GPU Computing for Atmospheric Modeling: Experience with a Small Kernel and Implications for a Full Model

Rory Kelly

doi:10.1109/MCSE.2010.26

GPU Computing for Atmospheric Modeling: Experience with a Small Kernel and Implications for a Full Model

Rory Kelly

Consulting Services Group

Research output: Contribution to journal › Article › peer-review

31 Scopus citations

Abstract

Modern graphics processing units offer enticing speedup potential for atmospheric modeling. However, accelerating such computationally expensive models using GPUs isn't straightforward. Porting a large atmospheric model's most expensive routine to an Nvidia 9800 GX2 let one researcher explore GPU use and glimpse the possibilities for accelerating an entire model.

Original language	English
Pages (from-to)	26-33
Number of pages	8
Journal	Computing in Science and Engineering
Volume	12
Issue number	4
DOIs	https://doi.org/10.1109/MCSE.2010.26
State	Published - 2010

Access to Document

10.1109/MCSE.2010.26

Cite this

@article{cf2356d5d4da4d8a94d60bca9fac7e77,

title = "GPU Computing for Atmospheric Modeling: Experience with a Small Kernel and Implications for a Full Model",

abstract = "Modern graphics processing units offer enticing speedup potential for atmospheric modeling. However, accelerating such computationally expensive models using GPUs isn't straightforward. Porting a large atmospheric model's most expensive routine to an Nvidia 9800 GX2 let one researcher explore GPU use and glimpse the possibilities for accelerating an entire model.",

author = "Rory Kelly",

year = "2010",

doi = "10.1109/MCSE.2010.26",

language = "English",

volume = "12",

pages = "26--33",

journal = "Computing in Science and Engineering",

issn = "1521-9615",

publisher = "IEEE Computer Society",

number = "4",

}

TY - JOUR

T1 - GPU Computing for Atmospheric Modeling

T2 - Experience with a Small Kernel and Implications for a Full Model

AU - Kelly, Rory

PY - 2010

Y1 - 2010

N2 - Modern graphics processing units offer enticing speedup potential for atmospheric modeling. However, accelerating such computationally expensive models using GPUs isn't straightforward. Porting a large atmospheric model's most expensive routine to an Nvidia 9800 GX2 let one researcher explore GPU use and glimpse the possibilities for accelerating an entire model.

AB - Modern graphics processing units offer enticing speedup potential for atmospheric modeling. However, accelerating such computationally expensive models using GPUs isn't straightforward. Porting a large atmospheric model's most expensive routine to an Nvidia 9800 GX2 let one researcher explore GPU use and glimpse the possibilities for accelerating an entire model.

UR - https://www.scopus.com/pages/publications/85008063764

U2 - 10.1109/MCSE.2010.26

DO - 10.1109/MCSE.2010.26

M3 - Article

AN - SCOPUS:85008063764

SN - 1521-9615

VL - 12

SP - 26

EP - 33

JO - Computing in Science and Engineering

JF - Computing in Science and Engineering

IS - 4

ER -

GPU Computing for Atmospheric Modeling: Experience with a Small Kernel and Implications for a Full Model

Abstract

Access to Document

Other files and links

Fingerprint

Cite this