A systolic array implementation of discrete relaxation algorithm

Update Item Information
Publication Type technical report
School or College College of Engineering
Department Computing, School of
Creator Henderson, Thomas C.
Other Author Wang, Wei; Gu, Jun
Title A systolic array implementation of discrete relaxation algorithm
Date 1986
Description Discrete Relaxation techniques have proven useful in solving a wide range of problems in digital signal processing, artificial intelligence, machine vision, and VLSI engineering, etc. A conventional hardware design for an 8-label 8-object Discrete Relaxation Algorithm (DRA) requires three 4K memory blocks and the maximum execution time of over seconds and minutes, which makes such a DRA hardware implementation infeasible. A highly parallel systolic array for the computation of an 8-label 8-object DRA problem has been developed. This realization eliminates the 12K memory requirement and performs DRA computation at the worst case in microseconds. The circuit requires about 6,382 transistors. Major design issues and chip descriptions are described in this paper.
Type Text
Publisher University of Utah
Subject Discrete Relaxation algorithm; Systolic array
Subject LCSH Relaxation methods (Mathematics)
Language eng
Bibliographic Citation Wang, W., Gu, J., & Henderson, T. C. (1986). A systolic array implementation of discrete relaxation algorithm. 1-55. UUCS-TR-86-008.
Series University of Utah Computer Science Technical Report
Relation is Part of ARPANET
Rights Management ©University of Utah
Format Medium application/pdf
Format Extent 26,381,544 bytes
Identifier ir-main,16310
ARK ark:/87278/s6rn3sb1
Setname ir_uspace
ID 705861
Reference URL https://collections.lib.utah.edu/ark:/87278/s6rn3sb1
Back to Search Results