN model on stream processing platform

AU - Füredi, László

AU - Szolgay, P.

PY - 2009

Y1 - 2009

N2 - Stream processor, Inc. (SPI) has introduced the Storm-1 stream processing System-on-a-Chip (SoC) which contains 16 data-parallel processing arrays with five Very Long Instruction Word (VLIW[1]) Arithmetic Logic Units (ALUs) in each lane (altogether 80 processing units). In the paper the implementation of emulated-digital Cellular Neural Networks Universal Machine (CNN-UM) simulation kernel on the Storm-1 Stream Processing Platform (SPP) is described and optimized according to the special requirements of SPP. The area/speed/power tradeoffs of our solution and other emulated digital CNN implementations are also compared.

AB - Stream processor, Inc. (SPI) has introduced the Storm-1 stream processing System-on-a-Chip (SoC) which contains 16 data-parallel processing arrays with five Very Long Instruction Word (VLIW[1]) Arithmetic Logic Units (ALUs) in each lane (altogether 80 processing units). In the paper the implementation of emulated-digital Cellular Neural Networks Universal Machine (CNN-UM) simulation kernel on the Storm-1 Stream Processing Platform (SPP) is described and optimized according to the special requirements of SPP. The area/speed/power tradeoffs of our solution and other emulated digital CNN implementations are also compared.

UR - http://www.scopus.com/inward/record.url?scp=71249108957&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=71249108957&partnerID=8YFLogxK

U2 - 10.1109/ECCTD.2009.5275115

DO - 10.1109/ECCTD.2009.5275115

M3 - Conference contribution

AN - SCOPUS:71249108957

SN - 9781424438969

SP - 843

EP - 846

BT - ECCTD 2009 - European Conference on Circuit Theory and Design Conference Program

ER -