Efficient pipelined architecture for competitive learning

Hui Ya Li, Chia Lung Hung, Wen-Jyi Hwang, Yi Tsan Hung

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This paper presents a novel pipelined architecture for competitive learning (CL). The architecture is implemented by the field programmable gate array (FPGA). It is used as a hardware accelerator in a system on programmable chip (SOPC) for reducing the computation time. In the architecture, a novel codeword swapping scheme is adopted so that neuron competitions for different training vectors can be operated concurrently. The neuron updating process is based on a hardware divider with simple table lookup operations. The divider performs finite precision calculations for area cost reduction at the expense of slight degradation in training performance. The CPU time of the NIOS processor executing the CL training with the proposed architecture as an accelerator is measured. Experimental results show that the NIOS processor with the proposed architecture as an accelerator can achieve up to a speedup of 254 over its software counterpart running on a general purpose processor Pentium IV without hardware support.

Original languageEnglish
Pages (from-to)236-244
Number of pages9
JournalJournal of Parallel and Distributed Computing
Volume71
Issue number2
DOIs
Publication statusPublished - 2011 Feb 1

Fingerprint

Competitive Learning
Particle accelerators
Hardware
Neurons
Program processors
Accelerator
Table lookup
Neuron
Cost reduction
Field programmable gate arrays (FPGA)
Hardware Accelerator
Look-up Table
Degradation
CPU Time
Field Programmable Gate Array
Updating
Speedup
Chip
Architecture
Software

Keywords

  • Clustering analysis
  • Competitive learning
  • FPGA
  • Reconfigurable computing
  • System on programmable chip
  • WTA

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computer Networks and Communications
  • Artificial Intelligence

Cite this

Efficient pipelined architecture for competitive learning. / Li, Hui Ya; Hung, Chia Lung; Hwang, Wen-Jyi; Hung, Yi Tsan.

In: Journal of Parallel and Distributed Computing, Vol. 71, No. 2, 01.02.2011, p. 236-244.

Research output: Contribution to journalArticle

Li, Hui Ya ; Hung, Chia Lung ; Hwang, Wen-Jyi ; Hung, Yi Tsan. / Efficient pipelined architecture for competitive learning. In: Journal of Parallel and Distributed Computing. 2011 ; Vol. 71, No. 2. pp. 236-244.
@article{27d2e8c57cf34b20905870babdb45024,
title = "Efficient pipelined architecture for competitive learning",
abstract = "This paper presents a novel pipelined architecture for competitive learning (CL). The architecture is implemented by the field programmable gate array (FPGA). It is used as a hardware accelerator in a system on programmable chip (SOPC) for reducing the computation time. In the architecture, a novel codeword swapping scheme is adopted so that neuron competitions for different training vectors can be operated concurrently. The neuron updating process is based on a hardware divider with simple table lookup operations. The divider performs finite precision calculations for area cost reduction at the expense of slight degradation in training performance. The CPU time of the NIOS processor executing the CL training with the proposed architecture as an accelerator is measured. Experimental results show that the NIOS processor with the proposed architecture as an accelerator can achieve up to a speedup of 254 over its software counterpart running on a general purpose processor Pentium IV without hardware support.",
keywords = "Clustering analysis, Competitive learning, FPGA, Reconfigurable computing, System on programmable chip, WTA",
author = "Li, {Hui Ya} and Hung, {Chia Lung} and Wen-Jyi Hwang and Hung, {Yi Tsan}",
year = "2011",
month = "2",
day = "1",
doi = "10.1016/j.jpdc.2010.09.001",
language = "English",
volume = "71",
pages = "236--244",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",
number = "2",

}

TY - JOUR

T1 - Efficient pipelined architecture for competitive learning

AU - Li, Hui Ya

AU - Hung, Chia Lung

AU - Hwang, Wen-Jyi

AU - Hung, Yi Tsan

PY - 2011/2/1

Y1 - 2011/2/1

N2 - This paper presents a novel pipelined architecture for competitive learning (CL). The architecture is implemented by the field programmable gate array (FPGA). It is used as a hardware accelerator in a system on programmable chip (SOPC) for reducing the computation time. In the architecture, a novel codeword swapping scheme is adopted so that neuron competitions for different training vectors can be operated concurrently. The neuron updating process is based on a hardware divider with simple table lookup operations. The divider performs finite precision calculations for area cost reduction at the expense of slight degradation in training performance. The CPU time of the NIOS processor executing the CL training with the proposed architecture as an accelerator is measured. Experimental results show that the NIOS processor with the proposed architecture as an accelerator can achieve up to a speedup of 254 over its software counterpart running on a general purpose processor Pentium IV without hardware support.

AB - This paper presents a novel pipelined architecture for competitive learning (CL). The architecture is implemented by the field programmable gate array (FPGA). It is used as a hardware accelerator in a system on programmable chip (SOPC) for reducing the computation time. In the architecture, a novel codeword swapping scheme is adopted so that neuron competitions for different training vectors can be operated concurrently. The neuron updating process is based on a hardware divider with simple table lookup operations. The divider performs finite precision calculations for area cost reduction at the expense of slight degradation in training performance. The CPU time of the NIOS processor executing the CL training with the proposed architecture as an accelerator is measured. Experimental results show that the NIOS processor with the proposed architecture as an accelerator can achieve up to a speedup of 254 over its software counterpart running on a general purpose processor Pentium IV without hardware support.

KW - Clustering analysis

KW - Competitive learning

KW - FPGA

KW - Reconfigurable computing

KW - System on programmable chip

KW - WTA

UR - http://www.scopus.com/inward/record.url?scp=78650416374&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78650416374&partnerID=8YFLogxK

U2 - 10.1016/j.jpdc.2010.09.001

DO - 10.1016/j.jpdc.2010.09.001

M3 - Article

AN - SCOPUS:78650416374

VL - 71

SP - 236

EP - 244

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 2

ER -