WIEN2k
===================================

.. attention::

   WIEN2k is available for users who have a license. If you have the license and want to use WIEN2k, contact our `center <https://www.sc.imr.tohoku.ac.jp/eng/contact.html>`_ . After checking the license, WIEN2K is ready for you. 

Following version is available. 

.. csv-table:: Large-scale Parallel Computing Server
   :header: "version", "module","execution queue"
   :widths: 20, 79, 1

   "24.1", "/work/app/WIEN2k/WIEN2k_24.1/run_lapw", "P_030 TP_002 MP_001 CP_001 DP_002 S_001 CS_001"

.. attention::
  Execute following command in advance.

  module load oneapi/2025.0.1

  WORKDIR=/work/scratch/$USER/$PBS_JOBID
  
  mkdir -p $WORKDIR

  export SCRATCH=$WORKDIR 

  export TMPDIR=$WORKDIR

  export WIENROOT=/work/app/WIEN2k/WIEN2k_24.1 

  export PATH=$WIENROOT:$PATH 



- Job Submission Script

.. code-block :: none

   #!/bin/sh 
   #PBS -l select=nodes
   #PBS -q queue
   #PBS -N jobname

   module load oneapi/2025.0.1 2> /dev/null

   WORKDIR=/work/scratch/$USER/$PBS_JOBID
   mkdir -p $WORKDIR

   export SCRATCH=$WORKDIR  
   export TMPDIR=$WORKDIR
   export WIENROOT=/work/app/WIEN2k/WIEN2k_24.1/
   export PATH=$WIENROOT:$PATH 

   cd ${PBS_O_WORKDIR}

   wien2k execution script option > output file 2> error file


- Example

.. code-block :: none

   #!/bin/sh 
   #PBS -l select=1 
   #PBS -q P_030 
   #PBS -N wien2k 

   module load oneapi/2025.0.1 2> /dev/null

   WORKDIR=/work/scratch/$USER/$PBS_JOBID
   mkdir -p $WORKDIR

   export SCRATCH=${PBS_O_WORKDIR} 
   export TMPDIR=$WORKDIR 
   export WIENROOT=/work/app/WIEN2k/WIEN2k_24.1/
   export PATH=$WIENROOT:$PATH 

   cd ${PBS_O_WORKDIR}
   
   run_lapw -p -cc 0.0001 -NI > wien2k.out 2> wien2k.err


You can use k-Point parallelization by using -p option in script file and set up .machines file in the current work directory.

.. code-block :: none

   $ cat .machines 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost
   1:localhost 
   1:localhost 
   1:localhost 
   1:localhost 
   granularity:1 
   extrafine:1

.. attention::

   If you want to run processes in parallel, you need to have the same number of localhost entries in the .machines file as the number of parallel processes.