July 2010 - mpi-forum - mailman3.cels.anl.gov

2011 meeting schedule - proposed
by Graham, Richard L. 12 Jul '10

12 Jul '10

Thanks for all the feedback on the proposed meeting schedule for 2011. Based on all the input, I put up a firmer schedule for 2011 at the meetings web site. I am trying to get more details on the Sept Euro MPI meeting, so we can schedule the Forum meeting and will get these up ASAP. Feedback is welcome. Thanks, Rich

1 0

Sept meeting agenda
by Graham, Richard L. 12 Jul '10

12 Jul '10

The first cut at a proposed Sept meeting has been posted on the web site. Please take a look the agenda and let me know of additions or subtractions you would like to see. Thanks, Rich

1 0

Re: [Mpi-forum] 2011 meetings
by Fab Tillier 09 Jul '10

09 Jul '10

If people really want an east-coast venue, I might be able to set things up in our Cambridge, MA office. I'm also hoping to be able to host people in Portland (direct flights from both Germany and Japan) if people would be interested in that. Is this something people would be interested in? Personally, California is a much easier trip for me (no red eye to Chicago). Boston requires just as long a trip as Chicago, but I have friends and family there, so the extra travel time can be amortized. Portland would be ideal of course... -Fab -----Original Message----- From: Bronis R. de Supinski Sent: Thursday, July 08, 2010 3:34 PM Rolf: Re: > 5 meetings in San Francisco and Chicago means > 3+2 or 2+3. The problem with this math is that it treats each year as a unit when the actual unit is individual meetings or the overall process. In 2010, we have: ATL, SFO, ORD, SFO, EURO, ORD, SFO. You propose to follow this in 2011 with: SFO, ORD, SFO, ORD, EURO, SFO My claim is we then in 2012 would be led to have: SFO, ORD, SFO, ORD, EURO, SFO Dropping the EuroMPI meetings we then have 9 out of 15 meetings in California. That is hardly equal. More importantly, we have had several people express that costs make it so they can attend Chicago meetings more easily than California (yes, many are fine with California meetings but not all). The arguments against Chicago are all weather and none about participation (the only thing close is the "significantly better for Japanese attendees" claim, which I think is a dubious: with United, I find 3 nonstops a day between SFO and Narita each way a day and those flights take about ten hours while I find 2 non-stops a day each way between ORD and Narita that take about 12 hours, so we are talking about a two flight time difference; yes, I know they are long flights but either way the cost is really a day on each end so the time difference is negligible). > With Chicago in winter time, the question is not the temperature, > it is the snow which sometimes continuously falls 3 days > (as I read). I do not recal EVER seeing a three day snow storm in the eight years I lived in Chicago. Yes, the weather is better in California, I cannot argue that but I think participation is more important. I think the more important question to raise here is the one Torsten was initiating: can we get the NSF funding renewed? $750 is tight for most of these trips but it much better than nothing (I don't use it although we do have Martin use so I am somewhat neutral here; I think it is essential for many of the US academics). Bronis > It is bad to make a 48h meeting slots > and the airplanes have problems to arrive. > Okay, some may have problems at their starting point > but not the whole forum. > Therefore, the sequence SFO - ORD - SFO - ORD - (SFO in Dec?) > seems quite natural. > And the Dec.