GPU acceleration of cutoff pair potentials for molecular modeling applications

Christopher I. Rodrigues, David J. Hardy, John E. Stone, Klaus Schulten, Wen Mei W. Hwu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The advent of systems biology requires the simulation of everlarger biomolecular systems, demanding a commensurate growth in computational power. This paper examines the use of the NVIDIA Tesla C870 graphics card programmed through the CUDA toolkit to accelerate the calculation of cutoff pair potentials, one of the most prevalent computations required by many different molecular modeling applications. We present algorithms to calculate electrostatic potential maps for cutoff pair potentials. Whereas a straightforward approach for decomposing atom data leads to low compute efficiency, a newer strategy enables fine-grained spatial decomposition of atom data that maps efficiently to the C870's memory system while increasing work-efficiency of atom data traversal by a factor of 5. The memory addressing flexibility exposed through CUDA's SPMD programming model is crucial in enabling this new strategy. An implementation of the new algorithm provides a greater than threefold performance improvement over our previously published implementation and runs 12 to 20 times faster than optimized CPU-only code. The lessons learned are generally applicable to algorithms accelerated by uniform grid spatial decomposition.

Original languageEnglish (US)
Title of host publicationConference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08
Pages273-282
Number of pages10
DOIs
StatePublished - Dec 1 2008
Event2008 Conference on Computing Frontiers, CF'08 - Ischia, Italy
Duration: May 5 2008May 7 2008

Publication series

NameConference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08

Other

Other2008 Conference on Computing Frontiers, CF'08
CountryItaly
CityIschia
Period5/5/085/7/08

Fingerprint

Molecular modeling
Atoms
Decomposition
Data storage equipment
Computer programming
Program processors
Electrostatics
Graphics processing unit

Keywords

  • CUDA
  • GPGPU
  • Graphics processors
  • Molecular dynamics

ASJC Scopus subject areas

  • Computer Science Applications
  • Hardware and Architecture
  • Software
  • Electrical and Electronic Engineering

Cite this

Rodrigues, C. I., Hardy, D. J., Stone, J. E., Schulten, K., & Hwu, W. M. W. (2008). GPU acceleration of cutoff pair potentials for molecular modeling applications. In Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08 (pp. 273-282). (Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08). https://doi.org/10.1145/1366230.1366277

GPU acceleration of cutoff pair potentials for molecular modeling applications. / Rodrigues, Christopher I.; Hardy, David J.; Stone, John E.; Schulten, Klaus; Hwu, Wen Mei W.

Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. p. 273-282 (Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rodrigues, CI, Hardy, DJ, Stone, JE, Schulten, K & Hwu, WMW 2008, GPU acceleration of cutoff pair potentials for molecular modeling applications. in Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08, pp. 273-282, 2008 Conference on Computing Frontiers, CF'08, Ischia, Italy, 5/5/08. https://doi.org/10.1145/1366230.1366277
Rodrigues CI, Hardy DJ, Stone JE, Schulten K, Hwu WMW. GPU acceleration of cutoff pair potentials for molecular modeling applications. In Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. p. 273-282. (Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08). https://doi.org/10.1145/1366230.1366277
Rodrigues, Christopher I. ; Hardy, David J. ; Stone, John E. ; Schulten, Klaus ; Hwu, Wen Mei W. / GPU acceleration of cutoff pair potentials for molecular modeling applications. Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. pp. 273-282 (Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08).
@inproceedings{06634d551d4143b28fc99c91d62a51b1,
title = "GPU acceleration of cutoff pair potentials for molecular modeling applications",
abstract = "The advent of systems biology requires the simulation of everlarger biomolecular systems, demanding a commensurate growth in computational power. This paper examines the use of the NVIDIA Tesla C870 graphics card programmed through the CUDA toolkit to accelerate the calculation of cutoff pair potentials, one of the most prevalent computations required by many different molecular modeling applications. We present algorithms to calculate electrostatic potential maps for cutoff pair potentials. Whereas a straightforward approach for decomposing atom data leads to low compute efficiency, a newer strategy enables fine-grained spatial decomposition of atom data that maps efficiently to the C870's memory system while increasing work-efficiency of atom data traversal by a factor of 5. The memory addressing flexibility exposed through CUDA's SPMD programming model is crucial in enabling this new strategy. An implementation of the new algorithm provides a greater than threefold performance improvement over our previously published implementation and runs 12 to 20 times faster than optimized CPU-only code. The lessons learned are generally applicable to algorithms accelerated by uniform grid spatial decomposition.",
keywords = "CUDA, GPGPU, Graphics processors, Molecular dynamics",
author = "Rodrigues, {Christopher I.} and Hardy, {David J.} and Stone, {John E.} and Klaus Schulten and Hwu, {Wen Mei W.}",
year = "2008",
month = "12",
day = "1",
doi = "10.1145/1366230.1366277",
language = "English (US)",
isbn = "9781605580777",
series = "Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08",
pages = "273--282",
booktitle = "Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08",

