Floating point CORDIC for matrix computations

Author

Cavallaro, Joseph R. ; Luk, Franklin T.

Author_Institution

Dept. of Electr. & Comput. Eng., Rice Univ., Houston, TX, USA

fYear

1988

fDate

3-5 Oct 1988

Firstpage

40

Lastpage

42

Abstract

The Coordinate Rotation Digital Computer (CORDIC) algorithms provide a VLSI hardware technique for computing the inverse tangents and vector rotations needed by many matrix decomposition algorithms. A novel simplified CORDIC processor composed of floating-point data paths with a fixed-point angle calculation is proposed. This hybrid processor possesses sufficient accuracy for matrix computations such as the QRD, eigenvalue decomposition, and the singular value decomposition. The simplified structure allows for efficient VLSI implementation

Keywords

computer architecture; matrix algebra; microprocessor chips; CORDIC processor; Coordinate Rotation Digital Computer; QRD; eigenvalue decomposition; fixed-point angle calculation; floating point CORDIC processor; floating-point data paths; hybrid processor; inverse tangents; matrix computations; matrix decomposition algorithms; singular value decomposition; vector rotations; Adders; Equations; Floating-point arithmetic; Hardware; Matrix decomposition; Read only memory; Registers; Signal processing algorithms; Singular value decomposition; Very large scale integration;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Design: VLSI in Computers and Processors, 1988. ICCD '88., Proceedings of the 1988 IEEE International Conference on

Conference_Location

Rye Brook, NY

Print_ISBN

0-8186-0872-2

Type

conf

DOI

10.1109/ICCD.1988.25655

Filename

25655