Improving the Performance of CP2K on the Cray XT

Iain Bethune

Improving the Performance of CP2K on the Cray XT

Iain Bethune

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

Abstract

CP2K is a freely available and increasingly popular Density Functional Theory code for the simulation of a wide range of systems. It is heavily used on many Cray XT systems, including ‘HECToR’ in the UK and ‘Monte Rosa’ in Switzerland. We describe performance optimisations made to the code in several key areas, including 3D Fourier Transforms, and present the implementation of a load balancing scheme for multi-grids. These result in performance gains of around 30% on 256 cores (for a generally representative benchmark) and up to 300% on 1024 cores (for non-homogeneous systems). Early results from the implementation of hybrid MPI/OpenMP parallelism in the code are also presented.

Original language	English
Title of host publication	Proceedings of the Cray User Group, 2010
Publication status	Published - May 2010
Event	Cray User Groups (CUG) 2010 - Edinburgh, United Kingdom Duration: 24 May 2010 → 27 May 2010

Conference

Conference	Cray User Groups (CUG) 2010
Country/Territory	United Kingdom
City	Edinburgh
Period	24/05/10 → 27/05/10

Keywords

CP2K
Cray XT
MPI
OpenMP
FFT
Load Balancing

Access to Document

Improving the Performance of CP2K on the Cray XTFinal published version, 155 KB

https://cug.org/5-publications/proceedings_attendee_lists/CUG10CD/pages/1-program/final_program/CUG10_Proceedings/pages/authors/16-18Thursday/17B-bethune.pdf

Cite this

@inproceedings{4994309f760b455c95f2f7cb0f64b0ad,

title = "Improving the Performance of CP2K on the Cray XT",

abstract = "CP2K is a freely available and increasingly popular Density Functional Theory code for the simulation of a wide range of systems. It is heavily used on many Cray XT systems, including {\textquoteleft}HECToR{\textquoteright} in the UK and {\textquoteleft}Monte Rosa{\textquoteright} in Switzerland. We describe performance optimisations made to the code in several key areas, including 3D Fourier Transforms, and present the implementation of a load balancing scheme for multi-grids. These result in performance gains of around 30% on 256 cores (for a generally representative benchmark) and up to 300% on 1024 cores (for non-homogeneous systems). Early results from the implementation of hybrid MPI/OpenMP parallelism in the code are also presented.",

keywords = "CP2K, Cray XT, MPI, OpenMP, FFT, Load Balancing",

author = "Iain Bethune",

year = "2010",

month = may,

language = "English",

booktitle = "Proceedings of the Cray User Group, 2010",

note = "Cray User Groups (CUG) 2010 ; Conference date: 24-05-2010 Through 27-05-2010",

}

TY - GEN

T1 - Improving the Performance of CP2K on the Cray XT

AU - Bethune, Iain

PY - 2010/5

Y1 - 2010/5

N2 - CP2K is a freely available and increasingly popular Density Functional Theory code for the simulation of a wide range of systems. It is heavily used on many Cray XT systems, including ‘HECToR’ in the UK and ‘Monte Rosa’ in Switzerland. We describe performance optimisations made to the code in several key areas, including 3D Fourier Transforms, and present the implementation of a load balancing scheme for multi-grids. These result in performance gains of around 30% on 256 cores (for a generally representative benchmark) and up to 300% on 1024 cores (for non-homogeneous systems). Early results from the implementation of hybrid MPI/OpenMP parallelism in the code are also presented.

AB - CP2K is a freely available and increasingly popular Density Functional Theory code for the simulation of a wide range of systems. It is heavily used on many Cray XT systems, including ‘HECToR’ in the UK and ‘Monte Rosa’ in Switzerland. We describe performance optimisations made to the code in several key areas, including 3D Fourier Transforms, and present the implementation of a load balancing scheme for multi-grids. These result in performance gains of around 30% on 256 cores (for a generally representative benchmark) and up to 300% on 1024 cores (for non-homogeneous systems). Early results from the implementation of hybrid MPI/OpenMP parallelism in the code are also presented.

KW - CP2K

KW - Cray XT

KW - MPI

KW - OpenMP

KW - FFT

KW - Load Balancing

M3 - Conference contribution

BT - Proceedings of the Cray User Group, 2010

T2 - Cray User Groups (CUG) 2010

Y2 - 24 May 2010 through 27 May 2010

ER -

Improving the Performance of CP2K on the Cray XT

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this