}

TY - GEN

T1 - GPU acceleration of cutoff pair potentials for molecular modeling applications

AU - Rodrigues, Christopher I.

AU - Hardy, David J.

AU - Stone, John E.

AU - Schulten, Klaus

AU - Hwu, Wen Mei W.

PY - 2008/12/1

Y1 - 2008/12/1

N2 - The advent of systems biology requires the simulation of everlarger biomolecular systems, demanding a commensurate growth in computational power. This paper examines the use of the NVIDIA Tesla C870 graphics card programmed through the CUDA toolkit to accelerate the calculation of cutoff pair potentials, one of the most prevalent computations required by many different molecular modeling applications. We present algorithms to calculate electrostatic potential maps for cutoff pair potentials. Whereas a straightforward approach for decomposing atom data leads to low compute efficiency, a newer strategy enables fine-grained spatial decomposition of atom data that maps efficiently to the C870's memory system while increasing work-efficiency of atom data traversal by a factor of 5. The memory addressing flexibility exposed through CUDA's SPMD programming model is crucial in enabling this new strategy. An implementation of the new algorithm provides a greater than threefold performance improvement over our previously published implementation and runs 12 to 20 times faster than optimized CPU-only code. The lessons learned are generally applicable to algorithms accelerated by uniform grid spatial decomposition.

AB - The advent of systems biology requires the simulation of everlarger biomolecular systems, demanding a commensurate growth in computational power. This paper examines the use of the NVIDIA Tesla C870 graphics card programmed through the CUDA toolkit to accelerate the calculation of cutoff pair potentials, one of the most prevalent computations required by many different molecular modeling applications. We present algorithms to calculate electrostatic potential maps for cutoff pair potentials. Whereas a straightforward approach for decomposing atom data leads to low compute efficiency, a newer strategy enables fine-grained spatial decomposition of atom data that maps efficiently to the C870's memory system while increasing work-efficiency of atom data traversal by a factor of 5. The memory addressing flexibility exposed through CUDA's SPMD programming model is crucial in enabling this new strategy. An implementation of the new algorithm provides a greater than threefold performance improvement over our previously published implementation and runs 12 to 20 times faster than optimized CPU-only code. The lessons learned are generally applicable to algorithms accelerated by uniform grid spatial decomposition.

KW - CUDA

KW - GPGPU

KW - Graphics processors

KW - Molecular dynamics

UR - http://www.scopus.com/inward/record.url?scp=53749106683&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=53749106683&partnerID=8YFLogxK

U2 - 10.1145/1366230.1366277

DO - 10.1145/1366230.1366277

M3 - Conference contribution

AN - SCOPUS:53749106683

SN - 9781605580777

T3 - Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08

SP - 273

EP - 282

BT - Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08

ER -