Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Ralph Castain via users
toph - Original Message - From: "Open MPI Users" mailto:users@lists.open-mpi.org> > To: "Open MPI Users" mailto:users@lists.open-mpi.org> > Cc: "Carlo Nervi" mailto:carlo.ne...@unito.it> > Sent: Thursday, 20 August, 2020 12:17:21 Subject: [OMPI u

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Carlo Nervi via users
I'm using VASP, Quantum Espresso, DFTB+, Gulp, Tinker, Crystal and Gaussian. VASP, QE and G16 are not a problem (the latter is using threads up to 48 cores). QE sometimes slows down, but nothing to be much worry. DFTB+ is often used with several jobs and mpi. In any cases jobs x mpi <= 48 I'm

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Carlo Nervi via users
t all your simulations in a single script this would look >> like >> >> mpirun -n 6 --cpu-list "$(seq -s, 0 5)" --bind-to cpu-list:ordered $app >> mpirun -n 6 --cpu-list "$(seq -s, 6 11)" --bind-to cpu-list:ordered $app >> ... >> mpirun -

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Jeff Hammond via users
On Thu, Aug 20, 2020 at 3:22 AM Carlo Nervi via users < users@lists.open-mpi.org> wrote: > Dear OMPI community, > I'm a simple end-user with no particular experience. > I compile quantum chemical programs and use them in parallel. > Which code? Some QC codes behave differently than traditional

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Ralph Castain via users
--cpu-list "$(seq -s, 42 47)" --bind-to cpu-list:ordered  $app Best Christoph - Original Message - From: "Open MPI Users" mailto:users@lists.open-mpi.org> > To: "Open MPI Users" mailto:users@lists.open-mpi.org> > Cc: "Carlo Nervi" mailto

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Carlo Nervi via users
t; Best > Christoph > > > - Original Message - > From: "Open MPI Users" > To: "Open MPI Users" > Cc: "Carlo Nervi" > Sent: Thursday, 20 August, 2020 12:17:21 > Subject: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa > archi

Re: [OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Christoph Niethammer via users
, 6 11)" --bind-to cpu-list:ordered $app ... mpirun -n 6 --cpu-list "$(seq -s, 42 47)" --bind-to cpu-list:ordered $app Best Christoph - Original Message - From: "Open MPI Users" To: "Open MPI Users" Cc: "Carlo Nervi" Sent: Thursday, 20 A

[OMPI users] OMPI 4.0.4 how to use mpirun properly in numa architecture

2020-08-20 Thread Carlo Nervi via users
Dear OMPI community, I'm a simple end-user with no particular experience. I compile quantum chemical programs and use them in parallel. My system is a 4 socket, 12 core per socket Opteron 6168 system for a total of 48 cores and 64 Gb of RAM. It has 8 NUMA nodes: openmpi $ hwloc-info depth 0: