VASP requires that the BLAS package is installed on the computer.

VASP-gnu

Not installed currently

VASP-gnu Install Directory

No Format
cdmodule load /sw/VASP/AIBN/5.2/vasp.5.lib-gnu

VASP-mpi-intel

Usage:

No Format
module load vasp/5.2vasp-mpi-intel

Sample pbs script

No Format

#!/bin/bash -l
#PBS -m abe
#PBS -V
### Mail to user
#PBS -M emailAddress@griffith.edu.au
### Job name
#PBS -N test_vasp
#PBS -W group_list=vasp
#PBS -l walltime=00:30:00
### Number of nodes:Number of CPUs:Number of threads per node
#PBS -l select=2:ncpus=2:mpiprocs=2:mem=8g
# This job's working directory
echo Working directory is $PBS_O_WORKDIR
cd $PBS_O_WORKDIR
source $HOME/.bashrc
module load vasp/5.2vasp-mpi-intel
module list
echo "Starting job"
echo Running on host `hostname`
echo Time is `date`
echo Directory is `pwd`
#echo This jobs runs on the following processors:
echo `cat $PBS_NODEFILE`
##mpirun  -machinefile   $PBS_NODEFILE  /export/home/snumber/pbs/mpi/p1.out
/sw/sdev/intel/mpi/4.0.0.027/x86_64/intel64/bin/mpirun -r ssh -f  $PBS_NODEFILE  vasp
echo "Done with job"

Install instructions

Instructions
============

change the Makefile for vasp and libs to use ifort and mpiifort
make sure the right links to mkl libraries are put in the appropriate place
load the following modules
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel
Just do make first in the lib and then in the vasp directory.

vasp.5.lib-mpi-intel

No Format
cd /sw/VASP/5.2.12vasp/vasp541-gnu-openmpi To use vasp on the cluster, one needs to be in the unix group named "vasp". Please contact the HPC admin to request this.

VASP-gnu

Not installed currently

VASP-gnu Install Directory

No Format
cd /sw/VASP/AIBN/5.2/vasp.5.lib-gnu

VASP-mpi-intel

Usage:

No Format
module load vasp/5.2vasp-mpi-intel

Sample pbs script

No Format

#!/bin/bash -l
#PBS -m abe
#PBS -V
### Mail to user
#PBS -M emailAddress@griffith.edu.au
### Job name
#PBS -N test_vasp
#PBS -W group_list=vasp
#PBS -l walltime=00:30:00
### Number of nodes:Number of CPUs:Number of threads per node
#PBS -l select=2:ncpus=2:mpiprocs=2:mem=8g
# This job's working directory
echo Working directory is $PBS_O_WORKDIR
cd $PBS_O_WORKDIR
source $HOME/.bashrc
module load vasp/5.2vasp-mpi-intel
module list
echo "Starting job"
echo Running on host `hostname`
echo Time is `date`
echo Directory is `pwd`
#echo This jobs runs on the following processors:
echo `cat $PBS_NODEFILE`
##mpirun  -machinefile   $PBS_NODEFILE  /export/home/snumber/pbs/mpi/p1.out
/sw/sdev/intel/mpi/4.0.0.027/x86_64/intel64/bin/mpirun -r ssh -f  $PBS_NODEFILE  vasp
echo "Done with job"

Install instructions

Instructions
============

change the Makefile for vasp and libs to use ifort and mpiifort
make sure the right links to mkl libraries are put in the appropriate place
load the following modules
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel
Just do make first in the lib and then in the vasp directory.

vasp.5.lib-mpi-intel

No Format

cd /sw/VASP/5.2.12/vasp.5.lib-mpi-intel
cp makefile.linux_ifc_P4 Makefile

module purge
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel

 module list
Currently Loaded Modulefiles:
  1) intel-mpi/4.0.0.027       2) intel-cc-11/11.1.072      3) intel-fc-11/11.1.072      4) intel-cmkl-11/11.1.072    5) intel-tools-11/11.1.072   6) fftw/3.2.2-intel


vi Makefile

Change the following to....
>>>>>>>>>>>>>
# C-preprocessor
CPP     = mpiicc -E -P -C $*.F >$*.f
FC=mpiifort
>>>>>>>>>>>>>>

make 2>&1 |tee make_vasp_lib_mpi_intel.txt

vasp.5.2mpi-intel

No Format

module purge
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel

cd /sw/VASP/5.2.12/vasp.5.2mpi-intel
cp makefile.linux_ifc_P4 Makefile



comment out the following in Makefile:

