Processor Arrays for Problems in Computational Physics (Parallel)

As the cost of hardware components drops, the design and development of processor array systems, consisting of thousands of relatively simple processing elements, becomes possible. An example of such an architecture is the Goodyear Aerospace Massively Parallel Processor, which was originally conceived as a machine to support high-speed image processing. Starting from this existing system configuration, we show how such a design is used very effectively to solve large-scale scientific problems with heavy floating-point computation requirements. Results from the implementation of algorithms for the fast solution of equations occurring in numerical weather prediction and computational fluid dynamics on such systems are presented.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2014

Author dc:creator

• Gallopoulos, Efstratios

• Computer Science

Identifier

(UMI)AAI8511610

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/69541