Author(s)
Ernest Kent, S Tanimoto
Abstract
HCL is a hierarchical cellular logic, in which operations are applied to objects called bit-pyramids which themselves are functions on spaces called hierarchical domains. HCL provides an algebra for computations involving hierarchical, multiple-resolution representations of image data. PIPE is a newly-developed parallel architecture which supports multiple-resolution pyramid operations. This paper establishes that HCL is functionally equivalent to a subset of PIPE's instruction set, and that every HCL primitive operation corresponds to a single machine instruction in PIPE and executes in a single machine cycle. Further, the connectivity of HCL data-objects is embedded in the data-paths of the PIPE architecture. Thus, PIPE can operate upon the data-objects of HCL directly, without using extra storage for links or pointers, and without computation of storage addresses. As a result, PIPE programs implementing HCL may be expected to run enormously faster than corresponding programs for von Neumann machines, or for other parallel machines which do not share PIPE's architectural correspondence to the structures of HCL.
Proceedings Title
IEEE Computer Society Workshop on Computer Architecture for Pattern Analysis and Image Database Management - Capaidm
Conference Dates
July 1, 1985
Conference Location
Capaidm, FL, USA
Conference Title
Pattern Analysis and Image Database Management
Citation
Kent, E.
and Tanimoto, S.
(1985),
Hierarchical Cellular Logic and the Pipe Processor: Structural and Functional Correspondence, IEEE Computer Society Workshop on Computer Architecture for Pattern Analysis and Image Database Management - Capaidm, Capaidm, FL, USA, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=820204 (Accessed April 30, 2026)
Additional citation formats
Issues
If you have any questions about this publication or are having problems accessing it, please contact [email protected].