Item #

Pre Change

Post Change

1

-

FC=mpiifort

2

-

No Format
BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread

3

-

No Format
LAPACK= -lmkl_lapack

4

-

No Format
FFT3D = fft3dfurth.o fft3dlib.o

5

-

No Format
FC=mpiifort

No Format
FCL=$(FC)

6

-

No Format
CPP = $(CPP_) -DMPI -DHOST=\"LinuxIFC\" -DIFC \

No Format
-DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \

No Format
-DMPI_BLOCK=8000

7

-

No Format
LIB = -L../vasp.5.lib-mpi-intel

...

-ldmy  \

...

vasp.5.2mpi-intel

No Format

module purge
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel

cd /sw/VASP/5.2.12/vasp.5.2mpi-intel
cp makefile.linux_ifc_P4 Makefile



comment out the following in Makefile:

Item #

Pre Change

Post Change

1

-

FC=mpiifort

2

-

No Format
BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread

3

-

No Format
LAPACK= -lmkl_lapack

4

-

No Format
FFT3D = fft3dfurth.o fft3dlib.o

5

-

No Format
FC=mpiifort

No Format
FCL=$(FC)

6

-

No Format

     ../vasp.5.lib-mpi-intel/linpack_double.o $(LAPACK) \
      $(SCA) $(BLAS)

8

-

No Format
FFT3D = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o

No Format

 diff Makefile makefile.linux_ifc_P4
53c53
< FC=mpiifort
---
> FC=ifc
126,128d125
< #####BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl -lpthread
< ##BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl -lpthread
< BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
133c130
< ##BLAS=  /opt/libs/libgoto/libgoto.so
---
> BLAS=  /opt/libs/libgoto/libgoto.so
136c133
< ##LAPACK= ../vasp.5.lib/lapack_double.o
---
> LAPACK= ../vasp.5.lib/lapack_double.o
139c136
< LAPACK= -lmkl_lapack
---
> #LAPACK= -lmkl_lapack
143,144c140,141
< LIB  = -L../vasp.5.lib-intel -ldmy \
<      ../vasp.5.lib-intel/linpack_double.o $(LAPACK) \
---
> LIB  = -L../vasp.5.lib -ldmy \
>      ../vasp.5.lib/linpack_double.o $(LAPACK) \
175,177c172,173
< ##FC=mpif77
< FC=mpiifort
< FCL=$(FC)
---
> #FC=mpif77
> #FCL=$(FC)
194,197c190,193
< CPP    = $(CPP_) -DMPI  -DHOST=\"LinuxIFC\" -DIFC \

...


<      -DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \
<

...

   -DMPI_BLOCK=8000

...

7

...

-

...


< #    -DRPROMU_DGEMV  -DRACCMU_DGEMV
---
> #CPP    =

...

 $(CPP_) -DMPI  -DHOST=\"LinuxIFC\" -DIFC \
> #

...

8

...

-

...