-meeting should be not needed. > We wanted to finish before SC, i.e., at EuroMPI meeting. > > And SFO is a bit worse for Europe, but significantly better > for Japanese attendees. > > Rich, what is your opinion now? > > Best regards > Rolf > > > ----- Original Message ----- >> From: "Bronis R. de Supinski" <bronis(a)llnl.gov> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> Sent: Thursday, July 8, 2010 5:18:35 PM >> Subject: Re: [Mpi-forum] 2011 meetings >> Rolf: >> >> Re: >> > For me, Atlanta is also a bad option. >> > >> > Is there any reason not to make the schedule >> > starting with CA? >> >> I believe the reason was that we are in CA in December. The >> problem with making the swap is that we then end up likely >> to do it again since we are in CA in December by the schedule >> below. We then end up in California 50% more often... >> >> > Feb 07-09 California >> > Mar 28-30 Chicago >> > May 09-10 California >> > (not May 16-17 due to IPDPS) >> > (not May 23-25 due to CUG) >> > Jul 18-20 Chicago >> > Sep xx-yy together with EuroMPI >> > (not Nov 13-18 ??? due to SC - who knows exact date?) >> >> The exact dates are Saturday November 12 - Friday November 18. >> Since the first day is just the education program, your dates >> are probably close to accurate for most. >> >> Bronis >> >> >> >> >> > Dec 05-07 California >> > >> > I hope, I have not moved some dates without need, >> > i.e., correctly representing the state of current >> > discussion. >> > >> > Best regards >> > Rolf >> > >> > ----- Original Message ----- >> >> From: "Jeff Hammond" <jeff.science(a)gmail.com> >> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> >> Sent: Thursday, July 8, 2010 4:24:36 PM >> >> Subject: Re: [Mpi-forum] 2011 meetings >> >> I find Chicago quite pleasant except for December 15 to January 5, >> >> when I know the Forum will not meet. While it may be cold, there is >> >> frequently plenty of sunshine. If you're inside a building, it >> >> might >> >> as well be Hawaii outside. >> >> >> >> I plan to never attend a meeting in Atlanta because my past >> >> experiences with Delta and/or that airport have all been terrible. >> >> >> >> Jeff >> >> >> >> On Thu, Jul 8, 2010 at 9:11 AM, Jeff Squyres <jsquyres(a)cisco.com> >> >> wrote: >> >> > On Jul 8, 2010, at 10:01 AM, Torsten Hoefler wrote: >> >> > >> >> >> > At this stage I will keep that meeting in Atlanta, unless more >> >> >> > people >> >> >> > speak up. >> >> >> I think it is unreasonable to expect that a majority of the >> >> >> Forum >> >> >> will >> >> >> speak up via email (or even read this thread :-). It would be >> >> >> reasonable >> >> >> to do statistical sampling. I heard 4 people against Atlanta and >> >> >> 0 >> >> >> in >> >> >> favor so far. Speak up! >> >> > >> >> > I prefer California and Chicago to Atlanta. Atlanta is certainly >> >> > not a deal-breaker for me, but I do prefer the other two. >> >> > >> >> > -- >> >> > Jeff Squyres >> >> > jsquyres(a)cisco.com >> >> > For corporate legal information go to: >> >> > http://**www.**cisco.com/web/about/doing_business/legal/cri/ >> >> > >> >> > >> >> > _______________________________________________ >> >> > mpi-forum mailing list >> >> > mpi-forum(a)lists.mpi-forum.org >> >> > http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> >> > >> >> >> >> >> >> >> >> -- >> >> Jeff Hammond >> >> Argonne Leadership Computing Facility >> >> jhammond(a)mcs.anl.gov / (630) 252-5381 >> >> http://**www.**linkedin.com/in/jeffhammond >> >> >> >> _______________________________________________ >> >> mpi-forum mailing list >> >> mpi-forum(a)lists.mpi-forum.org >> >> http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> > >> > -- >> > Dr. Rolf Rabenseifner . . . . . . . . . .. email >> > rabenseifner(a)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) >> > >> > -- >> > Dr. Rolf Rabenseifner . . . . . . . . . .. email >> > rabenseifner(a)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(a)lists.mpi-forum.org >> > http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > -- > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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(a)lists.mpi-forum.org > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] 2011 meetings
by Bronis R. de Supinski 08 Jul '10

