An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures

Jun Ma; K. K. Parhi; E. F. Deprettere

doi:10.1109/ACSSC.1999.831936

An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures

Jun Ma, K. K. Parhi, E. F. Deprettere

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

5 Scopus citations

Abstract

Singular value decomposition (SVD) has become a standard linear algebra tool in modern digital signal processing. CORDIC based SVD algorithms are among the most popular SVD algorithms which exhibit good numerical properties. The speed of the sequential algorithms is however limited by the recursive feedback loops in the underlying signal flow graph. The critical loop computation time is proportional to the size of the problem which prohibits pipelined processing. This paper addresses the derivation of parallel architectures for SVD updating algorithms. An algorithm transformation approach based on re-Timing and matrix associativity is presented to drive parallel SVD updating architectures. These architectures have critical loop computation time independent of the problem size and are suitable for CORDIC arithmetic based VLSI implementations.

Original language	English (US)
Title of host publication	Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers
Editors	Michael B. Matthews
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1401-1405
Number of pages	5
ISBN (Electronic)	0780357000, 9780780357006
DOIs	https://doi.org/10.1109/ACSSC.1999.831936
State	Published - 1999
Externally published	Yes
Event	33rd Asilomar Conference on Signals, Systems, and Computers, ACSSC 1999 - Pacific Grove, United States Duration: Oct 24 1999 → Oct 27 1999

Publication series

Name	Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers
Volume	2

Other

Other	33rd Asilomar Conference on Signals, Systems, and Computers, ACSSC 1999
Country/Territory	United States
City	Pacific Grove
Period	10/24/99 → 10/27/99

Bibliographical note

Publisher Copyright:
© 1999 IEEE.

Access

10.1109/ACSSC.1999.831936

OpenUrl availability

Full text

Cite this

Ma, J., Parhi, K. K., & Deprettere, E. F. (1999). An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures. In M. B. Matthews (Ed.), Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers (pp. 1401-1405). Article 831936 (Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers; Vol. 2). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ACSSC.1999.831936

An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures. / Ma, Jun; Parhi, K. K.; Deprettere, E. F.
Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers. ed. / Michael B. Matthews. Institute of Electrical and Electronics Engineers Inc., 1999. p. 1401-1405 831936 (Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers; Vol. 2).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Ma, J, Parhi, KK & Deprettere, EF 1999, An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures. in MB Matthews (ed.), Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers., 831936, Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers, vol. 2, Institute of Electrical and Electronics Engineers Inc., pp. 1401-1405, 33rd Asilomar Conference on Signals, Systems, and Computers, ACSSC 1999, Pacific Grove, United States, 10/24/99. https://doi.org/10.1109/ACSSC.1999.831936

Ma J, Parhi KK, Deprettere EF. An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures. In Matthews MB, editor, Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers. Institute of Electrical and Electronics Engineers Inc. 1999. p. 1401-1405. 831936. (Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers). doi: 10.1109/ACSSC.1999.831936

Ma, Jun ; Parhi, K. K. ; Deprettere, E. F. / An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures. Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers. editor / Michael B. Matthews. Institute of Electrical and Electronics Engineers Inc., 1999. pp. 1401-1405 (Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers).

@inproceedings{6a5d3633792c4dbeab23fd81c3b2b8ac,

title = "An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures",

abstract = "Singular value decomposition (SVD) has become a standard linear algebra tool in modern digital signal processing. CORDIC based SVD algorithms are among the most popular SVD algorithms which exhibit good numerical properties. The speed of the sequential algorithms is however limited by the recursive feedback loops in the underlying signal flow graph. The critical loop computation time is proportional to the size of the problem which prohibits pipelined processing. This paper addresses the derivation of parallel architectures for SVD updating algorithms. An algorithm transformation approach based on re-Timing and matrix associativity is presented to drive parallel SVD updating architectures. These architectures have critical loop computation time independent of the problem size and are suitable for CORDIC arithmetic based VLSI implementations.",

author = "Jun Ma and Parhi, {K. K.} and Deprettere, {E. F.}",

note = "Publisher Copyright: {\textcopyright} 1999 IEEE.; 33rd Asilomar Conference on Signals, Systems, and Computers, ACSSC 1999 ; Conference date: 24-10-1999 Through 27-10-1999",

year = "1999",

doi = "10.1109/ACSSC.1999.831936",

language = "English (US)",

series = "Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1401--1405",

editor = "Matthews, {Michael B.}",

booktitle = "Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers",

}

TY - GEN

T1 - An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures

AU - Ma, Jun

AU - Parhi, K. K.

AU - Deprettere, E. F.

PY - 1999

Y1 - 1999

N2 - Singular value decomposition (SVD) has become a standard linear algebra tool in modern digital signal processing. CORDIC based SVD algorithms are among the most popular SVD algorithms which exhibit good numerical properties. The speed of the sequential algorithms is however limited by the recursive feedback loops in the underlying signal flow graph. The critical loop computation time is proportional to the size of the problem which prohibits pipelined processing. This paper addresses the derivation of parallel architectures for SVD updating algorithms. An algorithm transformation approach based on re-Timing and matrix associativity is presented to drive parallel SVD updating architectures. These architectures have critical loop computation time independent of the problem size and are suitable for CORDIC arithmetic based VLSI implementations.

AB - Singular value decomposition (SVD) has become a standard linear algebra tool in modern digital signal processing. CORDIC based SVD algorithms are among the most popular SVD algorithms which exhibit good numerical properties. The speed of the sequential algorithms is however limited by the recursive feedback loops in the underlying signal flow graph. The critical loop computation time is proportional to the size of the problem which prohibits pipelined processing. This paper addresses the derivation of parallel architectures for SVD updating algorithms. An algorithm transformation approach based on re-Timing and matrix associativity is presented to drive parallel SVD updating architectures. These architectures have critical loop computation time independent of the problem size and are suitable for CORDIC arithmetic based VLSI implementations.

UR - http://www.scopus.com/inward/record.url?scp=0033307134&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033307134&partnerID=8YFLogxK

U2 - 10.1109/ACSSC.1999.831936

DO - 10.1109/ACSSC.1999.831936

M3 - Conference contribution

AN - SCOPUS:0033307134

T3 - Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers

SP - 1401

EP - 1405

BT - Conference Record of the 33rd Asilomar Conference on Signals, Systems, and Computers

A2 - Matthews, Michael B.

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 33rd Asilomar Conference on Signals, Systems, and Computers, ACSSC 1999

Y2 - 24 October 1999 through 27 October 1999

ER -

An algorithm transformation approach to CORDIC based parallel singular value decompositions architectures

Abstract

Publication series

Other

Bibliographical note

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this