No Format
FFT3D = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o

No Format

 diff Makefile makefile.linux_ifc_P4
53c53
< FC=mpiifort
---
> FC=ifc
126,128d125
< #####BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl -lpthread
< ##BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl -lpthread
< BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
133c130
< ##BLAS=  /opt/libs/libgoto/libgoto.so-DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \
> #     -DMPI_BLOCK=8000
> ##    -DRPROMU_DGEMV  -DRACCMU_DGEMV
216,218c212,214
< LIB     = -L../vasp.5.lib-mpi-intel -ldmy  \
<       ../vasp.5.lib-mpi-intel/linpack_double.o $(LAPACK) \
<       $(SCA) $(BLAS)
---
> BLAS=#LIB  /opt/libs/libgoto/libgoto.so 136c133 < ##LAPACK= -L../vasp.5.lib/lapack_double.o ---ldmy  \
> LAPACK= #      ../vasp.5.lib/lapacklinpack_double.o 139c136
< LAPACK= -lmkl_lapack
---$(LAPACK) \
> #LAPACK= -lmkl_lapack
143,144c140,141
< LIB  = -L../vasp.5.lib-intel -ldmy \
<      ../vasp.5.lib-intel/linpack_double.o $(LAPACK) \#      $(SCA) $(BLAS)
221c217
< FFT3D   = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o
---
> LIB#FFT3D   = -L../vasp.5.lib -ldmy \
>      ../vasp.5.lib/linpack_double.o $(LAPACK) \
175,177c172,173
< ##FC=mpif77
< FC=mpiifort
< FCL=$(FC)
---
> #FC=mpif77
> #FCL=$(FC)
194,197c190,193
< CPP    = $(CPP_) -DMPI  -DHOST=\"LinuxIFC\" -DIFC \
<      -DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \
<      -DMPI_BLOCK=8000
< #    -DRPROMU_DGEMV  -DRACCMU_DGEMV
---
> #CPP    = $(CPP_) -DMPI  -DHOST=\"LinuxIFC\" -DIFC \
> #     -DCACHE_SIZE=4000 -DPGF90 -Davoidalloc -DNGZhalf \
> #     -DMPI_BLOCK=8000
> ##    -DRPROMU_DGEMV  -DRACCMU_DGEMV
216,218c212,214
< LIB     = -L../vasp.5.lib-mpi-intel -ldmy  \
<       ../vasp.5.lib-mpi-intel/linpack_double.o $(LAPACK) \
<       $(SCA) $(BLAS)
---
> #LIB     = -L../vasp.5.lib -ldmy  \
> #      ../vasp.5.lib/linpack_double.o $(LAPACK) \
> #      $(SCA) $(BLAS)
221c217
< FFT3D   = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o
---
> #FFT3D   = fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o

No Format
make 2>&1 \|tee make_vasp52_mpi_intel.txt

vasp-serial

vasp.5.lib-intel

No Format
fftmpi.o fftmpi_map.o fft3dfurth.o fft3dlib.o

No Format
make 2>&1 \|tee make_vasp52_mpi_intel.txt

vasp-serial

vasp.5.lib-intel

No Format

cd /sw/VASP/5.2.12
tar -zxvf /data1/vasp.5.lib.tar.gz
mv vasp.5.lib vasp.5.lib-intel
cd /sw/VASP/5.2.12/vasp.5.lib-intel
cp makefile.linux_ifc_P4 Makefile

module purge
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel

vi Makefile

Change the following ....
>>>>>>>>>>>>>
 diff Makefile makefile.linux_ifc_P4
18,19c18,19
< CPP     = icc -E -P -C $*.F >$*.f
< FC=ifort
---
> CPP     = gcc -E -P -C $*.F >$*.f
> FC=ifc
51c51
<       icc $(CFLAGS) -c $*.c
---
>       $(CC) $(CFLAGS) -c $*.c
>>>>>>>>>>>>>>

make 2>&1 |tee make_vasp_lib_intel.txt

vasp.5.2-intel

No Format

module purge
module load intel-cc-11/11.1.072
module load fftw/3.2.2-intel
module load  intel-cmkl-11/11.1.072


module load intel-tools-11/11.1.072
module load  fftw/3.3-alpha-intel
module unload  intel-mpi/4.0.0.027
 module list
Currently Loaded Modulefiles:
  1) intel-cc-11/11.1.072      2) intel-fc-11/11.1.072      3) intel-cmkl-11/11.1.072    4) intel-tools-11/11.1.072   5) fftw/3.3-alpha-intel

cd /sw/VASP/5.2.12
tar -zxvf /data1/vasp.5.lib2.12.tar.gz
mv vasp.5.lib2 vasp.5.lib2-intel
cd /sw/VASP/5.2.12/vasp.5.lib2-intel
cp makefile.linux_ifc_P4 Makefile

