Hi Hui,

Here is the updated list with your request:
  1. One mpi job with mpiexec.hydra -n 64 myprog => 57.32s
  2. One mpi job with mpiexec.hydra -bind-to numa -n 64 myprog => 50.52s
  3. Two mpi jobs with mpiexec.hydra -n 64 myprog => 99.77s
  4. Two mpi jobs with mpiexec.hydra -bind-to numa -n 64 myprog => 72.23s
  5. Four mpi jobs with mpiexec.hydra -bind-to numa -n 64 myprog => 159.2s
  6. Four mpi jobs with mpiexec.hydra -n 64 myprog => 159.9s
Case 6 is the same speed as Case 5.  The code is a CFD code based on a subdomain approach.   There is communication between neighboring subdomains.  The machine is a supermicro server with 512GB of memory.  These small test cases use about 3% of the available memory.   I want to run 4 larger MPI jobs concurrently.  


From: Zhou, Hui <zhouh@anl.gov>
Sent: Wednesday, November 13, 2019 8:55 AM
To: discuss@mpich.org
Cc: Douglas Dommermuth
Subject: Re: [mpich-discuss] Optimizing runtime for 4 mpiexec.hydra jobs
 
Hi Doug, 

What is your number running 4 mpi jobs without -bind-to numa?

— 
Hui Zhou









On Nov 13, 2019, at 10:39 AM, Douglas Dommermuth via discuss <discuss@mpich.org> wrote:

I am running Ubuntu 18.04.3 with MPICH 3.3~a2-4 and GFortran 4:7.4.0-1ubuntu2.3 and GCC 4:7.4.0-1ubuntu2.3CC on dual AMD EPYC 7742 processors with hyper threading enabled.  My codes are written in MPI and Fortran.  The dual AMD processors have 128 cores and 256 threads.   I want to optimize the runtime for 4 mpi jobs running concurrently with 64 threads each.  Some timings are provided here:
  1. One mpi job with mpiexec.hydra -n 64 myprog => 57.32s 
  2. One mpi job with mpiexec.hydra -bind-to numa -n 64 => 50.52s
  3. Two mpi jobs with mpiexec.hydra -n 64 myprog => 99.77s
  4. Two mpi jobs with mpiexec.hydra -bind-to numa -n 64 => 72.23s
  5. Four mpi jobs with mpiexec.hydra -bind-to numa -n 64 => 159.2s
The option "-bind-to numa" helps, but even so,  running four mpi jobs concurrently with 64 threads each is considerably slower than running one mpi job with 64 threads.  I can almost run four mpi jobs sequentially and match the time for running four mpi jobs concurrently.   How can I improve on the result for running 4 mpi jobs concurrently?   Thanks, Doug.
_______________________________________________
discuss mailing list     discuss@mpich.org
To manage subscription options or unsubscribe:
https://lists.mpich.org/mailman/listinfo/discuss