[mvapich-discuss] Re: Algorithms for MPI_Gather, MPI_Scatter and MPI_Bcast in MVAPICH2

[vaibhav dutt]

Hi,

I have seen leader based implementation of MPI_Bcast in a paper
http://sensorweb.cs.gsu.edu/~dej/docs/ccgrid08.pdf  in which all the data
is sent to a leader in a node and then the leader undergoes intra node
communication
to provide data to all processes within a node. Is this leader based
implementation also used
for MPI_Scatter and MPI_Gather in anyway?

Thank You,
Vaibhav

On Thu, Jul 26, 2012 at 11:31 AM, vaibhav dutt <
vaibhavsupersaiyan9 at gmail.com> wrote:

> Hi,
>
> I have been doing some experiments with OSU Microbenchmarks, determining
> the time spent in the collective operations like MPI_Gather, MPI_Scatter
> and MPI_Bcast.
> I was wondering which algorithms are used in MVAPICH2 to implement these.
> Can anyone help me regarding this?
>
> Thank You,
> Vaibhav
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.cse.ohio-state.edu/pipermail/mvapich-discuss/attachments/20120726/a55eef72/attachment-0001.html