Proteus: A reconfigurable computational network for computer vision

Robert M. Haralick, Arun K. Somani, Craig Wittenbrink, Robert Johnson, Kenneth Cooper, Linda G. Shapiro, Ihsin T. Phillips, Jenq Neng Hwang, William Cheung, Yung Hsi Yao, Chung-Ho Chen, Larry Yang, Brian Daugherty, Bob Lorbeski, Kent Loving, Tom Miller, Larye Parkins, Steve Soos

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

The Proteus architecture is a highly parallel, multiple instruction, multiple data machine (MIMD) optimized for large granularity tasks such as machine vision and image processing. The system can achieve 20 gigaflops (80 gigaflops peak). It accepts data via multiple serial links at a rate of up to 640 MB/S. The system employs a hierarchical reconfigurable interconnection network with the highest level being a circuit-switched enhanced hypercube, serial interconnection network for internal data transfers. The system is designed to use 256 to 1024 RISC processors. The processors use 1-MB external read/write allocating caches for reduced multiprocessor contention. The system detects, locates, and replaces faulty subsystems using redundant hardware to facilitate fault tolerance. The parallelism is directly controllable through an advanced software system for partitioning, scheduling, and development. System software includes a translator for the INSIGHT language, a parallel debugger, lowand high-level simulators, and a message-passing system for all control needs. Image-processing application software includes a variety of point operators, neighborhood operators, convolution, and the mathematical morphology operations of binary and gray-scale dilation, erosion, opening, and closing.

Original languageEnglish
Pages (from-to)85-100
Number of pages16
JournalMachine Vision and Applications
Volume8
Issue number2
DOIs
Publication statusPublished - 1995 Mar 1

Fingerprint

Computer vision
Mathematical operators
Image processing
Hypercube networks
Reduced instruction set computing
Mathematical morphology
Interconnection networks (circuit switching)
Message passing
Data transfer
Fault tolerance
Convolution
Application programs
Erosion
Computer systems
Simulators
Scheduling
Hardware
Networks (circuits)

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Haralick, R. M., Somani, A. K., Wittenbrink, C., Johnson, R., Cooper, K., Shapiro, L. G., ... Soos, S. (1995). Proteus: A reconfigurable computational network for computer vision. Machine Vision and Applications, 8(2), 85-100. https://doi.org/10.1007/BF01213474
Haralick, Robert M. ; Somani, Arun K. ; Wittenbrink, Craig ; Johnson, Robert ; Cooper, Kenneth ; Shapiro, Linda G. ; Phillips, Ihsin T. ; Hwang, Jenq Neng ; Cheung, William ; Yao, Yung Hsi ; Chen, Chung-Ho ; Yang, Larry ; Daugherty, Brian ; Lorbeski, Bob ; Loving, Kent ; Miller, Tom ; Parkins, Larye ; Soos, Steve. / Proteus : A reconfigurable computational network for computer vision. In: Machine Vision and Applications. 1995 ; Vol. 8, No. 2. pp. 85-100.
@article{33eb69e8998d43eca9bb1b3d69cfbed6,
title = "Proteus: A reconfigurable computational network for computer vision",
abstract = "The Proteus architecture is a highly parallel, multiple instruction, multiple data machine (MIMD) optimized for large granularity tasks such as machine vision and image processing. The system can achieve 20 gigaflops (80 gigaflops peak). It accepts data via multiple serial links at a rate of up to 640 MB/S. The system employs a hierarchical reconfigurable interconnection network with the highest level being a circuit-switched enhanced hypercube, serial interconnection network for internal data transfers. The system is designed to use 256 to 1024 RISC processors. The processors use 1-MB external read/write allocating caches for reduced multiprocessor contention. The system detects, locates, and replaces faulty subsystems using redundant hardware to facilitate fault tolerance. The parallelism is directly controllable through an advanced software system for partitioning, scheduling, and development. System software includes a translator for the INSIGHT language, a parallel debugger, lowand high-level simulators, and a message-passing system for all control needs. Image-processing application software includes a variety of point operators, neighborhood operators, convolution, and the mathematical morphology operations of binary and gray-scale dilation, erosion, opening, and closing.",
author = "Haralick, {Robert M.} and Somani, {Arun K.} and Craig Wittenbrink and Robert Johnson and Kenneth Cooper and Shapiro, {Linda G.} and Phillips, {Ihsin T.} and Hwang, {Jenq Neng} and William Cheung and Yao, {Yung Hsi} and Chung-Ho Chen and Larry Yang and Brian Daugherty and Bob Lorbeski and Kent Loving and Tom Miller and Larye Parkins and Steve Soos",
year = "1995",
month = "3",
day = "1",
doi = "10.1007/BF01213474",
language = "English",
volume = "8",
pages = "85--100",
journal = "Machine Vision and Applications",
issn = "0932-8092",
publisher = "Springer Verlag",
number = "2",

