[petsc-dev] testexamples using too many processes

Satish Balay balay at mcs.anl.gov
Mon Mar 23 10:27:14 CDT 2015


nemesis has this performance degradation issue [when oversubscribing
cpu-cores] - so --download-mpich defaults using
-download-mpich-device=ch3:sock [this mode doesn't not have
performance degradation]

Satish

On Mon, 23 Mar 2015, Matthew Knepley wrote:

> On Mon, Mar 23, 2015 at 8:21 AM, Lisandro Dalcin <dalcinl at gmail.com> wrote:
> 
> > At least one test for DMDA executes in 32 processes. My desktop
> > machine has 8 cores, my laptop just 4. Running testexamples make them
> > totally unresponsive for a while. Any ideas about how to improve this?
> >
> 
> It should not do that. What MPICH device do you use? Maybe try fork?
> 
>   Thanks,
> 
>      Matt
> 
> 
> > mpiexec -n 32 ./ex36 -mx 5 -my 5 -mz 5 -nl 3 -dim 3 -cmap 1
> > -da_refine_x 1 -da_refine_y 3 -da_refine_z 4
> >
> > --
> > Lisandro Dalcin
> > ============
> > Research Scientist
> > Computer, Electrical and Mathematical Sciences & Engineering (CEMSE)
> > Numerical Porous Media Center (NumPor)
> > King Abdullah University of Science and Technology (KAUST)
> > http://numpor.kaust.edu.sa/
> >
> > 4700 King Abdullah University of Science and Technology
> > al-Khawarizmi Bldg (Bldg 1), Office # 4332
> > Thuwal 23955-6900, Kingdom of Saudi Arabia
> > http://www.kaust.edu.sa
> >
> > Office Phone: +966 12 808-0459
> >
> 
> 
> 
> 




More information about the petsc-dev mailing list