English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
Add to Basket
  Local TagsStatisticsRelease HistoryDetailsSummary

Released
Journal Article

Arbitrary Angular Momentum Electron Repulsion Integrals with Graphical Processing Units: Application to the Resolution of Identity Hartree–Fock Method

MPS-Authors
/persons/resource/persons237807

Kalinowski,  Jaroslaw
Research Department Schlögl, Max Planck Institute for Chemical Energy Conversion, Max Planck Society;

/persons/resource/persons216844

Wennmohs,  Frank
Research Department Neese, Max Planck Institute for Chemical Energy Conversion, Max Planck Society;

/persons/resource/persons216825

Neese,  Frank
Research Department Neese, Max Planck Institute for Chemical Energy Conversion, Max Planck Society;

External Ressource
No external resources are shared
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Kalinowski, J., Wennmohs, F., & Neese, F. (2017). Arbitrary Angular Momentum Electron Repulsion Integrals with Graphical Processing Units: Application to the Resolution of Identity Hartree–Fock Method. Journal of Chemical Theory and Computation, 13(7), 1360-1370. doi:10.1021/acs.jctc.7b00030.


Cite as: http://hdl.handle.net/21.11116/0000-0007-6F35-2
Abstract
A resolution of identity based implementation of the Hartree–Fock method on graphical processing units (GPUs) is presented that is capable of handling basis functions with arbitrary angular momentum. For practical reasons, only functions up to (ff|f) angular momentum are presently calculated on the GPU, thus leaving the calculation of higher angular momenta integrals on the CPU of the hybrid CPU-GPU environment. Speedups of up to a factor of 30 are demonstrated relative to state-of-the-art serial and parallel CPU implementations. Benchmark calculations with over 3500 contracted basis functions (def2-SVP or def2-TZVP basis sets) are reported. The presented implementation supports all devices with OpenCL support and is capable of utilizing multiple GPU cards over either MPI or OpenCL itself.