A guide for implementing tridiagonal solvers on GPUs

Li Wen Chang; Wen Mei W. Hwu

doi:10.1007/978-3-319-06548-9_2

A guide for implementing tridiagonal solvers on GPUs

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Abstract

Tridiagonal solvers are important building blocks for many applications on GPUs. Although a wide range of algorithms and optimizations have been proposed for tridiagonal solvers, there are no comprehensive guidelines for building a high-performance tridiagonal solver for GPUs. In this chapter, we review and summarize several existing optimization strategies, and systematically analyze the purpose for each optimization. Finally, a case study, called SPIKE-CR, is given to demonstrate how to apply the guidelines to build a high-performance GPU tridiagonal solver.

Original language	English (US)
Title of host publication	Numerical Computations with GPUs
Publisher	Springer
Pages	29-44
Number of pages	16
ISBN (Electronic)	9783319065489
ISBN (Print)	9783319065472
DOIs	https://doi.org/10.1007/978-3-319-06548-9_2
State	Published - Jan 1 2014

ASJC Scopus subject areas

General Computer Science

Online availability

10.1007/978-3-319-06548-9_2

Library availability

Discover UIUC Full Text

Cite this

@inbook{8435bfbd82e146269b4d5314bd7af032,

title = "A guide for implementing tridiagonal solvers on GPUs",

abstract = "Tridiagonal solvers are important building blocks for many applications on GPUs. Although a wide range of algorithms and optimizations have been proposed for tridiagonal solvers, there are no comprehensive guidelines for building a high-performance tridiagonal solver for GPUs. In this chapter, we review and summarize several existing optimization strategies, and systematically analyze the purpose for each optimization. Finally, a case study, called SPIKE-CR, is given to demonstrate how to apply the guidelines to build a high-performance GPU tridiagonal solver.",

author = "Chang, {Li Wen} and Hwu, {Wen Mei W.}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2014.",

year = "2014",

month = jan,

day = "1",

doi = "10.1007/978-3-319-06548-9_2",

language = "English (US)",

isbn = "9783319065472",

pages = "29--44",

booktitle = "Numerical Computations with GPUs",

publisher = "Springer",

address = "Germany",

}

TY - CHAP

T1 - A guide for implementing tridiagonal solvers on GPUs

AU - Chang, Li Wen

AU - Hwu, Wen Mei W.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Tridiagonal solvers are important building blocks for many applications on GPUs. Although a wide range of algorithms and optimizations have been proposed for tridiagonal solvers, there are no comprehensive guidelines for building a high-performance tridiagonal solver for GPUs. In this chapter, we review and summarize several existing optimization strategies, and systematically analyze the purpose for each optimization. Finally, a case study, called SPIKE-CR, is given to demonstrate how to apply the guidelines to build a high-performance GPU tridiagonal solver.

AB - Tridiagonal solvers are important building blocks for many applications on GPUs. Although a wide range of algorithms and optimizations have been proposed for tridiagonal solvers, there are no comprehensive guidelines for building a high-performance tridiagonal solver for GPUs. In this chapter, we review and summarize several existing optimization strategies, and systematically analyze the purpose for each optimization. Finally, a case study, called SPIKE-CR, is given to demonstrate how to apply the guidelines to build a high-performance GPU tridiagonal solver.

UR - http://www.scopus.com/inward/record.url?scp=84946477815&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84946477815&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-06548-9_2

DO - 10.1007/978-3-319-06548-9_2

M3 - Chapter

AN - SCOPUS:84946477815

SN - 9783319065472

SP - 29

EP - 44

BT - Numerical Computations with GPUs

PB - Springer

ER -

A guide for implementing tridiagonal solvers on GPUs

Abstract

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this