Dear Paolo,
Once again, thank you very much for the quick reply. I considered your last
suggestion, but before of doing so, I tried to compile qe using intel compilers
and (to my surprise) it finally succeeded!.
Since the procedure that worked for me include details not described so far in
the web, here I try to share a detailed explanation for the compilation of
qe5.1-lbeef (libbeef) that worked for me, in case somebody else face similar
issues in the future:
OS: Ubuntu v.18.04.1 (64bit Virtual machine guest on Windows 10 host –
VirtualBox6.0)
icc : Intel® Parallel Studio XE 2019 Update 4 Cluster Eddition for Linux
1. (after installing intel compilers) Initialize the intel compiler:
Initiallize the icc, ifortran in the working terminal. Stepped at e.g.
"/intel/bin", type
source compilervars.sh intel64
source iccvars.sh intel64
source ifortvars.sh intel64
2. lbeef configuration-installation:
Check the icc compiler is active in the working terminal (should be icc..)
icc --version
Configuration of lbeef 0.1.1.
Installation/compilation of lbeef
./configure CC=icc
--prefix=/home/jorge/aist-qe-installations/37/libbeef-0.1.1
make
make install
3. Compilation of qe5.1-patched/lbeef
Configuration (A detailed tutorial about the compilation of qe using intel
compilers can be find in: https://www.youtube.com/watch?v=doudMLEaq3w)
./configure MPIF90=mpiifort CC=mpiicc F90=ifort F77=mpiifort
-enable-parallel
BEEF_LIBS="-L/home/jorge/aist-qe-installations/37/libbeef-0.1.1/lib -lbeef"
Edit the make.sys file generated after the configuration with the corresponding
Intel MKL libraries:
BLAS_LIBS = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64
-lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64
-lpthread -lm -ldl
LAPACK_LIBS = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64
-lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -ldl
SCALAPACK_LIBS = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64
-lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -ldl
FFT_LIBS = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64
-lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -ldl
MPI_LIBS = -L/home/intel/impi/2019.4.243/intel64/lib/ -lmpi
MASS_LIBS = -L${MKLROOT}/lib/intel64 -lmkl_scalapack_lp64 -lmkl_intel_lp64
-lmkl_sequential -lmkl_core -lmkl_blacs_intelmpi_lp64 -lpthread -lm -ldl
Note: the Intel Math Kernel Libraries are specific for the system used
(version if icc, OS, etc), please see to the following video for a detailed
explanation of this step: https://www.youtube.com/watch?v=doudMLEaq3w);
remember to save the file!
compilation of qe:
make pw
generation of kernel tables
bin/generate_rVV10_kernel_table.x
bin/generate_vdW_kernel_table.x
4. Running a calculation
Note: The generated "generate_vdW_kernel_table", table (located also in the qe
folder) must be copied into the folder containig the input file to be run
Note: Remember to have intialized icc/intel compiler and libraries in the
working terminal
check the intel mpirun, must be icc …
which mpirun
run (synthax)
mpirun
'/home/jorge/aist-qe-installations/37/espresso.5.1.r11289.pybeef/espresso/bin/pw.x'
-in test20.in> test20.out
Regards,
Jorge
-----
Jorge A. Delgado, Ph.D.
Research Scientist - Catalysis and Nanomaterials
Flow Chemistry Team, Catalytic Chemistry Research Center (IRC3)
National Institute of Advanced Industrial Science and Technology (AIST)
Central 5-2, 1-1-1 Higashi, Tsukuba, Ibaraki, 305-8565, Japan
Phone: +81 80-2151-8731
Email: [email protected]<mailto:[email protected]>
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list [email protected]
https://lists.quantum-espresso.org/mailman/listinfo/users
