A numerical solver package for a wide range of quantum lattice models based on many-variable Variational Monte Carlo method
====
mVMC (many-variable Variational Monte Carlo method)
is a software for performing the highly-accurate
variational Monte Carlo calculations
with the simple and flexible user interface.
mVMC also supports the large-scale parallelization.
For the conventional models in strongly correlated electron systems such as the Hubbard model, the Heisenberg model, and the Kondo-lattice model,
users can perform the calculation by preparing the one input files whose length is shorter than ten lines.
By using the same input file, users can perform the exact diagonalization through HPhi.
Thus, it is easy to examine the accuracy of the variational calculation for small system sizes
and to perform the calculations
for large system sizes that can not be treated
by the exact diagonalization.
A broad spectrum of users including experimental scientists is cordially welcome.
many-variable variational Monte Carlo method
Hubbard model, Heisenberg model, Kondo lattice model, multi-orbital Hubbard model
specific heat, susceptibility, ground state energy, structure factors
- C compiler (intel, Fujitsu, GNU, etc. )
- ScaLAPACK library (intel MKL, Fujitsu, ATLAS, etc.)
- MPI library
You can install mVMC and also get a manual for mVMC from a release note.
GNU General Public License version 3 (GPL v3).
The mVMC package is developed based on the mVMC-mini program. The license of mVMC-mini is "The BSD 3-Clause License".
We hope that you cite the following URL when you publish the results using mVMC (after writing the paper for mVMC, we will replace the URL to the paper's reference):
URL: https://github.com/issp-center-dev/mVMC
Lecture notes and sample scripts used in Hands-on are available at mVMC-tutorial
Takahiro Misawa, Satoshi Morita, Takahiro Ohgoe, Kota Ido, Yuichi Motoyama, Mitsuaki Kawamura, Kazuyoshi Yoshimi, Takeo Kato, Masatoshi Imada.