Mixed-precision models for calculation of high-order virial coefficients on GPUs

Chao Feng; Andrew Schultz; Vipin Chaudhary; David Kofke

doi:10.1109/HiPC.2014.7116898

Mixed-precision models for calculation of high-order virial coefficients on GPUs

Chao Feng
, Andrew Schultz
, Vipin Chaudhary
, David Kofke

Department of Chemical and Biological Engineering

SUNY Buffalo

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

Abstract

The virial equation of state (VEOS) is a density expansion of the thermodynamic pressure with respect to an ideal-gas reference. Its coefficients can be computed from a molecular model, and become more expensive to calculate at higher order. In this paper, we use GPU to calculate the 8^th, 9^th and 10^th virial coefficients of the Lennard-Jones (LJ) potential model by the Mayer Sampling Monte Carlo (MSMC) method and Wheatley's algorithm. Two mixed-precision models are proposed to overcome a potential precision limitation of current GPUs while maintaining the performance benefit. On the latest Kepler architecture GPU Tesla K40, an average speedup of 20 to 40 is achieved for these calculations.

Original language	English
Title of host publication	2014 21st International Conference on High Performance Computing, HiPC 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781479959761
DOIs	https://doi.org/10.1109/HiPC.2014.7116898
State	Published - 2014
Event	2014 21st International Conference on High Performance Computing, HiPC 2014 - Goa, India Duration: Dec 17 2014 → Dec 20 2014

Publication series

Name	2014 21st International Conference on High Performance Computing, HiPC 2014

Conference

Conference	2014 21st International Conference on High Performance Computing, HiPC 2014
Country/Territory	India
City	Goa
Period	12/17/14 → 12/20/14

Keywords

GPU
Lennard-Jones Potential
Mayer Sampling Monte Carlo
Mixed-Precision
Virial Equation of State
Wheatley's Algorithm

Access to Document

10.1109/HiPC.2014.7116898

Cite this

Feng, C., Schultz, A., Chaudhary, V., & Kofke, D. (2014). Mixed-precision models for calculation of high-order virial coefficients on GPUs. In 2014 21st International Conference on High Performance Computing, HiPC 2014 Article 7116898 (2014 21st International Conference on High Performance Computing, HiPC 2014). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/HiPC.2014.7116898

@inproceedings{68b10eb00b41477eb90986ce55598b36,

title = "Mixed-precision models for calculation of high-order virial coefficients on GPUs",

abstract = "The virial equation of state (VEOS) is a density expansion of the thermodynamic pressure with respect to an ideal-gas reference. Its coefficients can be computed from a molecular model, and become more expensive to calculate at higher order. In this paper, we use GPU to calculate the 8th, 9th and 10th virial coefficients of the Lennard-Jones (LJ) potential model by the Mayer Sampling Monte Carlo (MSMC) method and Wheatley's algorithm. Two mixed-precision models are proposed to overcome a potential precision limitation of current GPUs while maintaining the performance benefit. On the latest Kepler architecture GPU Tesla K40, an average speedup of 20 to 40 is achieved for these calculations.",

keywords = "GPU, Lennard-Jones Potential, Mayer Sampling Monte Carlo, Mixed-Precision, Virial Equation of State, Wheatley's Algorithm",

author = "Chao Feng and Andrew Schultz and Vipin Chaudhary and David Kofke",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 21st International Conference on High Performance Computing, HiPC 2014 ; Conference date: 17-12-2014 Through 20-12-2014",

year = "2014",

doi = "10.1109/HiPC.2014.7116898",

language = "English",

series = "2014 21st International Conference on High Performance Computing, HiPC 2014",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2014 21st International Conference on High Performance Computing, HiPC 2014",

address = "United States",

}

Feng, C, Schultz, A, Chaudhary, V & Kofke, D 2014, Mixed-precision models for calculation of high-order virial coefficients on GPUs. in 2014 21st International Conference on High Performance Computing, HiPC 2014., 7116898, 2014 21st International Conference on High Performance Computing, HiPC 2014, Institute of Electrical and Electronics Engineers Inc., 2014 21st International Conference on High Performance Computing, HiPC 2014, Goa, India, 12/17/14. https://doi.org/10.1109/HiPC.2014.7116898

Mixed-precision models for calculation of high-order virial coefficients on GPUs. / Feng, Chao; Schultz, Andrew; Chaudhary, Vipin et al.
2014 21st International Conference on High Performance Computing, HiPC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. 7116898 (2014 21st International Conference on High Performance Computing, HiPC 2014).

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

TY - GEN

T1 - Mixed-precision models for calculation of high-order virial coefficients on GPUs

AU - Feng, Chao

AU - Schultz, Andrew

AU - Chaudhary, Vipin

AU - Kofke, David

PY - 2014

Y1 - 2014

N2 - The virial equation of state (VEOS) is a density expansion of the thermodynamic pressure with respect to an ideal-gas reference. Its coefficients can be computed from a molecular model, and become more expensive to calculate at higher order. In this paper, we use GPU to calculate the 8th, 9th and 10th virial coefficients of the Lennard-Jones (LJ) potential model by the Mayer Sampling Monte Carlo (MSMC) method and Wheatley's algorithm. Two mixed-precision models are proposed to overcome a potential precision limitation of current GPUs while maintaining the performance benefit. On the latest Kepler architecture GPU Tesla K40, an average speedup of 20 to 40 is achieved for these calculations.

AB - The virial equation of state (VEOS) is a density expansion of the thermodynamic pressure with respect to an ideal-gas reference. Its coefficients can be computed from a molecular model, and become more expensive to calculate at higher order. In this paper, we use GPU to calculate the 8th, 9th and 10th virial coefficients of the Lennard-Jones (LJ) potential model by the Mayer Sampling Monte Carlo (MSMC) method and Wheatley's algorithm. Two mixed-precision models are proposed to overcome a potential precision limitation of current GPUs while maintaining the performance benefit. On the latest Kepler architecture GPU Tesla K40, an average speedup of 20 to 40 is achieved for these calculations.

KW - GPU

KW - Lennard-Jones Potential

KW - Mayer Sampling Monte Carlo

KW - Mixed-Precision

KW - Virial Equation of State

KW - Wheatley's Algorithm

UR - https://www.scopus.com/pages/publications/84988299622

U2 - 10.1109/HiPC.2014.7116898

DO - 10.1109/HiPC.2014.7116898

M3 - Conference contribution

AN - SCOPUS:84988299622

T3 - 2014 21st International Conference on High Performance Computing, HiPC 2014

BT - 2014 21st International Conference on High Performance Computing, HiPC 2014

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 21st International Conference on High Performance Computing, HiPC 2014

Y2 - 17 December 2014 through 20 December 2014

ER -

Feng C, Schultz A, Chaudhary V, Kofke D. Mixed-precision models for calculation of high-order virial coefficients on GPUs. In 2014 21st International Conference on High Performance Computing, HiPC 2014. Institute of Electrical and Electronics Engineers Inc. 2014. 7116898. (2014 21st International Conference on High Performance Computing, HiPC 2014). doi: 10.1109/HiPC.2014.7116898

Mixed-precision models for calculation of high-order virial coefficients on GPUs

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this