Implementation and testing of a parallel layer peeling algorithm /

(RILA). The results from the parallel implementation were

Bibliographic Details
Main Author: Gandapur, Tasneem Kausar
Format: Thesis eBook
Language:English
Published: [Place of publication not identified] : [publisher not identified] ; 1996.
Subjects:
Online Access:Link to OAKTrust copy
Description
Summary:(RILA). The results from the parallel implementation were
algorithm applied invariant imbedding methods to the two
allocation, or the total workload breakdown for assignment
always between nearest neighbors.
and analysis. These included issues such as task
are used to move deeper, into the ith layer. The parallel
Assignment (BLA), and Rotated Individual Layer Assignment
between processors, and synchronization of the
computations from different processors.
considered in the parallel layer peeling implementation
described, and the characteristic tracing method has been
ensures that the address of the processor working on any
equation describing radiative transfer. The published
equation, resulting in an integrodifferential equation that
implementation of the layer peeling algorithm was considered.
implementation was considered because of the anticipated
implementation was done by using Reflected Gray Codes for
implementations of the above mentioned serial algorithm viz.,
implemented in parallel. For this thesis, the parallel
In order to achieve better performance, the parallel
In this context, layer peeling implies that pertinent
input of the direct solution.
its results as input to the inverse solution. The results
layer differs in at most a single bit from the processor
Layer Peeling and Characteristic Tracing have been previously
obtained from the inverse solution were then compared to the
parallel algorithm were implemented viz., Block Layer
parallelizing a previously published sequential algorithm.
problem associated with the first order, partial differential
processor assignment. Reflected Gray Code ordering
quantities are known at some layer, say the (i-I)th, which
Several issues specific to parallel computation were
speedup in the solution of the problem under consideration.
stream approximation to a one dimensional radiative transfer
takes multiple scattering events into account. Two different
The algorithm was implemented on an NCUBE 2 machine, with a
The objective of this thesis is to examine the feasibility of
This algorithm uses invariant imbedding to solve an inverse
to different processors, communication of interim results
total of sixty-four processors. Two variations of the
verified for accuracy by solving the direct problem and using
working on the neighboring layer, so the communication is
Item Description:"Major subject: Computer Science".
Vita.
Physical Description:xi, 82 leaves : illustrations ; 28 cm.
Also available online.
Issued also on microfiche from Lange Micrographics.
Bibliography:Includes bibliographical references: pages 76-78.