Date of Award

Fall 2004

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical & Computer Engineering

Program/Concentration

Electrical Engineering

Committee Director

James F. Leathrum, Jr.

Committee Member

Lee A. Belfore II

Committee Member

Vijayan Asari

Call Number for Print

Special Collections LD4331.E55 F65 2004

Abstract

Vector calculations are very prevalent today. Though the vector-processing computer is quite an old concept, superscalar processors lack hardware support for vector operations. This thesis investigates whether an ordinary superscalar computer architecture can be designed to include hardware support for improved vector operations without drastically changing the existing superscalar design and behavior. A computer architecture design was created and implemented that included the vector multiply (dot product) operation. The design includes a Vector Operations Unit that captures incoming vector operations and generates the necessary set of machine instructions to complete the vector operation internally. It then delivers these instructions to the dispatch and decode logic of the original superscalar architecture to be executed. This method of support for vector operations results in an extremely high density of instructions generated for each instruction that is fetched from memory. The results of this design show a significant speedup when tested against certain software implementations of the vector multiply algorithm.

Rights

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

DOI

10.25777/ehwv-xg02

Share

COinS