[mpich-discuss] reading data into a parallel program

[Prashanth]

Jayesh,
  Thanks again for getting back to me. How do I broadcast "non-standard"
datasets using MPI? For example, how do i broadcast unstructured grids to
all the nodes? Also, should I use a barrier to hold all the processes till
i'm done broadcasting?
Prashanth

On Tue, Dec 7, 2010 at 1:27 PM, Jayesh Krishna <jayesh at mcs.anl.gov> wrote:

> Hi,
>  In your case it looks like you might be better off reading from one MPI
> process and broadcasting to others.
>
> Regards,
> Jayesh
> ----- Original Message -----
> From: Prashanth <prashanth.dumpuri at gmail.com>
> To: Jayesh Krishna <jayesh at mcs.anl.gov>, mpich-discuss at mcs.anl.gov
> Sent: Tue, 07 Dec 2010 11:10:07 -0600 (CST)
> Subject: Re: [mpich-discuss] reading data into a parallel program
>
> Jayesh,
>  Thanks for getting back to me. I'm not using MPI-IO functions. My datasets
> are unstructured grids and I have to use another toolkit (VTK) to read the
> files. I can store them in the binary format and they shouldn't be more
> than
> a few megabytes.
>  My MPI processes are launched after i read the data and every processor
> requires knowledge of the complete dataset (in other words i cannot
> partition the data and split it across the available processors). Hope I
> answered your question correctly.
>   So, given this should I let all the nodes read the input files?
> Thanks
> Prashanth
>
> On Tue, Dec 7, 2010 at 10:45 AM, Jayesh Krishna <jayesh at mcs.anl.gov>
> wrote:
>
> > Hi,
> >  This will depend on how much data is read and where your MPI processes
> are
> > launched. Are you using MPI-IO functions (like MPI_File_read_all()) for
> > reading the data ?
> >
> > Regards,
> > Jayesh
> > ----- Original Message -----
> > From: Prashanth <prashanth.dumpuri at gmail.com>
> > To: mpich-discuss at mcs.anl.gov
> > Sent: Mon, 06 Dec 2010 10:44:53 -0600 (CST)
> > Subject: [mpich-discuss] reading data into a parallel program
> >
> > All,
> >   I have a main program that reads input data from files and 3
> > filters/classes instantiated within the main program that process the
> > datasets. I have parallelized 2 of those 3 classes (using MPICH2). Is it
> > faster to read data on the root processor and then broadcast it to all
> the
> > child nodes or should I let all the nodes read the data?
> >    I'm running my main program on a Windows XP machine with an Intel Quad
> > core processor.
> > Thanks
> > Prashanth
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20101207/3ec1514b/attachment.htm>