08 Jul '10

Rolf: Re: > 5 meetings in San Francisco and Chicago means > 3+2 or 2+3. The problem with this math is that it treats each year as a unit when the actual unit is individual meetings or the overall process. In 2010, we have: ATL, SFO, ORD, SFO, EURO, ORD, SFO. You propose to follow this in 2011 with: SFO, ORD, SFO, ORD, EURO, SFO My claim is we then in 2012 would be led to have: SFO, ORD, SFO, ORD, EURO, SFO Dropping the EuroMPI meetings we then have 9 out of 15 meetings in California. That is hardly equal. More importantly, we have had several people express that costs make it so they can attend Chicago meetings more easily than California (yes, many are fine with California meetings but not all). The arguments against Chicago are all weather and none about participation (the only thing close is the "significantly better for Japanese attendees" claim, which I think is a dubious: with United, I find 3 nonstops a day between SFO and Narita each way a day and those flights take about ten hours while I find 2 non-stops a day each way between ORD and Narita that take about 12 hours, so we are talking about a two flight time difference; yes, I know they are long flights but either way the cost is really a day on each end so the time difference is negligible). > With Chicago in winter time, the question is not the temperature, > it is the snow which sometimes continuously falls 3 days > (as I read). I do not recal EVER seeing a three day snow storm in the eight years I lived in Chicago. Yes, the weather is better in California, I cannot argue that but I think participation is more important. I think the more important question to raise here is the one Torsten was initiating: can we get the NSF funding renewed? $750 is tight for most of these trips but it much better than nothing (I don't use it although we do have Martin use so I am somewhat neutral here; I think it is essential for many of the US academics). Bronis > It is bad to make a 48h meeting slots > and the airplanes have problems to arrive. > Okay, some may have problems at their starting point > but not the whole forum. > Therefore, the sequence SFO - ORD - SFO - ORD - (SFO in Dec?) > seems quite natural. > And the Dec.-meeting should be not needed. > We wanted to finish before SC, i.e., at EuroMPI meeting. > > And SFO is a bit worse for Europe, but significantly better > for Japanese attendees. > > Rich, what is your opinion now? > > Best regards > Rolf > > > ----- Original Message ----- >> From: "Bronis R. de Supinski" <bronis(a)llnl.gov> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> Sent: Thursday, July 8, 2010 5:18:35 PM >> Subject: Re: [Mpi-forum] 2011 meetings >> Rolf: >> >> Re: >> > For me, Atlanta is also a bad option. >> > >> > Is there any reason not to make the schedule >> > starting with CA? >> >> I believe the reason was that we are in CA in December. The >> problem with making the swap is that we then end up likely >> to do it again since we are in CA in December by the schedule >> below. We then end up in California 50% more often... >> >> > Feb 07-09 California >> > Mar 28-30 Chicago >> > May 09-10 California >> > (not May 16-17 due to IPDPS) >> > (not May 23-25 due to CUG) >> > Jul 18-20 Chicago >> > Sep xx-yy