[Mpi-forum] Meeting schedule for 2013

Fri Sep 28 09:11:03 CDT 2012

Le 27 sept. 2012 à 08:47, "Bronis R. de Supinski" <bronis at llnl.gov> a écrit :

> 
> Rolf:
> 
> Re:
>>> even years: 3US+euroMPI
>>> odd years: 2US,1Asia+EuroMPI.
>> 
>> Yes, this was the last position at the Forum and also discussed at the EuroMPI business meeting.
> 
> As I understand the situation, this option was not discussed
> at the Forum. As stated previously, the matter was decided in
> Chicago and reopening it without an announcement was inappropriate.
> 
> In addition, what is discussed at the EuroMPI business meeting
> is not relevant. It is not their decision.
> 
Bronis, the matter has not been decided, as we are obviously still discussing it. Nothing was decided either in Vienna as it felt like there was no consensus still (straw votes where very divided, and many options came close to one another, we should be driven by consensus right?). No need to feel so strong on "rules" again, we are just expressing wishes so far. 

It so happens that many people are going to attend EuroMPI anyway, so collocating is a clear way of cutting cost for many attendees; however, it could be interesting to make so that we don't have to spend a WE in a foreign country when attending both the conference and the Forum, as this is expensive and annoying. 

Also, it is to be noted that european institutions participation is declining recently. Locating meetings in Asia was taken as a step to address the same issue when Asian participation was low. We might consider this in the equation, or risk that the forum becomes an US only body, and not an international body anymore. 

So the question is not in terms of us against them (the conference), it is in terms of what is convenient for the Forum members (many of whom do attend the conference, many of whom are not US based institutions) and makes sense for an -international- standardization body. 

