Optimization of a GPU implantation of multi-dimensional RF pulse design algorithm

Date

2011-05

Authors

Contributor

Advisor

Department

Instructor

Depositor

Speaker

Researcher

Consultant

Interviewer

Narrator

Transcriber

Annotator

Journal Title

Journal ISSN

Volume Title

Publisher

University of Hawaii at Manoa

Volume

Number/Issue

Starting Page

Ending Page

Alternative Title

Abstract

Pulse sequences are commonly used for both routine clinical diagnosis and advanced applications. The purpose of the pulse sequence is to manipulate the magnetization in order to produce the desired signal. However, the main drawback is that the calculation takes too much time. This project shows that modern graphics processing units work efficiently during the RF Pulse generation, which could improve the time calculation speed enormously. This paper introduces the use of GPU techniques such as multithreads, sharedMem, coalesce, and constMem to optimize the performance of the conjugate gradient least square algorithm for RF pulse. Compared to a CPU implementation, a version utilizing the conjugate gradient least square algorithm on the NVIDIA Tesla C 1060 is significantly faster. The total time is reduced by a factor of 129 relative to the CPU version.

Description

Keywords

GPU

Citation

Extent

Format

Geographic Location

Time Period

Related To

Theses for the degree of Master of Science (University of Hawaii at Manoa). Electrical Engineering.

Related To (URI)

Table of Contents

Rights

Rights Holder

Local Contexts

Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.