together with EuroMPI >> > (not Nov 13-18 ??? due to SC - who knows exact date?) >> >> The exact dates are Saturday November 12 - Friday November 18. >> Since the first day is just the education program, your dates >> are probably close to accurate for most. >> >> Bronis >> >> >> >> >> > Dec 05-07 California >> > >> > I hope, I have not moved some dates without need, >> > i.e., correctly representing the state of current >> > discussion. >> > >> > Best regards >> > Rolf >> > >> > ----- Original Message ----- >> >> From: "Jeff Hammond" <jeff.science(a)gmail.com> >> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> >> Sent: Thursday, July 8, 2010 4:24:36 PM >> >> Subject: Re: [Mpi-forum] 2011 meetings >> >> I find Chicago quite pleasant except for December 15 to January 5, >> >> when I know the Forum will not meet. While it may be cold, there is >> >> frequently plenty of sunshine. If you're inside a building, it >> >> might >> >> as well be Hawaii outside. >> >> >> >> I plan to never attend a meeting in Atlanta because my past >> >> experiences with Delta and/or that airport have all been terrible. >> >> >> >> Jeff >> >> >> >> On Thu, Jul 8, 2010 at 9:11 AM, Jeff Squyres <jsquyres(a)cisco.com> >> >> wrote: >> >> > On Jul 8, 2010, at 10:01 AM, Torsten Hoefler wrote: >> >> > >> >> >> > At this stage I will keep that meeting in Atlanta, unless more >> >> >> > people >> >> >> > speak up. >> >> >> I think it is unreasonable to expect that a majority of the >> >> >> Forum >> >> >> will >> >> >> speak up via email (or even read this thread :-). It would be >> >> >> reasonable >> >> >> to do statistical sampling. I heard 4 people against Atlanta and >> >> >> 0 >> >> >> in >> >> >> favor so far. Speak up! >> >> > >> >> > I prefer California and Chicago to Atlanta. Atlanta is certainly >> >> > not a deal-breaker for me, but I do prefer the other two. >> >> > >> >> > -- >> >> > Jeff Squyres >> >> > jsquyres(a)cisco.com >> >> > For corporate legal information go to: >> >> > http://**www.**cisco.com/web/about/doing_business/legal/cri/ >> >> > >> >> > >> >> > _______________________________________________ >> >> > mpi-forum mailing list >> >> > mpi-forum(a)lists.mpi-forum.org >> >> > http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> >> > >> >> >> >> >> >> >> >> -- >> >> Jeff Hammond >> >> Argonne Leadership Computing Facility >> >> jhammond(a)mcs.anl.gov / (630) 252-5381 >> >> http://**www.**linkedin.com/in/jeffhammond >> >> >> >> _______________________________________________ >> >> mpi-forum mailing list >> >> mpi-forum(a)lists.mpi-forum.org >> >> http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> > >> > -- >> > Dr. Rolf Rabenseifner . . . . . . . . . .. email >> > rabenseifner(a)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) >> > >> > -- >> > Dr. Rolf Rabenseifner . . . . . . . . . .. email >> > rabenseifner(a)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(a)lists.mpi-forum.org >> > http://**lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > -- > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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(a)lists.mpi-forum.org > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] 2011 meetings
by Snir, Marc 08 Jul '10

