[mpiwg-tools] Call cancelled today
hermanns at jara.rwth-aachen.de
Thu Mar 8 07:04:56 CST 2018
I won't make the call today and will have to cancel.
Nevertheless, I want to give a short summary of the WG topics at the
- Ticket #6: MPIR line numbers has passed its the vote, thus we will
soon have to prepare and release MPIR 1.1.
- Ticket #77: Integer-based access to MPI_Wtime
This sparked an interesting discussion that we should pick up in the
next call. While a straw vote showed that the majority likes the fact
that the new calls use MPI_Wtime, looking at Jim's measurements on the
ticket discussion, I wondered whether an API that enabled the use of
the low-overhead counter for our event timestamps would be more
beneficial for the interface.
- Ticket #67: Consistent reference of MPI process
We found a couple more places where we should adapt the text. I will
fix this, and this will go up for (a) a no-no vote and (b) potentially
a first vote (or another reading) at the next meeting.
- Ticket #79: MPI_T events
I presented the current proposal and there were no big objections, yet
the role of sources was discussed briefly, where I will update the
text a bit to enable an easier and more consistent interface to
sources in general.
I couldn't attend the QMPI discussion, but in the wrap-up, I think
Martin concluded that we are also at a point, where we should produce
a prototype to see if the API is what we need and want.
Jülich Aachen Research Alliance,
High Performance Computing (JARA-HPC)
Jülich Supercomputing Centre (JSC)
Phone: +49 2461 61 2509 | +49 241 80 24381
Fax: +49 2461 80 6 99753
email: hermanns at jara.rwth-aachen.de
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 4899 bytes
Desc: S/MIME Cryptographic Signature
More information about the mpiwg-tools