Regards, 
Aurelien

>> The MPI Forum was re-started by the activities at several EuroMPI conferences, the last one at Paris 2007.
>> 
>> Therefore, there is a strong relationship and resposibility
>> and to have the MPI Forum meeting in US and the EuroMPI in Europe
>> both in fall, would not reflect this responsibility.
>> 
>> As far as I see, the colleagues in Madrid would like to host the MPI Forum meeting together with the EuroMPI and therefore, I expect - based on the previous meeting rules - we will have
>> 
>> MPI-next Forum meeting Sep. 11-13, 2013,
>> (instead of Sep. 9-11 in San Jose), and
>> EuroMPI, Sep. 15-18, 2013, both in Madrid, Spain
>> 
>> This may change in later years if the EuroMPI may go to Asia,
>> but this is another discussion within the EuroMPI organizers.
> 
> What I hear is that holding the MPI Forum meeting
> separately from EuroMPI would damage EuroMPI. Well,
> forcing the Forum to co-locate is putting the cart
> before the horse. If EuroMPI cannot withstand a
> separate meeting, as decided by the Forum, then
> EuroMPI needs to adapt -- not the other way around.
> 
> As was previously suggested, EuroMPI is free to rebrand
> as International MPI and to co-locate with Forum meetings.
> If the conference demonstrated that flexibility, I would
> be more willing to consider their desires as an MPI
> Forum voting member.
> 
> Bronis
> 
> 
> 
> 
>> Best regards
>> Rolf 
>> 
>> ----- Original Message -----
>>> From: "Aurélien Bouteiller" <bouteill at eecs.utk.edu>
>>> To: "Main MPI Forum mailing list" <mpi-forum at lists.mpi-forum.org>
>>> Sent: Thursday, September 27, 2012 2:41:08 AM
>>> Subject: Re: [Mpi-forum] Meeting schedule for 2013
>>> The 3-1 with 1 alternate between asia and europe collected the most
>>> vote. However there was significant disappointment about not
>>> collocating a meeting with EuroMPI, and after the straw vote closed,
>>> another interesting possibility was envisioned:
>>> even years: 3US+euroMPI
>>> odd years: 2US,1Asia+EuroMPI.
>>> This last proposal has not been straw voted, so it is not possible to
>>> know if it has support, but it seemed from the reception it had that
>>> the idea is not unbearable for most US institutions.
>>> Aurelien
>>> Le 26 sept. 2012 à 15:17, Jeff Squyres <jsquyres at cisco.com> a écrit :
>>> > I don't remember there being a majority for the 2-2 option here in
>>> > Vienna -- am I remembering last week incorrectly?
>>> >
>>> > I definitely can't do 2 international meetings next week.
>>> >
>>> >
>>> >
>>> >
>>> > On Sep 26, 2012, at 8:02 PM, Richard Graham wrote:
>>> >
>>> >> This was discussed and there was another vote at the meeting in
>>> >> Vienna. This note is just a follow up to the discussions at that
>>> >> meeting.
>>> >>
>>> >> Rich
>>> >>
>>> >> -----Original Message-----
>>> >> From: mpi-forum-bounces at lists.mpi-forum.org
>>> >> [mailto:mpi-forum-bounces at lists.mpi-forum.org] On Behalf Of Schulz,
>>> >> Martin
>>> >> Sent: Wednesday, September 26, 2012 7:36 PM
>>> >> To: Bronis R. de Supinski; Main MPI Forum mailing list
>>> >> Subject: Re: [Mpi-forum] Meeting schedule for 2013
>>> >>
>>> >> Hi Rich,
>>> >>
>>> >> I have the same recollection as Bronis and Mohamad. I thought there
>>> >> was a strong majority (at least at the Chicago meeting) for a 3-1
>>> >> solution with alternating non-US locations (one year Asia, one year
>>> >> Europe, but not both in one year - as it also shows on the current
>>> >> tentative meeting calendar for next year). I know this is not fair
>>> >> to everyone, but the group (again, at least in Chicago, which was
>>> >> pretty complete, though, since we were wrapping up 3.0) felt that
>>> >> this would maximize active participation.
>>> >>
>>> >> Martin
>>> >>
>>> >>
>>> >> On Sep 26, 2012, at 9:39 AM, Bronis R. de Supinski wrote:
>>> >>
>>> >>>
>>> >>> Rich:
>>> >>>
>>> >>> I would have to disagree. At best, you are splitting hairs to say
>>> >>> this
>>> >>> option was not voted on. In the "alternate"
>>> >>> years, it is simply the 2-2 option that was voted on and did not
>>> >>> come
>>> >>> anywhere near the 3-1 option in terms of votes.
>>> >>>
>>> >>> The arguments against 2-2 (even in alternating years) still apply.
>>> >>> It
>>> >>> significantly lowers attendance by active participants.
>>> >>>
>>> >>> Bronis
>>> >>>
>>> >>>
>>> >>>
>>> >>>
>>> >>> On Wed, 26 Sep 2012, Richard Graham wrote:
>>> >>>
>>> >>>> Such decisions are not made lightly - I assure you this. This was
>>> >>>> discussed at the last meeting, and this was an option that was
>>> >>>> not voted on. However, this was discussed again at the end of the
>>> >>>> meeting, and some at EuroMPI.
>>> >>>>
>>> >>>> Rich
>>> >>>>
>>> >>>> -----Original Message-----
>>> >>>> From: mpi-forum-bounces at lists.mpi-forum.org
>>> >>>> [mailto:mpi-forum-bounces at lists.mpi-forum.org] On Behalf Of
>>> >>>> Mohamad
>>> >>>> Chaarawi
>>> >>>> Sent: Wednesday, September 26, 2012 6:30 PM
>>> >>>> To: mpi-forum at lists.mpi-forum.org
>>> >>>> Subject: Re: [Mpi-forum] Meeting schedule for 2013
>>> >>>>
>>> >>>> Hi Rich,
>>> >>>>
>>> >>>> On 9/26/2012 10:30 AM, Richard Graham wrote:
>>> >>>>> Therefore, I propose to keep the European meeting of the Forum
>>> >>>>> in
>>> >>>>> conjunction with EuroMPI, and every other year also have a
>>> >>>>> meeting
>>> >>>>> in Japan. With this in mind, here is the proposed schedule for
>>> >>>>> 2012
>>> >>>>
>>> >>>> The majority at the last meeting in Chicago were against having 2
>>> >>>> meetings internationally. We had several votes on different
>>> >>>> formations (4-0, 3-1, and 2-2), and 3-1 was the majority vote.
>>> >>>>
>>> >>>> It would have been fair to have a similar vote in Europe and see
>>> >>>> how the participants vote..
>>> >>>> Did the attendance at the Vienna meeting vote on a different
>>> >>>> schedule, or is your change coming from just some concerns (which
>>> >>>> are also valid) from some people?
>>> >>>>
>>> >>>> Can we have a formal organizational Vote by eligible
>>> >>>> organizations on this instead of just scheduling meetings based
>>> >>>> on suggestions that vary from meeting to meeting?
>>> >>>> This vote can be done over the phone.
>>> >>>>
>>> >>>> Thanks,
>>> >>>> Mohamad
>>> >>>> _______________________________________________
>>> >>>> mpi-forum mailing list
>>> >>>> mpi-forum at lists.mpi-forum.org
>>> >>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> >>>>
>>> >>>> _______________________________________________
>>> >>>> mpi-forum mailing list
>>> >>>> mpi-forum at lists.mpi-forum.org
>>> >>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> >>>>
>>> >>> _______________________________________________
>>> >>> mpi-forum mailing list
>>> >>> mpi-forum at lists.mpi-forum.org
>>> >>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> >>
>>> >> ________________________________________________________________________
>>> >> Martin Schulz, schulzm at llnl.gov, http://people.llnl.gov/schulzm
>>> >> CASC @ Lawrence Livermore National Laboratory, Livermore, USA
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> _______________________________________________
>>> >> mpi-forum mailing list
>>> >> mpi-forum at lists.mpi-forum.org
>>> >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> >>
>>> >> _______________________________________________
>>> >> mpi-forum mailing list
>>> >> mpi-forum at lists.mpi-forum.org
>>> >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> >
>>> >
>>> > --
>>> > Jeff Squyres
>>> > jsquyres at cisco.com
>>> > For corporate legal information go to:
>>> > http://www.cisco.com/web/about/doing_business/legal/cri/
>>> >
>>> >
>>> > _______________________________________________
>>> > mpi-forum mailing list
>>> > mpi-forum at lists.mpi-forum.org
>>> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>>> --
>>> * Dr. Aurélien Bouteiller
>>> * Researcher at Innovative Computing Laboratory
>>> * University of Tennessee
>>> * 1122 Volunteer Boulevard, suite 309b
>>> * Knoxville, TN 37996
>>> * 865 974 9375
>>> _______________________________________________
>>> mpi-forum mailing list
>>> mpi-forum at lists.mpi-forum.org
>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
>> 
>> -- 
>> Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner at hlrs.de
>> High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530
>> University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832
>> Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner
>> Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30)
>> 
>> _______________________________________________
>> mpi-forum mailing list
>> mpi-forum at lists.mpi-forum.org
>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum
> _______________________________________________
> mpi-forum mailing list
> mpi-forum at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

--
* Dr. Aurélien Bouteiller
* Researcher at Innovative Computing Laboratory
* University of Tennessee
* 1122 Volunteer Boulevard, suite 309b
* Knoxville, TN 37996
* 865 974 9375