====== Quantum Espresso 6.5 ======
==== description ====
Quantum ESPRESSO is an integrated suite of Open-Source computer codes for electronic-structure calculations and materials modeling at the nanoscale.
It is based on density-functional theory, plane waves, and pseudopotentials. ( https://www.quantum-espresso.org/project/manifesto ) .

==== software version ====
Quantum Espresso 6.5, https://github.com/QEF/q-e/releases . 

==== prepare software ====
To prepare Quantum Espresso (ver. 6.5) software login to tcr.cent.uw.edu.pl.

Then open interactive session on any computing node with :
  srun -n16 -N1 --pty bash -l

When interactive session is started go through Quantum Espresso installation process described with bellow commands. It will take about 12 minutes.
  #folder for source files
  mkdir -p ~/downloads/quantum_espresso_6.5
  #folder for compiled binares
  mkdir -p ~/soft/qe-6.5
  
  cd ~/downloads/quantum_espresso_6.5
  wget https://github.com/QEF/q-e/releases/download/qe-6.5/qe-6.5-ReleasePack.tgz
  tar xvzf qe-6.5-ReleasePack.tgz
  cd qe-6.5
  module load mpi/openmpi-x86_64
  
  ./configure --prefix="/home/users/${USER}/soft/qe-6.5"
  
  time make -j${SLURM_NTASKS} all 
  make install

remember to end interactive session with ''exit'' command.

If no errors occurred, compiled Quantum Espresso binaries are available in ''/home/users/${USER}/soft/qe-6.5/bin'' .

==== sbatch example ====
Use Quantum Espresso 6.5. This description assumes that path to binaries is ''/home/users/${USER}/soft/qe-6.5/bin'' . Use bellow ''qe-test.sbatch'' file to run computation.
  #!/bin/bash -l
  #SBATCH --job-name="qe-test_N2_n32"
  #SBATCH --nodes=2                   # number of computing_nodes
  #SBATCH --ntasks=32                 # number of CPU's ( 16*computing_nodes )
  #SBATCH --mem-per-cpu=2G
  #SBATCH --partition=short
  #SBATCH --constraint=intel
  #SBATCH --exclusive
  #SBATCH --time=2:00:00
  
  WORKDIR="/home/users/${USER}/soft_tests/qe_run_`date +%s`_${RANDOM}/"
  mkdir -p ${WORKDIR}
  cd ${WORKDIR}
  
  export BIN_DIR="/home/users/${USER}/soft/qe-6.5/bin"
  export PATH=${BIN_DIR}:$PATH
  export PSEUDO_DIR=${WORKDIR}
  export TMP_DIR="/tmp"
     
  #copy input files and pseudo files to ${WORKDIR} 
  cp /home/users/${USER}/downloads/quantum_espresso_input_files/* ${WORKDIR}
    
  module load mpi/openmpi-x86_64
  
  T1=`date +%s`
    
  mpirun -np ${SLURM_NTASKS} pw.x -npool ${SLURM_NNODES} -inp Ti2N.in > Ti2N.out
    
  T2=`date +%s`
  echo -e "stop ${T2}\t start ${T1}\t ${SLURM_NNODES}"

==== performance tests ====
Bellow results show time of computation in function of used resources (computation scalability) for a specific computational task done with ''pw.x'' program.

Assigning a larger amount of computing nodes does not always lead to a (efficient) reduction in computing time (wall-time of the job). To find the most appropriate number of nodes for a specific type of job, it is essential to run one's own benchmarks. In general, parallel <color #22b14c>jobs should scale to at least 70% efficiency for the sake of other TCR users</color>. One user using twice the resources to squeeze out 10% more performance may be keeping other users from working at all.

Bellow results should be consider as results of this specific computational task on this specific hardware (TCR cluster) and not overall benchmark for Quantum Espresso software suite.

{{ tcr:qe65_tcr_result_001.png?nolink|}}
^nodes	^min [s] 	^avg [s] 	^median [s] 	^max [s] 	^efficiency [%] ^
|1	|1862	|2065.75	|2126	|2149	|<color #22b14c>100.00%</color>|
|2	|1105	|1192.75	|1157.5	|1351	|<color #22b14c>84.25%</color>|
|3	|763	|770	|768.5	|780	|<color #22b14c>81.35%</color>|
|4	|798	|1026.25	|868.5	|1570	|58.33%|
|5	|571	|589.75	|574	|640	|65.22%|
|6	|488	|536	|525.5	|605	|63.59%|
|7	|372	|502.25	|414.5	|808	|<color #22b14c>71.51%</color>|
|8	|375	|489.75	|408.5	|767	|62.07%|
|9	|324	|403.75	|346.5	|598	|63.85%|
|10	|327	|456.25	|465.5	|567	|56.94%|
|12	|285	|429.4	|375	|609	|54.44%|
|16	|228	|274.25	|281.5	|306	|51.04%|




*) values (min, avg, median, max, efficiency) do not include failed runs \\
*) efficiency as ''t1 / ( nodes * tn )'' ( where t1 is min computation time at one node, tn is min computation time on N nodes ) \\