08 Jul '10

If we care about airport performance then we should meet in Salt Lake City or Phoenix... http://www.bts.gov/programs/airline_information/airline_ontime_tables/2010_… http://www.bts.gov/programs/airline_information/airline_ontime_tables/2010_…

1 0

Re: [Mpi-forum] 2011 meetings
by Jeff Hammond 08 Jul '10

08 Jul '10

It would be good to look at actual data on airport closures if that's how you want to decide. December weather at SFO may not include snow but can still delay many flights. I've been victim to plenty of winter delays in Chicago due to problems at LAX, SFO, SeaTac, etc. not ORD/ MDW. I've heard ORD has more snow plows than any airport on earth. Jeff Sent from my iPhone On Jul 8, 2010, at 1:52 PM, Rolf Rabenseifner <rabenseifner(a)hlrs.de> wrote: > 5 meetings in San Francisco and Chicago means > 3+2 or 2+3. > With Chicago in winter time, the question is not the temperature, > it is the snow which sometimes continuously falls 3 days > (as I read). It is bad to make a 48h meeting slots > and the airplanes have problems to arrive. > Okay, some may have problems at their starting point > but not the whole forum. > Therefore, the sequence SFO - ORD - SFO - ORD - (SFO in Dec?) > seems quite natural. > And the Dec.-meeting should be not needed. > We wanted to finish before SC, i.e., at EuroMPI meeting. > > And SFO is a bit worse for Europe, but significantly better > for Japanese attendees. > > Rich, what is your opinion now? > > Best regards > Rolf > > > ----- Original Message ----- >> From: "Bronis R. de Supinski" <bronis(a)llnl.gov> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> Sent: Thursday, July 8, 2010 5:18:35 PM >> Subject: Re: [Mpi-forum] 2011 meetings >> Rolf: >> >> Re: >>> For me, Atlanta is also a bad option. >>> >>> Is there any reason not to make the schedule >>> starting with CA? >> >> I believe the reason was that we are in CA in December. The >> problem with making the swap is that we then end up likely >> to do it again since we are in CA in December by the schedule >> below. We then end up in California 50% more often... >> >>> Feb 07-09 California >>> Mar 28-30 Chicago >>> May 09-10 California >>> (not May 16-17 due to IPDPS) >>> (not May 23-25 due to CUG) >>> Jul 18-20 Chicago >>> Sep xx-yy together with EuroMPI >>> (not Nov 13-18 ??? due to SC - who knows exact date?) >> >> The exact dates are Saturday November 12 - Friday November 18. >> Since the first day is just the education program, your dates >> are probably close to accurate for most. >> >> Bronis >> >> >> >> >>> Dec 05-07 California >>> >>> I hope, I have not moved some dates without need, >>> i.e., correctly representing the state of current >>> discussion. >>> >>> Best regards >>> Rolf >>> >>> ----- Original Message ----- >>>> From: "Jeff Hammond" <jeff.science(a)gmail.com> >>>> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >>>> Sent: Thursday, July 8, 2010 4:24:36 PM >>>> Subject: Re: [Mpi-forum] 2011 meetings >>>> I find Chicago quite pleasant except for December 15 to January 5, >>>> when I know the Forum will not meet. While it may be cold, there is >>>> frequently plenty of sunshine. If you're inside a building, it >>>> might >>>> as well be Hawaii outside. >>>> >>>> I plan to never attend a meeting in Atlanta because my past >>>> experiences with Delta and/or that airport have all been terrible. >>>> >>>> Jeff >>>> >>>> On Thu, Jul 8, 2010 at 9:11 AM, Jeff Squyres <jsquyres(a)cisco.com> >>>> wrote: >>>>> On Jul 8, 2010, at 10:01 AM, Torsten Hoefler wrote: >>>>> >>>>>>> At this stage I will keep that meeting in Atlanta, unless more >>>>>>> people >>>>>>> speak up. >>>>>> I think it is unreasonable to expect that a majority of the >>>>>> Forum >>>>>> will >>>>>> speak up via email (or even read this thread :-). It would be >>>>>> reasonable >>>>>> to do statistical sampling. I heard 4 people against Atlanta and >>>>>> 0 >>>>>> in >>>>>> favor so far. Speak up! >>>>> >>>>> I prefer California and Chicago to Atlanta. Atlanta is certainly >>>>> not a deal-breaker for me, but I do prefer the other two. >>>>> >>>>> -- >>>>> Jeff Squyres >>>>> jsquyres(a)cisco.com >>>>> For corporate legal information go to: >>>>> http://*www.*cisco.com/web/about/doing_business/legal/cri/ >>>>> >>>>> >>>>> _______________________________________________ >>>>> mpi-forum mailing list >>>>> mpi-forum(a)lists.mpi-forum.org >>>>> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >>>>> >>>> >>>> >>>> >>>> -- >>>> Jeff Hammond >>>> Argonne Leadership Computing Facility >>>> jhammond(a)mcs.anl.gov / (630) 252-5381 >>>> http://*www.*linkedin.com/in/jeffhammond >>>> >>>> _______________________________________________ >>>> mpi-forum mailing list >>>> mpi-forum(a)lists.mpi-forum.org >>>> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >>> >>> -- >>> Dr. Rolf Rabenseifner . . . . . . . . . .. email >>> rabenseifner(a)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) >>> >>> -- >>> Dr. Rolf Rabenseifner . . . . . . . . . .. email >>> rabenseifner(a)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(a)lists.mpi-forum.org >>> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > -- > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] 2011 meetings
by Rolf Rabenseifner 08 Jul '10