>>>>>>>
 diff Makefile makefile.linux_ifc_P4
53c53

module purge
module load intel-tools-11< FC=ifort
---
> FC=ifc
126d125
< BLAS=-L/sw/sdev/intel/Compiler/11.1.072
module load fftw/3.2.2-intel

vi Makefile

Change the following ....
>>>>>>>>>>>>>
 diff Makefile makefile.linux_ifc_P4
18,19c18,19
< CPP     = icc -E -P -C $*.F >$*.f
< FC=ifort
---
> CPP     = gcc -E -P -C $*.F >$*.f
> FC=ifc
51c51
<       icc $(CFLAGS) -c $*.c
---/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
131c130
< #BLAS=  /opt/libs/libgoto/libgoto.so
---
> BLAS=  /opt/libs/libgoto/libgoto.so
134c133
< LAPACK= ../vasp.5.lib-intel/lapack_double.o
---
> LAPACK= ../vasp.5.lib/lapack_double.o
141,142c140,141
< LIB  = -L../vasp.5.lib-intel -ldmy \
<      ../vasp.5.lib-intel/linpack_double.o $(LAPACK) \
---
> LIB  = -L../vasp.5.lib -ldmy \
>      ../vasp.5.lib/linpack_double.o $(CCLAPACK) $(CFLAGS) -c $*.c
>>>>>>>>>>>>>>\

>>>>>>>>>

make 2>&1 |tee make_vaspvasp52_libserial_intel.txt

...

VASP-5.

...

4.

...

1

No Format

module purge
module load intel-ccmpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel
module load  intel-cmkl-11/11.1.072


module load intel-tools-11/11.1.072
module load  fftw/3.3-alpha-intel
module unload  intel-mpi/4.0.0.027
 module list
Currently Loaded Modulefiles:
  1) intel-cc-11/11.1.072      2) intel-fc-11/11.1.072      3) intel-cmkl-11/11.1.072    4) intel-tools-11/11.1.072   5) fftw/3.3-alpha-intel

cd /sw/VASP/5.2.12
tar -zxvf /data1/vasp.5.2.12.tar.gz
mv vasp.5.2 vasp.5.2-intel
cd /sw/VASP/5.2.12/vasp.5.2-intel
cp makefile.linux_ifc_P4 Makefile

>>>>>>>
 diff Makefile makefile.linux_ifc_P4
53c53
< FC=ifort
---
> FC=ifc
126d125
< BLAS=-L/sw/sdev/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_lapack -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread
131c130
< #BLAS=  /opt/libs/libgoto/libgoto.so
---
> BLAS=  /opt/libs/libgoto/libgoto.so
134c133
< LAPACK= ../vasp.5.lib-intel/lapack_double.o
---
> LAPACK= ../vasp.5.lib/lapack_double.o
141,142c140,141
< LIB  = -L../vasp.5.lib-intel -ldmy \
<      ../vasp.5.lib-intel/linpack_double.o $(LAPACK) \
---
> LIB  = -L../vasp.5.lib -ldmy \
>      ../vasp.5.lib/linpack_double.o $(LAPACK) \

>>>>>>>>>

make 2>&1 |tee make_vasp52_serial_intel.txt

VASP-5.4.1

No Format

module purge
module load intel-mpi/4.0.0.027
module load intel-tools-11/11.1.072
module load fftw/3.2.2-intel
cd /sw/VASP/5.4.1/intel-serial
cp arch/makefile.include.linux_intel_serial makefile.include
vi makefile.include

Add this: MKLROOT=/sw/sdev/intel/Compiler/11.1/072/mkl
make

This will build the standard, gamma-only, and non-collinear version
of VASP one after the other.

Patch
=====
module load  vasp/5.4.1vasp-intel-serial
cd /sw/VASP/5.4.1/intel-serial
gunzip the patch
e.g gunzip patch.5.4.1.03082016.gz
patch -p0 < patch.5.4.1.14032016
patch -p0 <patch.5.4.1.03082016 .1.072
module load fftw/3.2.2-intel
cd /sw/VASP/5.4.1/intel-serial
cp arch/makefile.include.linux_intel_serial makefile.include
vi makefile.include

Add this: MKLROOT=/sw/sdev/intel/Compiler/11.1/072/mkl
make

This will build the standard, gamma-only, and non-collinear version
of VASP one after the other.

