High performance computation and display of molecular orbitals on and multi-core cpus

John E. Stone, Jan Saam, David J. Hardy, Kirby L. Vandivort, Wen-Mei W Hwu, Klaus J Schulten

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

Abstract

The visualization of molecular orbitals (MOs) is important for analyzing the results of quantum chemistry simulations. The functions describing the MOs are computed on a threedimensional lattice, and the resulting data can then be used for plotting isocontours or isosurfaces for visualization as well as for other types of analyses. Existing software packages that render MOs perform calculations on the CPU and require runtimes of tens to hundreds of seconds depending on the complexity of the molecular system. We present novel data-parallel algorithms for computing lattices of MOs on modern graphics processing units (GPUs) and multi-core CPUs. The fastest GPU algorithm achieves up to a 125-fold speedup over an optimized CPU implementation running on one CPU core. We also demonstrate possible benefits of dynamic GPU kernel generation and just-intime compilation for MO calculation. We have implemented these algorithms within the popular molecular visualization program VMD, which can now produce high quality MO renderings for large systems in less than a second, and achieves the first-ever interactive animations of quantum chemistry simulation trajectories using only on-the-y calculation.

Original languageEnglish (US)
Title of host publicationProceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2
Number of pages1
DOIs
StatePublished - Jul 23 2009
Event2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2 - Washington, D.C., United States
Duration: Mar 8 2009Mar 8 2009

Publication series

NameProceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2

Other

Other2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2
CountryUnited States
CityWashington, D.C.
Period3/8/093/8/09

Fingerprint

Molecular orbitals
Display devices
Program processors
Quantum chemistry
Orbital calculations
Visualization
Describing functions
Animation
Parallel algorithms
Software packages
Trajectories
Graphics processing unit

Keywords

  • Algorithms
  • CUDA
  • Design
  • GPGPU
  • Molecular orbital
  • Performance GPU computing

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Hardware and Architecture
  • Software

Cite this

Stone, J. E., Saam, J., Hardy, D. J., Vandivort, K. L., Hwu, W-M. W., & Schulten, K. J. (2009). High performance computation and display of molecular orbitals on and multi-core cpus. In Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2 (Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2). https://doi.org/10.1145/1513895.1513897

High performance computation and display of molecular orbitals on and multi-core cpus. / Stone, John E.; Saam, Jan; Hardy, David J.; Vandivort, Kirby L.; Hwu, Wen-Mei W; Schulten, Klaus J.

Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2. 2009. (Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2).

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

Stone, JE, Saam, J, Hardy, DJ, Vandivort, KL, Hwu, W-MW & Schulten, KJ 2009, High performance computation and display of molecular orbitals on and multi-core cpus. in Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2. Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2, 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2, Washington, D.C., United States, 3/8/09. https://doi.org/10.1145/1513895.1513897
Stone JE, Saam J, Hardy DJ, Vandivort KL, Hwu W-MW, Schulten KJ. High performance computation and display of molecular orbitals on and multi-core cpus. In Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2. 2009. (Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2). https://doi.org/10.1145/1513895.1513897
Stone, John E. ; Saam, Jan ; Hardy, David J. ; Vandivort, Kirby L. ; Hwu, Wen-Mei W ; Schulten, Klaus J. / High performance computation and display of molecular orbitals on and multi-core cpus. Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2. 2009. (Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2).
@inproceedings{d8efdf28b9064c31a17f2267abca9345,
title = "High performance computation and display of molecular orbitals on and multi-core cpus",
abstract = "The visualization of molecular orbitals (MOs) is important for analyzing the results of quantum chemistry simulations. The functions describing the MOs are computed on a threedimensional lattice, and the resulting data can then be used for plotting isocontours or isosurfaces for visualization as well as for other types of analyses. Existing software packages that render MOs perform calculations on the CPU and require runtimes of tens to hundreds of seconds depending on the complexity of the molecular system. We present novel data-parallel algorithms for computing lattices of MOs on modern graphics processing units (GPUs) and multi-core CPUs. The fastest GPU algorithm achieves up to a 125-fold speedup over an optimized CPU implementation running on one CPU core. We also demonstrate possible benefits of dynamic GPU kernel generation and just-intime compilation for MO calculation. We have implemented these algorithms within the popular molecular visualization program VMD, which can now produce high quality MO renderings for large systems in less than a second, and achieves the first-ever interactive animations of quantum chemistry simulation trajectories using only on-the-y calculation.",
keywords = "Algorithms, CUDA, Design, GPGPU, Molecular orbital, Performance GPU computing",
author = "Stone, {John E.} and Jan Saam and Hardy, {David J.} and Vandivort, {Kirby L.} and Hwu, {Wen-Mei W} and Schulten, {Klaus J}",
year = "2009",
month = "7",
day = "23",
doi = "10.1145/1513895.1513897",
language = "English (US)",
isbn = "9781605585178",
series = "Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2",
booktitle = "Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2",

}

TY - GEN

T1 - High performance computation and display of molecular orbitals on and multi-core cpus

AU - Stone, John E.

AU - Saam, Jan

AU - Hardy, David J.

AU - Vandivort, Kirby L.

AU - Hwu, Wen-Mei W

AU - Schulten, Klaus J

PY - 2009/7/23

Y1 - 2009/7/23

N2 - The visualization of molecular orbitals (MOs) is important for analyzing the results of quantum chemistry simulations. The functions describing the MOs are computed on a threedimensional lattice, and the resulting data can then be used for plotting isocontours or isosurfaces for visualization as well as for other types of analyses. Existing software packages that render MOs perform calculations on the CPU and require runtimes of tens to hundreds of seconds depending on the complexity of the molecular system. We present novel data-parallel algorithms for computing lattices of MOs on modern graphics processing units (GPUs) and multi-core CPUs. The fastest GPU algorithm achieves up to a 125-fold speedup over an optimized CPU implementation running on one CPU core. We also demonstrate possible benefits of dynamic GPU kernel generation and just-intime compilation for MO calculation. We have implemented these algorithms within the popular molecular visualization program VMD, which can now produce high quality MO renderings for large systems in less than a second, and achieves the first-ever interactive animations of quantum chemistry simulation trajectories using only on-the-y calculation.

AB - The visualization of molecular orbitals (MOs) is important for analyzing the results of quantum chemistry simulations. The functions describing the MOs are computed on a threedimensional lattice, and the resulting data can then be used for plotting isocontours or isosurfaces for visualization as well as for other types of analyses. Existing software packages that render MOs perform calculations on the CPU and require runtimes of tens to hundreds of seconds depending on the complexity of the molecular system. We present novel data-parallel algorithms for computing lattices of MOs on modern graphics processing units (GPUs) and multi-core CPUs. The fastest GPU algorithm achieves up to a 125-fold speedup over an optimized CPU implementation running on one CPU core. We also demonstrate possible benefits of dynamic GPU kernel generation and just-intime compilation for MO calculation. We have implemented these algorithms within the popular molecular visualization program VMD, which can now produce high quality MO renderings for large systems in less than a second, and achieves the first-ever interactive animations of quantum chemistry simulation trajectories using only on-the-y calculation.

KW - Algorithms

KW - CUDA

KW - Design

KW - GPGPU

KW - Molecular orbital

KW - Performance GPU computing

UR - http://www.scopus.com/inward/record.url?scp=67650652938&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650652938&partnerID=8YFLogxK

U2 - 10.1145/1513895.1513897

DO - 10.1145/1513895.1513897

M3 - Conference contribution

AN - SCOPUS:67650652938

SN - 9781605585178

T3 - Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2

BT - Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU-2

ER -