The mapping of two-dimensional array processors to one-dimensional processors

DOI10.1016/0167-8191(86)90021-9MaRDI QIDQ1079955zbMATH OpenOpenAlexFDO

Authors Chang-Biau Yang, R. C. T. Lee

Publication date 1986

Published in Parallel Computing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/0167-8191(86)90021-9

zbMATH Keywords

linear array processorlongest common subsequencespeech recognitionspoken word recognitionsystolic algorithmstwo-dimensional wavefront array processorVLSI

Mathematics Subject Classification ID

Pattern recognition, speech recognition (68T10) Algorithms in computer science (68W99)

Review text

We consider the case of a two-dimensional wavefront array processor where only one wavefront appears at any time. We show that in such a situation, this two-dimensional wavefront processor can be mapped to a linear array processor if the wavefronts never backtrack. The mapping will not increase the number of registers in each processor element. Two examples, the spoken word recognition problem and the longest common subsequence problem, are given to demonstrate the feasibility of this method.

Recommendations

• Mapping Homogeneous Graphs on Linear Arrays
• scientific article; zbMATH DE number 426143
• Mapping dynamic programming onto a linear systolic array
• Partitioning of processor arrays: a piecewise regular approach
• scientific article; zbMATH DE number 194950

Cited in

(4)

• Mapping of neural network models onto two-dimensional processor arrays
• Mapping Homogeneous Graphs on Linear Arrays
• A GENERALIZED MAPPING OF 2-D ARRAY PROCESSOR TO LINEAR ARRAY PROCESSOR AND ITS APPLICATIONS
• Two dimensional processor array with a reconfigurable bus system is at least as powerful as CRCW model

This page was built for publication: The mapping of two-dimensional array processors to one-dimensional processors

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1079955)