08 Jul '10

5 meetings in San Francisco and Chicago means 3+2 or 2+3. With Chicago in winter time, the question is not the temperature, it is the snow which sometimes continuously falls 3 days (as I read). It is bad to make a 48h meeting slots and the airplanes have problems to arrive. Okay, some may have problems at their starting point but not the whole forum. Therefore, the sequence SFO - ORD - SFO - ORD - (SFO in Dec?) seems quite natural. And the Dec.-meeting should be not needed. We wanted to finish before SC, i.e., at EuroMPI meeting. And SFO is a bit worse for Europe, but significantly better for Japanese attendees. Rich, what is your opinion now? Best regards Rolf ----- Original Message ----- > From: "Bronis R. de Supinski" <bronis(a)llnl.gov> > To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> > Sent: Thursday, July 8, 2010 5:18:35 PM > Subject: Re: [Mpi-forum] 2011 meetings > Rolf: > > Re: > > For me, Atlanta is also a bad option. > > > > Is there any reason not to make the schedule > > starting with CA? > > I believe the reason was that we are in CA in December. The > problem with making the swap is that we then end up likely > to do it again since we are in CA in December by the schedule > below. We then end up in California 50% more often... > > > Feb 07-09 California > > Mar 28-30 Chicago > > May 09-10 California > > (not May 16-17 due to IPDPS) > > (not May 23-25 due to CUG) > > Jul 18-20 Chicago > > Sep xx-yy together with EuroMPI > > (not Nov 13-18 ??? due to SC - who knows exact date?) > > The exact dates are Saturday November 12 - Friday November 18. > Since the first day is just the education program, your dates > are probably close to accurate for most. > > Bronis > > > > > > Dec 05-07 California > > > > I hope, I have not moved some dates without need, > > i.e., correctly representing the state of current > > discussion. > > > > Best regards > > Rolf > > > > ----- Original Message ----- > >> From: "Jeff Hammond" <jeff.science(a)gmail.com> > >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> > >> Sent: Thursday, July 8, 2010 4:24:36 PM > >> Subject: Re: [Mpi-forum] 2011 meetings > >> I find Chicago quite pleasant except for December 15 to January 5, > >> when I know the Forum will not meet. While it may be cold, there is > >> frequently plenty of sunshine. If you're inside a building, it > >> might > >> as well be Hawaii outside. > >> > >> I plan to never attend a meeting in Atlanta because my past > >> experiences with Delta and/or that airport have all been terrible. > >> > >> Jeff > >> > >> On Thu, Jul 8, 2010 at 9:11 AM, Jeff Squyres <jsquyres(a)cisco.com> > >> wrote: > >> > On Jul 8, 2010, at 10:01 AM, Torsten Hoefler wrote: > >> > > >> >> > At this stage I will keep that meeting in Atlanta, unless more > >> >> > people > >> >> > speak up. > >> >> I think it is unreasonable to expect that a majority of the > >> >> Forum > >> >> will > >> >> speak up via email (or even read this thread :-). It would be > >> >> reasonable > >> >> to do statistical sampling. I heard 4 people against Atlanta and > >> >> 0 > >> >> in > >> >> favor so far. Speak up! > >> > > >> > I prefer California and Chicago to Atlanta. Atlanta is certainly > >> > not a deal-breaker for me, but I do prefer the other two. > >> > > >> > -- > >> > Jeff Squyres > >> > jsquyres(a)cisco.com > >> > For corporate legal information go to: > >> > http://*www.*cisco.com/web/about/doing_business/legal/cri/ > >> > > >> > > >> > _______________________________________________ > >> > mpi-forum mailing list > >> > mpi-forum(a)lists.mpi-forum.org > >> > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > >> > > >> > >> > >> > >> -- > >> Jeff Hammond > >> Argonne Leadership Computing Facility > >> jhammond(a)mcs.anl.gov / (630) 252-5381 > >> http://*www.*linkedin.com/in/jeffhammond > >> > >> _______________________________________________ > >> mpi-forum mailing list > >> mpi-forum(a)lists.mpi-forum.org > >> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > > > -- > > Dr. Rolf Rabenseifner . . . . . . . . . .. email > > rabenseifner(a)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) > > > > -- > > Dr. Rolf Rabenseifner . . . . . . . . . .. email > > rabenseifner(a)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(a)lists.mpi-forum.org > > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum -- Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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)

1 0

Re: [Mpi-forum] 2011 meetings
by Pavan Balaji 08 Jul '10

08 Jul '10

I prefer Chicago and CA as well. I didn't go to the Atlanta meeting, but I see no reason to have a meeting there. -- Pavan On 07/08/2010 10:04 AM, Darius Buntinas wrote: > ...and the Chicago and CA meeting facilities were free IIRC. Also, while the wireless may have been flaky at times in Chicago and CA, but it's better than no wireless at all in ATL. > > I prefer Chicago or CA over ATL, but I'd still go if it were in ATL. > > -d > > On Jul 8, 2010, at 9:47 AM, Dave Goodell wrote: > >> On Jul 8, 2010, at 5:37 AM CDT, Graham, Richard L. wrote: >> >>> At this stage I will keep that meeting in Atlanta, unless more people speak up. >> The CA and Chicago venues that we normally use are better than the ATL hotel we had last time. I would prefer not meeting in Atlanta unless we can be sure it will work as well as the other two locations. >> >> -Dave >> >> >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [Mpi-forum] 2011 meetings
by Bronis R. de Supinski 08 Jul '10

