Skip to main content


eCommons@Cornell

eCommons@Cornell >
College of Engineering >
Computer Science >
Computer Science Technical Reports >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1813/6803
Title: Hypercube Algorithms on the Polymorphic Torus
Authors: Elster, Anne C.
Li, Hungwen
Keywords: computer science
technical report
Issue Date: May-1989
Publisher: Cornell University
Citation: http://techreports.library.cornell.edu:8081/Dienst/UI/1.0/Display/cul.cs/TR89-1003
Abstract: The Polymorphic Torus architecture is a reconfigurable, massively parallel finegrained system, which in its two-dimensional $(N^{2})$ case has a lower wiring complexity than, say, hypercubes. However, due to the dynamic connection features at run-time, it allows several parallel structures such as trees, rings, and hypercubes to be emulated efficiently. In this paper, we consider algorithms that are especially well-suited for hypercubes, i.e. algorithms that take advantage of the relatively high connectivity of the hypercube topology, and show how these algorithms attain comparable bounds on a 2-D Polymorphic Torus. In particular, algorithms for dense matrix vector multiplication (including using 2 orthogonal trees for the matrix-vector case), sparse matrix-vector multiplication, and the FFT are discussed.
URI: http://hdl.handle.net/1813/6803
Appears in Collections:Computer Science Technical Reports

Files in This Item:

File Description SizeFormat
89-1003.pdf1.77 MBAdobe PDFView/Open
89-1003.ps430.6 kBPostscriptView/Open

Refworks Export

Items in eCommons are protected by copyright, with all rights reserved, unless otherwise indicated.

 

© 2014 Cornell University Library Contact Us