Dear Stefan
I did an installation of QE 6.2 in a workstation with dual processors hyperthreaded 10 core each running Ubuntu version 16.04 Mate.
I did it using gnu compilers, mpich2 for mpi, and intel mkl.
It runs in parallel without any problems except for relativistic DOS calculations which for some reason only works using netlib math libraries. I can send you in few days the make file and the procedure.
I used the same procedure in installing many versions without any problems even in a cluster running on red hat linux. I always observe near linear scaling with core count.
Yasser Al Wahedi
Assistant Professor
Khalifa University
From: Stefan Seidl
Sent: Thursday, 4 January, 19:02
Subject: [Pw_forum] Installing Quantum Espresso version 6.2 with GFORTRAN and OPENMPI2
To: pw_forum@pwscf.org
Dear all,
can anyone confirm a successful installation of Quantum Espresso version 6.2 on a supercomputer by using gfortran and OpenMPI2 and afterwards running a parallel job successful.
I did the installation "for the impatient" (configure finds gfortran over mpif90):
./configure
make all
Afterwards running a job, provides following message in my output file filename.out:
--------------------------------------------------------------------------
mpiexec noticed that process rank 26 with PID 9977 on node node0389 exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------
Nothing else is written in my outfile. Where could be the mistake ?
But I installed Quantum Espresso version 6.1 "for the impatient":
./configure
make all
Running the same job (as in version 6.2) with the 6.1 binary was successful.
Best,
Stefan Seidl
_______________________________________________ Pw_forum mailing list Pw_forum@pwscf.org http://pwscf.org/mailman/listinfo/pw_forum