08 Jul '10

Rolf: Re: > For me, Atlanta is also a bad option. > > Is there any reason not to make the schedule > starting with CA? I believe the reason was that we are in CA in December. The problem with making the swap is that we then end up likely to do it again since we are in CA in December by the schedule below. We then end up in California 50% more often... > Feb 07-09 California > Mar 28-30 Chicago > May 09-10 California > (not May 16-17 due to IPDPS) > (not May 23-25 due to CUG) > Jul 18-20 Chicago > Sep xx-yy together with EuroMPI > (not Nov 13-18 ??? due to SC - who knows exact date?) The exact dates are Saturday November 12 - Friday November 18. Since the first day is just the education program, your dates are probably close to accurate for most. Bronis > Dec 05-07 California > > I hope, I have not moved some dates without need, > i.e., correctly representing the state of current > discussion. > > Best regards > Rolf > > ----- Original Message ----- >> From: "Jeff Hammond" <jeff.science(a)gmail.com> >> To: "Main MPI Forum mailing list" <mpi-forum(a)lists.mpi-forum.org> >> Sent: Thursday, July 8, 2010 4:24:36 PM >> Subject: Re: [Mpi-forum] 2011 meetings >> I find Chicago quite pleasant except for December 15 to January 5, >> when I know the Forum will not meet. While it may be cold, there is >> frequently plenty of sunshine. If you're inside a building, it might >> as well be Hawaii outside. >> >> I plan to never attend a meeting in Atlanta because my past >> experiences with Delta and/or that airport have all been terrible. >> >> Jeff >> >> On Thu, Jul 8, 2010 at 9:11 AM, Jeff Squyres <jsquyres(a)cisco.com> >> wrote: >> > On Jul 8, 2010, at 10:01 AM, Torsten Hoefler wrote: >> > >> >> > At this stage I will keep that meeting in Atlanta, unless more >> >> > people >> >> > speak up. >> >> I think it is unreasonable to expect that a majority of the Forum >> >> will >> >> speak up via email (or even read this thread :-). It would be >> >> reasonable >> >> to do statistical sampling. I heard 4 people against Atlanta and 0 >> >> in >> >> favor so far. Speak up! >> > >> > I prefer California and Chicago to Atlanta. Atlanta is certainly >> > not a deal-breaker for me, but I do prefer the other two. >> > >> > -- >> > Jeff Squyres >> > jsquyres(a)cisco.com >> > For corporate legal information go to: >> > http://*www.*cisco.com/web/about/doing_business/legal/cri/ >> > >> > >> > _______________________________________________ >> > mpi-forum mailing list >> > mpi-forum(a)lists.mpi-forum.org >> > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum >> > >> >> >> >> -- >> Jeff Hammond >> Argonne Leadership Computing Facility >> jhammond(a)mcs.anl.gov / (630) 252-5381 >> http://*www.*linkedin.com/in/jeffhammond >> >> _______________________________________________ >> mpi-forum mailing list >> mpi-forum(a)lists.mpi-forum.org >> http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum > > -- > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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) > > -- > Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)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(a)lists.mpi-forum.org > http://*lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

Re: [Mpi-forum] 2011 meetings
by Darius Buntinas 08 Jul '10

08 Jul '10

...and the Chicago and CA meeting facilities were free IIRC. Also, while the wireless may have been flaky at times in Chicago and CA, but it's better than no wireless at all in ATL. I prefer Chicago or CA over ATL, but I'd still go if it were in ATL. -d On Jul 8, 2010, at 9:47 AM, Dave Goodell wrote: > On Jul 8, 2010, at 5:37 AM CDT, Graham, Richard L. wrote: > >> At this stage I will keep that meeting in Atlanta, unless more people speak up. > > The CA and Chicago venues that we normally use are better than the ATL hotel we had last time. I would prefer not meeting in Atlanta unless we can be sure it will work as well as the other two locations. > > -Dave > > > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0