Patch
=====
module load  vasp/5.4.1vasp-intel-serial
cd /sw/VASP/5.4.1/intel-serial
gunzip the patch
e.g gunzip patch.5.4.1.03082016.gz
patch -p0 < patch.5.4.1.14032016
patch -p0 <patch.5.4.1.03082016 


Special Patch (for Tim):
Replace the files in src by the updated ones from patch_vasp.5.4.1.05Feb16.tgz (including the newly created file tim_parameters.F) and replace/modify also the "hidden" file .objects so as to contain the item tim_parameters.o right after subdft3.o.

VASP 5.4.4

Note: Awoonga is a different HPC managed by QRIS. Awoonga has been replaced by a newer HPC called Tinaroo. Information as of 18/01/23; Tinaroo is soon to be replaced by a newer HPC called Bunya. Bunya will have VASP 6. VASP 5.4.4 will not be available on Bunya. View FAQ Qs38 for details on Awoonga: FAQ - Griffith HPC Cluster#GriffithHPCCluster-Qs38%3AHowdoIgetstartedontheawoongacluster

No Format

awoonga node only
module load mkl/2017.3
module load intel/2017.4
module load lapack/3.6.0
cp arch/makefile.include.linux_intel_serial makefile.include
vi makefile.include

Add this: 
MKLROOT=/opt/intel/composer_xe_2017.4/mkl
make 2>&1 | tee makeLog.txt

Sample PBS script

No Format

#!/bin/bash
#PBS -q workq
#PBS -m bea
#PBS -M email@griffithuni.edu.au
#PBS -N VaspTest
#PBS -l select=1:ncpus=4:mpiprocs=4:mem=35gb,walltime=4:00:00
## load vasp module to setup env variables to run vasp
cd $PBS_O_WORKDIR
module load vasp/vasp541-gnu-openmpi
## get the number of allocated cpu cores
export nprocs=`cat $PBS_NODEFILE | wc -l`
## run vasp in parallel over $nprocs processors
###alias vasp="vasp_std"
mpiexec vasp_std < /dev/null > vasp.debug;
echo "Done with job"

Another Sample PBS Script

No Format

#!/bin/bash
#PBS -m abe
#PBS -M EMAIL@griffith.edu.au
#PBS -N WestPap
#PBS -q workq
#PBS -l select=1:ncpus=1:mem=5gb,walltime=4:00:00
cd  $PBS_O_WORKDIR
module load vasp/vasp541-gnu-openmpi
echo "Starting job"
vasp_std < /dev/null > vasp.debug
echo "Done with Job"

Versions Compared

Old Version 4

New Version Current

Key

VASP-gnu

VASP-gnu Install Directory

VASP-mpi-intel

Usage:

Sample pbs script

Install instructions

vasp.5.lib-mpi-intel

VASP-gnu

VASP-gnu Install Directory

VASP-mpi-intel

Usage:

Sample pbs script

Install instructions

vasp.5.lib-mpi-intel

vasp.5.2mpi-intel

vasp.5.2mpi-intel

vasp-serial

vasp.5.lib-intel

vasp-serial

vasp.5.lib-intel

vasp.5.2-intel

VASP-5.

4.

1

VASP-5.4.1

VASP 5.4.4

Sample PBS script

Another Sample PBS Script

Page Comparison

Versions Compared

Old Version 4

New Version Current

Key

VASP-gnu

VASP-gnu Install Directory

VASP-mpi-intel

Usage:

Sample pbs script

Install instructions

vasp.5.lib-mpi-intel

VASP-gnu

VASP-gnu Install Directory

VASP-mpi-intel

Usage:

Sample pbs script

Install instructions

vasp.5.lib-mpi-intel

vasp.5.2mpi-intel

vasp.5.2mpi-intel

vasp-serial

vasp.5.lib-intel

vasp-serial

vasp.5.lib-intel

vasp.5.2-intel

VASP-5.

4.

1

VASP-5.4.1

VASP 5.4.4

Sample PBS script

Another Sample PBS Script