TY - JOUR
T1 - mVMC—Open-source software for many-variable variational Monte Carlo method
AU - Misawa, Takahiro
AU - Morita, Satoshi
AU - Yoshimi, Kazuyoshi
AU - Kawamura, Mitsuaki
AU - Motoyama, Yuichi
AU - Ido, Kota
AU - Ohgoe, Takahiro
AU - Imada, Masatoshi
AU - Kato, Takeo
N1 - Funding Information:
We would like to express our sincere gratitude to Daisuke Tahara for providing us his code of variational Monte Carlo method. A part of mVMC is based on his original code. We also acknowledge Hiroshi Shinaoka, Youhei Yamaji, Moyuru Kurita, Ryui Kaneko, and Hui-Hai Zhao for their cooperations on developing mVMC. We would also like to thank the support from “Project for advancement of software usability in materials science” by Institute for Solid State Physics, University of Tokyo, for development of mVMC ver.1.0. This work was also supported by Grant-in-Aid for Scientific Research ( 16H06345 , 16K17746 , and 18K13477 ) and Building of Consortia for the Development of Human Resources in Science and Technology from the MEXT of Japan . We also thank numerical resources from the Supercomputer Center of Institute for Solid State Physics, University of Tokyo. KI was financially supported by Grant-in-Aid for JSPS Fellows (No. 17J07021 ) and Japan Society for the Promotion of Science through Program for Leading Graduate Schools (MERIT). We thank the computational resources of the K computer provided by the RIKEN Advanced Institute for Computational Science through the HPCI System Research project, as well as the project ”Social and scientific priority issue (Creation of new functional devices and high-performance materials to support next-generation industries; CDMSI)” to be tackled by using post-K computer, under the project number hp160201 , and hp170263 supported by Ministry of Education, Culture, Sports, Science and Technology, Japan (MEXT) .
Funding Information:
We would like to express our sincere gratitude to Daisuke Tahara for providing us his code of variational Monte Carlo method. A part of mVMC is based on his original code. We also acknowledge Hiroshi Shinaoka, Youhei Yamaji, Moyuru Kurita, Ryui Kaneko, and Hui-Hai Zhao for their cooperations on developing mVMC. We would also like to thank the support from “Project for advancement of software usability in materials science” by Institute for Solid State Physics, University of Tokyo, for development of mVMC ver.1.0. This work was also supported by Grant-in-Aid for Scientific Research (16H06345, 16K17746, and 18K13477) and Building of Consortia for the Development of Human Resources in Science and Technology from the MEXT of Japan. We also thank numerical resources from the Supercomputer Center of Institute for Solid State Physics, University of Tokyo. KI was financially supported by Grant-in-Aid for JSPS Fellows (No. 17J07021) and Japan Society for the Promotion of Science through Program for Leading Graduate Schools (MERIT). We thank the computational resources of the K computer provided by the RIKEN Advanced Institute for Computational Science through the HPCI System Research project, as well as the project ”Social and scientific priority issue (Creation of new functional devices and high-performance materials to support next-generation industries; CDMSI)” to be tackled by using post-K computer, under the project number hp160201, and hp170263 supported by Ministry of Education, Culture, Sports, Science and Technology, Japan (MEXT).
Publisher Copyright:
© 2018 The Authors
PY - 2019/2
Y1 - 2019/2
N2 - mVMC (many-variable Variational Monte Carlo) is an open-source software package based on the variational Monte Carlo method applicable for a wide range of Hamiltonians for interacting fermion systems. In mVMC, we introduce more than ten thousands variational parameters and simultaneously optimize them by using the stochastic reconfiguration (SR) method. In this paper, we explain basics and user interfaces of mVMC. By using mVMC, users can perform the calculation by preparing only one input file of about ten lines for widely studied quantum lattice models, and can also perform it for general Hamiltonians by preparing several additional input files. We show the benchmark results of mVMC for the Hubbard model, the Heisenberg model, and the Kondo-lattice model. These benchmark results demonstrate that mVMC provides ground-state and low-energy-excited-state wave functions for interacting fermion systems with high accuracy. Program summary: Program title: mVMC Program Files doi: http://dx.doi.org/10.17632/xhgyp6ncvt.1 Licensing provisions: GNU General Public License version 3 Programming language: C External routines/libraries: MPI, BLAS, LAPACK, Pfapack, ScaLAPACK (optional) Nature of problem: Physical properties (such as the charge/spin structure factors) of strongly correlated electrons at zero temperature. Solution method: Application software based on the variational Monte Carlo method for quantum lattice model such as the Hubbard model, the Heisenberg model and the Kondo model. Unusual features: It is possible to perform the highly-accurate calculations for ground states in a wide range of theoretical Hamiltonians in quantum many-body systems. In addition to the conventional orders such as magnetic and/or charge orders, user can treat the anisotropic superconductivities within the same framework. This flexibility is the main advantage of mVMC.
AB - mVMC (many-variable Variational Monte Carlo) is an open-source software package based on the variational Monte Carlo method applicable for a wide range of Hamiltonians for interacting fermion systems. In mVMC, we introduce more than ten thousands variational parameters and simultaneously optimize them by using the stochastic reconfiguration (SR) method. In this paper, we explain basics and user interfaces of mVMC. By using mVMC, users can perform the calculation by preparing only one input file of about ten lines for widely studied quantum lattice models, and can also perform it for general Hamiltonians by preparing several additional input files. We show the benchmark results of mVMC for the Hubbard model, the Heisenberg model, and the Kondo-lattice model. These benchmark results demonstrate that mVMC provides ground-state and low-energy-excited-state wave functions for interacting fermion systems with high accuracy. Program summary: Program title: mVMC Program Files doi: http://dx.doi.org/10.17632/xhgyp6ncvt.1 Licensing provisions: GNU General Public License version 3 Programming language: C External routines/libraries: MPI, BLAS, LAPACK, Pfapack, ScaLAPACK (optional) Nature of problem: Physical properties (such as the charge/spin structure factors) of strongly correlated electrons at zero temperature. Solution method: Application software based on the variational Monte Carlo method for quantum lattice model such as the Hubbard model, the Heisenberg model and the Kondo model. Unusual features: It is possible to perform the highly-accurate calculations for ground states in a wide range of theoretical Hamiltonians in quantum many-body systems. In addition to the conventional orders such as magnetic and/or charge orders, user can treat the anisotropic superconductivities within the same framework. This flexibility is the main advantage of mVMC.
KW - Lattice fermion models
KW - Numerical linear algebra
KW - Variational Monte Carlo method
UR - http://www.scopus.com/inward/record.url?scp=85053844476&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85053844476&partnerID=8YFLogxK
U2 - 10.1016/j.cpc.2018.08.014
DO - 10.1016/j.cpc.2018.08.014
M3 - Article
AN - SCOPUS:85053844476
SN - 0010-4655
VL - 235
SP - 447
EP - 462
JO - Computer Physics Communications
JF - Computer Physics Communications
ER -