}

Haralick, RM, Somani, AK, Wittenbrink, C, Johnson, R, Cooper, K, Shapiro, LG, Phillips, IT, Hwang, JN, Cheung, W, Yao, YH, Chen, C-H, Yang, L, Daugherty, B, Lorbeski, B, Loving, K, Miller, T, Parkins, L & Soos, S 1995, 'Proteus: A reconfigurable computational network for computer vision', Machine Vision and Applications, vol. 8, no. 2, pp. 85-100. https://doi.org/10.1007/BF01213474

Proteus : A reconfigurable computational network for computer vision. / Haralick, Robert M.; Somani, Arun K.; Wittenbrink, Craig; Johnson, Robert; Cooper, Kenneth; Shapiro, Linda G.; Phillips, Ihsin T.; Hwang, Jenq Neng; Cheung, William; Yao, Yung Hsi; Chen, Chung-Ho; Yang, Larry; Daugherty, Brian; Lorbeski, Bob; Loving, Kent; Miller, Tom; Parkins, Larye; Soos, Steve.

In: Machine Vision and Applications, Vol. 8, No. 2, 01.03.1995, p. 85-100.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Proteus

T2 - A reconfigurable computational network for computer vision

AU - Haralick, Robert M.

AU - Somani, Arun K.

AU - Wittenbrink, Craig

AU - Johnson, Robert

AU - Cooper, Kenneth

AU - Shapiro, Linda G.

AU - Phillips, Ihsin T.

AU - Hwang, Jenq Neng

AU - Cheung, William

AU - Yao, Yung Hsi

AU - Chen, Chung-Ho

AU - Yang, Larry

AU - Daugherty, Brian

AU - Lorbeski, Bob

AU - Loving, Kent

AU - Miller, Tom

AU - Parkins, Larye

AU - Soos, Steve

PY - 1995/3/1

Y1 - 1995/3/1

N2 - The Proteus architecture is a highly parallel, multiple instruction, multiple data machine (MIMD) optimized for large granularity tasks such as machine vision and image processing. The system can achieve 20 gigaflops (80 gigaflops peak). It accepts data via multiple serial links at a rate of up to 640 MB/S. The system employs a hierarchical reconfigurable interconnection network with the highest level being a circuit-switched enhanced hypercube, serial interconnection network for internal data transfers. The system is designed to use 256 to 1024 RISC processors. The processors use 1-MB external read/write allocating caches for reduced multiprocessor contention. The system detects, locates, and replaces faulty subsystems using redundant hardware to facilitate fault tolerance. The parallelism is directly controllable through an advanced software system for partitioning, scheduling, and development. System software includes a translator for the INSIGHT language, a parallel debugger, lowand high-level simulators, and a message-passing system for all control needs. Image-processing application software includes a variety of point operators, neighborhood operators, convolution, and the mathematical morphology operations of binary and gray-scale dilation, erosion, opening, and closing.

AB - The Proteus architecture is a highly parallel, multiple instruction, multiple data machine (MIMD) optimized for large granularity tasks such as machine vision and image processing. The system can achieve 20 gigaflops (80 gigaflops peak). It accepts data via multiple serial links at a rate of up to 640 MB/S. The system employs a hierarchical reconfigurable interconnection network with the highest level being a circuit-switched enhanced hypercube, serial interconnection network for internal data transfers. The system is designed to use 256 to 1024 RISC processors. The processors use 1-MB external read/write allocating caches for reduced multiprocessor contention. The system detects, locates, and replaces faulty subsystems using redundant hardware to facilitate fault tolerance. The parallelism is directly controllable through an advanced software system for partitioning, scheduling, and development. System software includes a translator for the INSIGHT language, a parallel debugger, lowand high-level simulators, and a message-passing system for all control needs. Image-processing application software includes a variety of point operators, neighborhood operators, convolution, and the mathematical morphology operations of binary and gray-scale dilation, erosion, opening, and closing.

UR - http://www.scopus.com/inward/record.url?scp=0029230434&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029230434&partnerID=8YFLogxK

U2 - 10.1007/BF01213474

DO - 10.1007/BF01213474

M3 - Article

AN - SCOPUS:0029230434

VL - 8

SP - 85

EP - 100

JO - Machine Vision and Applications

JF - Machine Vision and Applications

SN - 0932-8092

IS - 2

ER -

Haralick RM, Somani AK, Wittenbrink C, Johnson R, Cooper K, Shapiro LG et al. Proteus: A reconfigurable computational network for computer vision. Machine Vision and Applications. 1995 Mar 1;8(2):85-100. https://doi.org/10.1007/BF01213474