GPU-Accelerated Large-Scale Excited-State Simulation Based on Divide-and-Conquer Time-Dependent Density-Functional Tight-Binding

Research output: Contribution to journalArticle

Abstract

The present study implemented the divide-and-conquer time-dependent density-functional tight-binding (DC-TDDFTB) code on a graphical processing unit (GPU). The DC method, which is a linear-scaling scheme, divides a total system into several fragments. By separately solving local equations in individual fragments, the DC method could reduce slow central processing unit (CPU)-GPU memory access, as well as computational cost, and avoid shortfalls of GPU memory. Numerical applications confirmed that the present code on GPU significantly accelerated the TDDFTB calculations, while maintaining accuracy. Furthermore, the DC-TDDFTB simulation of 2-acetylindan-1,3-dione displays excited-state intramolecular proton transfer and provides reasonable absorption and fluorescence energies with the corresponding experimental values.

Original languageEnglish
JournalJournal of Computational Chemistry
DOIs
Publication statusAccepted/In press - 2019 Jan 1

Fingerprint

Tight-binding
Divide and conquer
Excited States
Excited states
Density Functional
Unit
Processing
Simulation
Fragment
Data storage equipment
Proton transfer
Program processors
Fluorescence
Divides
Computational Cost
Absorption
Graphics
Scaling
Costs
Energy

Keywords

  • divide-and-conquer method
  • excited-state theory
  • graphical processor unit
  • linear scaling
  • time-dependent density-functional tight-binding method

ASJC Scopus subject areas

  • Chemistry(all)
  • Computational Mathematics

Cite this

@article{1cbb27d81a2045f58d7963eb9658bee4,
title = "GPU-Accelerated Large-Scale Excited-State Simulation Based on Divide-and-Conquer Time-Dependent Density-Functional Tight-Binding",
abstract = "The present study implemented the divide-and-conquer time-dependent density-functional tight-binding (DC-TDDFTB) code on a graphical processing unit (GPU). The DC method, which is a linear-scaling scheme, divides a total system into several fragments. By separately solving local equations in individual fragments, the DC method could reduce slow central processing unit (CPU)-GPU memory access, as well as computational cost, and avoid shortfalls of GPU memory. Numerical applications confirmed that the present code on GPU significantly accelerated the TDDFTB calculations, while maintaining accuracy. Furthermore, the DC-TDDFTB simulation of 2-acetylindan-1,3-dione displays excited-state intramolecular proton transfer and provides reasonable absorption and fluorescence energies with the corresponding experimental values.",
keywords = "divide-and-conquer method, excited-state theory, graphical processor unit, linear scaling, time-dependent density-functional tight-binding method",
author = "Takeshi Yoshikawa and Nana Komoto and Yoshifumi Nishimura and Hiromi Nakai",
year = "2019",
month = "1",
day = "1",
doi = "10.1002/jcc.26053",
language = "English",
journal = "Journal of Computational Chemistry",
issn = "0192-8651",
publisher = "John Wiley and Sons Inc.",

}

TY - JOUR

T1 - GPU-Accelerated Large-Scale Excited-State Simulation Based on Divide-and-Conquer Time-Dependent Density-Functional Tight-Binding

AU - Yoshikawa, Takeshi

AU - Komoto, Nana

AU - Nishimura, Yoshifumi

AU - Nakai, Hiromi

PY - 2019/1/1

Y1 - 2019/1/1

N2 - The present study implemented the divide-and-conquer time-dependent density-functional tight-binding (DC-TDDFTB) code on a graphical processing unit (GPU). The DC method, which is a linear-scaling scheme, divides a total system into several fragments. By separately solving local equations in individual fragments, the DC method could reduce slow central processing unit (CPU)-GPU memory access, as well as computational cost, and avoid shortfalls of GPU memory. Numerical applications confirmed that the present code on GPU significantly accelerated the TDDFTB calculations, while maintaining accuracy. Furthermore, the DC-TDDFTB simulation of 2-acetylindan-1,3-dione displays excited-state intramolecular proton transfer and provides reasonable absorption and fluorescence energies with the corresponding experimental values.

AB - The present study implemented the divide-and-conquer time-dependent density-functional tight-binding (DC-TDDFTB) code on a graphical processing unit (GPU). The DC method, which is a linear-scaling scheme, divides a total system into several fragments. By separately solving local equations in individual fragments, the DC method could reduce slow central processing unit (CPU)-GPU memory access, as well as computational cost, and avoid shortfalls of GPU memory. Numerical applications confirmed that the present code on GPU significantly accelerated the TDDFTB calculations, while maintaining accuracy. Furthermore, the DC-TDDFTB simulation of 2-acetylindan-1,3-dione displays excited-state intramolecular proton transfer and provides reasonable absorption and fluorescence energies with the corresponding experimental values.

KW - divide-and-conquer method

KW - excited-state theory

KW - graphical processor unit

KW - linear scaling

KW - time-dependent density-functional tight-binding method

UR - http://www.scopus.com/inward/record.url?scp=85071013763&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071013763&partnerID=8YFLogxK

U2 - 10.1002/jcc.26053

DO - 10.1002/jcc.26053

M3 - Article

AN - SCOPUS:85071013763

JO - Journal of Computational Chemistry

JF - Journal of Computational Chemistry

SN - 0192-8651

ER -