I’ve clarified it in my local version, that’ll go out on the website. IBM BG/Q MPI had the MPI_T interface implemented, but it doesn’t expose any variables. So while it’s technically MPI-3 compliant, it’s not particularly useful to users. Hence they get an extra clarification on their “supported” check box. — Pavan On Jun 9, 2014, at 5:33 PM, Jeff Squyres (jsquyres) <[email protected]> wrote:

Just curious -- what does (check) mean?

On Jun 9, 2014, at 4:40 PM, "Balaji, Pavan" <[email protected]> wrote:

...

Folks,

Thanks for the corrections so far. Please see attached for the corrected version.

If there are no additional changes, we’ll upload this version to the Forum website in a couple of days.

Thanks,

— Pavan

On Jun 9, 2014, at 12:34 AM, Balaji, Pavan <[email protected]> wrote:

...

Folks,

With ISC coming up, I wanted to make sure the status of all MPI-3 implementations is accurately reflected on the attached implementations slide (which will be put up on the Forum website and will be used at ISC talks and BoFs).

If there is any inaccuracy in the status, please let us know. I know that some implementations are planning for a release at ISC which will change this status. In such cases, please send me a note with the change and the date when the change needs to be made public and I’ll take care of it.

Thanks,

— Pavan

<mpi-implementations.pptx>_______________________________________________ mpi-forum mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

<mpi-implementations.pptx>_______________________________________________ mpi-forum mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

-- Jeff Squyres [email protected] For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

_______________________________________________ mpi-forum mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum