March 2012 - mpi-forum - mailman3.cels.anl.gov

Re: [Mpi-forum] overloading MPI_Init etc.
by Jeffrey Squyres 05 Mar '12

05 Mar '12

Bronis -- I was thinking the same thing, but then I noticed that the user changed the signature of MPI_Init. I think that that use over C overloading (changing the signature) is "dodgy" at best, because they're technically infringing on the MPI_* namespace (i.e., they're changing it). Overloading to do profiling is really just that: overloading (or intercepting, if you will). It's not changing or adding to the MPI_* namespace. So yes, I think Fab's comment is that this guy is going against the MPI spec is correct here. And therefore the user's argument is moot. (plus, the argument is kinda silly in that any C-based API is going to run into the same problems in C++ -- singling out MPI is kinda meaningless here, IMHO) Patient: "Doc, it hurts when I move my arm like this." Doc: "Don't do that. $20 co-pay, please." On Mar 5, 2012, at 5:27 AM, Bronis R. de Supinski wrote: > > Um, it is standard practice when using the profiling > interface to redefine MPI_* functions. I think this > question is one for the C and C++ standards, not MPI... > > On Sun, 4 Mar 2012, Jeff Hammond wrote: > >> What was ambiguous to me - looking at 2.6.3-4 - was if using MPI_ was >> still a protected namespace in C++. The standard really only says >> explicitly that it is so in C. It is only with an indirect argument >> that mpi.h can define both C and C++ bindings that I can see C++ code >> not being able to define MPI_ functions. >> >> Note that MyMPI_ was ~exactly~ how I fixed the user's code :-) >> >> Jeff >> >> On Sun, Mar 4, 2012 at 9:39 PM, Fab Tillier <ftillier(a)microsoft.com> wrote: >>> I don't think defining one's own mpi.h is an issue from a compliance perspective. However, overloading MPI_Init definitely is - the standard reserves the MPI_ namespace, so if it isn't in the standard, it's not standard compliant. Now, if they named it MyMPI_Init, it'd be fine. >>> >>> -Fab >>> >>> Jeff Hammond wrote on Sun, 4 Mar 2012 at 18:22:40 >>> >>>> I have a user who insists it is MPI-compliant to do the following: >>>> >>>> user file mpi.h, which is included in the application with #include >>>> "mpi.h": ================================== #include <mpi.h> >>>> >>>> <snip> >>>> >>>> void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& >>>> Master) { >>>> MPI::Init(argc, argv); >>>> my_rank = MPI::COMM_WORLD.Get_rank(); >>>> mpi_size = MPI::COMM_WORLD.Get_size(); >>>> Master = 0; >>>> >>>> std::cout << "Hello World! I am " << my_rank << " of " << mpi_size >>>> << std::endl; >>>> } >>>> >>>> void MPI_finalize() >>>> {MPI::Finalize();} >>>> >>>> ================================== >>>> >>>> I believe that it is a terrible practice, if not a violation of the >>>> MPI standard to: >>>> (1) define one's own mpi.h, since this may confuse the preprocessor >>>> (or are there precedence rules in the ANSI C standard?) as to which to >>>> include first (especially if CC=mpicc or CXX=mpicxx), and >>>> (2) overload MPI_Init, even if one is using C++ and relying upon C++ >>>> name-mangling to avoid a symbol conflict. >>>> >>>> The user believes that the inability to overload MPI_Init in C++ means >>>> that MPI is not compliant with the C++ standard. >>>> >>>> Thanks, >>>> >>>> Jeff >>>> >>> >>> >>> _______________________________________________ >>> 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 >> University of Chicago Computation Institute >> jhammond(a)alcf.anl.gov / (630) 252-5381 >> http://www.linkedin.com/in/jeffhammond >> https://wiki.alcf.anl.gov/old/index.php/User:Jhammond >> https://wiki-old.alcf.anl.gov/index.php/User:Jhammond >> >> _______________________________________________ >> 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 -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [Mpi-forum] overloading MPI_Init etc.
by William Gropp 05 Mar '12

05 Mar '12

The issue is really one of whether the C++ (or Fortran) interfaces call the C MPI interface or not. There are both reasons to do so and reasons not. The reasons for is that a developer can write a profiling library that intercepts the C calls; on an implementation that routes everything through C, this simplifies their effort (at some cost in precision of call). In this case, the user should have make use of the PMPI::Init interface. Bill On Mar 5, 2012, at 4:27 AM, Bronis R. de Supinski wrote: > > Um, it is standard practice when using the profiling > interface to redefine MPI_* functions. I think this > question is one for the C and C++ standards, not MPI... > > On Sun, 4 Mar 2012, Jeff Hammond wrote: > >> What was ambiguous to me - looking at 2.6.3-4 - was if using MPI_ was >> still a protected namespace in C++. The standard really only says >> explicitly that it is so in C. It is only with an indirect argument >> that mpi.h can define both C and C++ bindings that I can see C++ code >> not being able to define MPI_ functions. >> >> Note that MyMPI_ was ~exactly~ how I fixed the user's code :-) >> >> Jeff >> >> On Sun, Mar 4, 2012 at 9:39 PM, Fab Tillier <ftillier(a)microsoft.com> wrote: >>> I don't think defining one's own mpi.h is an issue from a compliance perspective. However, overloading MPI_Init definitely is - the standard reserves the MPI_ namespace, so if it isn't in the standard, it's not standard compliant. Now, if they named it MyMPI_Init, it'd be fine. >>> >>> -Fab >>> >>> Jeff Hammond wrote on Sun, 4 Mar 2012 at 18:22:40 >>> >>>> I have a user who insists it is MPI-compliant to do the following: >>>> >>>> user file mpi.h, which is included in the application with #include >>>> "mpi.h": ================================== #include <mpi.h> >>>> >>>> <snip> >>>> >>>> void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& >>>> Master) { >>>> MPI::Init(argc, argv); >>>> my_rank = MPI::COMM_WORLD.Get_rank(); >>>> mpi_size = MPI::COMM_WORLD.Get_size(); >>>> Master = 0; >>>> >>>> std::cout << "Hello World! I am " << my_rank << " of " << mpi_size >>>> << std::endl; >>>> } >>>> >>>> void MPI_finalize() >>>> {MPI::Finalize();} >>>> >>>> ================================== >>>> >>>> I believe that it is a terrible practice, if not a violation of the >>>> MPI standard to: >>>> (1) define one's own mpi.h, since this may confuse the preprocessor >>>> (or are there precedence rules in the ANSI C standard?) as to which to >>>> include first (especially if CC=mpicc or CXX=mpicxx), and >>>> (2) overload MPI_Init, even if one is using C++ and relying upon C++ >>>> name-mangling to avoid a symbol conflict. >>>> >>>> The user believes that the inability to overload MPI_Init in C++ means >>>> that MPI is not compliant with the C++ standard. >>>> >>>> Thanks, >>>> >>>> Jeff >>>> >>> >>> >>> _______________________________________________ >>> 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 >> University of Chicago Computation Institute >> jhammond(a)alcf.anl.gov / (630) 252-5381 >> http://www.linkedin.com/in/jeffhammond >> https://wiki.alcf.anl.gov/old/index.php/User:Jhammond >> https://wiki-old.alcf.anl.gov/index.php/User:Jhammond >> >> _______________________________________________ >> 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 William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Paul and Cynthia Saylor Professor of Computer Science University of Illinois Urbana-Champaign

1 0

Re: [Mpi-forum] overloading MPI_Init etc.
by Bronis R. de Supinski 05 Mar '12

05 Mar '12

Um, it is standard practice when using the profiling interface to redefine MPI_* functions. I think this question is one for the C and C++ standards, not MPI... On Sun, 4 Mar 2012, Jeff Hammond wrote: > What was ambiguous to me - looking at 2.6.3-4 - was if using MPI_ was > still a protected namespace in C++. The standard really only says > explicitly that it is so in C. It is only with an indirect argument > that mpi.h can define both C and C++ bindings that I can see C++ code > not being able to define MPI_ functions. > > Note that MyMPI_ was ~exactly~ how I fixed the user's code :-) > > Jeff > > On Sun, Mar 4, 2012 at 9:39 PM, Fab Tillier <ftillier(a)microsoft.com> wrote: >> I don't think defining one's own mpi.h is an issue from a compliance perspective. ï¿½However, overloading MPI_Init definitely is - the standard reserves the MPI_ namespace, so if it isn't in the standard, it's not standard compliant. ï¿½Now, if they named it MyMPI_Init, it'd be fine. >> >> -Fab >> >> Jeff Hammond wrote on Sun, 4 Mar 2012 at 18:22:40 >> >>> I have a user who insists it is MPI-compliant to do the following: >>> >>> user file mpi.h, which is included in the application with #include >>> "mpi.h": ================================== #include <mpi.h> >>> >>> <snip> >>> >>> void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& >>> Master) { >>> ï¿½ MPI::Init(argc, argv); >>> ï¿½ my_rank = MPI::COMM_WORLD.Get_rank(); >>> ï¿½ mpi_size = MPI::COMM_WORLD.Get_size(); >>> ï¿½ Master = 0; >>> >>> ï¿½ std::cout << "Hello World! I am " << my_rank << " of " << mpi_size >>> << std::endl; >>> } >>> >>> void MPI_finalize() >>> {MPI::Finalize();} >>> >>> ================================== >>> >>> I believe that it is a terrible practice, if not a violation of the >>> MPI standard to: >>> (1) define one's own mpi.h, since this may confuse the preprocessor >>> (or are there precedence rules in the ANSI C standard?) as to which to >>> include first (especially if CC=mpicc or CXX=mpicxx), and >>> (2) overload MPI_Init, even if one is using C++ and relying upon C++ >>> name-mangling to avoid a symbol conflict. >>> >>> The user believes that the inability to overload MPI_Init in C++ means >>> that MPI is not compliant with the C++ standard. >>> >>> Thanks, >>> >>> Jeff >>> >> >> >> _______________________________________________ >> 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 > University of Chicago Computation Institute > jhammond(a)alcf.anl.gov / (630) 252-5381 > http://www.linkedin.com/in/jeffhammond > https://wiki.alcf.anl.gov/old/index.php/User:Jhammond > https://wiki-old.alcf.anl.gov/index.php/User:Jhammond > > _______________________________________________ > 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] overloading MPI_Init etc.
by Jeff Hammond 04 Mar '12

04 Mar '12

What was ambiguous to me - looking at 2.6.3-4 - was if using MPI_ was still a protected namespace in C++. The standard really only says explicitly that it is so in C. It is only with an indirect argument that mpi.h can define both C and C++ bindings that I can see C++ code not being able to define MPI_ functions. Note that MyMPI_ was ~exactly~ how I fixed the user's code :-) Jeff On Sun, Mar 4, 2012 at 9:39 PM, Fab Tillier <ftillier(a)microsoft.com> wrote: > I don't think defining one's own mpi.h is an issue from a compliance perspective. ï¿½However, overloading MPI_Init definitely is - the standard reserves the MPI_ namespace, so if it isn't in the standard, it's not standard compliant. ï¿½Now, if they named it MyMPI_Init, it'd be fine. > > -Fab > > Jeff Hammond wrote on Sun, 4 Mar 2012 at 18:22:40 > >> I have a user who insists it is MPI-compliant to do the following: >> >> user file mpi.h, which is included in the application with #include >> "mpi.h": ================================== #include <mpi.h> >> >> <snip> >> >> void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& >> Master) { >> ï¿½ MPI::Init(argc, argv); >> ï¿½ my_rank = MPI::COMM_WORLD.Get_rank(); >> ï¿½ mpi_size = MPI::COMM_WORLD.Get_size(); >> ï¿½ Master = 0; >> >> ï¿½ std::cout << "Hello World! I am " << my_rank << " of " << mpi_size >> << std::endl; >> } >> >> void MPI_finalize() >> {MPI::Finalize();} >> >> ================================== >> >> I believe that it is a terrible practice, if not a violation of the >> MPI standard to: >> (1) define one's own mpi.h, since this may confuse the preprocessor >> (or are there precedence rules in the ANSI C standard?) as to which to >> include first (especially if CC=mpicc or CXX=mpicxx), and >> (2) overload MPI_Init, even if one is using C++ and relying upon C++ >> name-mangling to avoid a symbol conflict. >> >> The user believes that the inability to overload MPI_Init in C++ means >> that MPI is not compliant with the C++ standard. >> >> Thanks, >> >> Jeff >> > > > _______________________________________________ > 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 University of Chicago Computation Institute jhammond(a)alcf.anl.gov / (630) 252-5381 http://www.linkedin.com/in/jeffhammond https://wiki.alcf.anl.gov/old/index.php/User:Jhammond https://wiki-old.alcf.anl.gov/index.php/User:Jhammond

1 0

Re: [Mpi-forum] overloading MPI_Init etc.
by Fab Tillier 04 Mar '12

04 Mar '12

I don't think defining one's own mpi.h is an issue from a compliance perspective. However, overloading MPI_Init definitely is - the standard reserves the MPI_ namespace, so if it isn't in the standard, it's not standard compliant. Now, if they named it MyMPI_Init, it'd be fine. -Fab Jeff Hammond wrote on Sun, 4 Mar 2012 at 18:22:40 > I have a user who insists it is MPI-compliant to do the following: > > user file mpi.h, which is included in the application with #include > "mpi.h": ================================== #include <mpi.h> > > <snip> > > void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& > Master) { > MPI::Init(argc, argv); > my_rank = MPI::COMM_WORLD.Get_rank(); > mpi_size = MPI::COMM_WORLD.Get_size(); > Master = 0; > > std::cout << "Hello World! I am " << my_rank << " of " << mpi_size > << std::endl; > } > > void MPI_finalize() > {MPI::Finalize();} > > ================================== > > I believe that it is a terrible practice, if not a violation of the > MPI standard to: > (1) define one's own mpi.h, since this may confuse the preprocessor > (or are there precedence rules in the ANSI C standard?) as to which to > include first (especially if CC=mpicc or CXX=mpicxx), and > (2) overload MPI_Init, even if one is using C++ and relying upon C++ > name-mangling to avoid a symbol conflict. > > The user believes that the inability to overload MPI_Init in C++ means > that MPI is not compliant with the C++ standard. > > Thanks, > > Jeff >

1 0

overloading MPI_Init etc.
by Jeff Hammond 04 Mar '12

04 Mar '12

I have a user who insists it is MPI-compliant to do the following: user file mpi.h, which is included in the application with #include "mpi.h": ================================== #include <mpi.h> <snip> void MPI_Init(int argc, char* argv[], int& my_rank, int& mpi_size, int& Master) { MPI::Init(argc, argv); my_rank = MPI::COMM_WORLD.Get_rank(); mpi_size = MPI::COMM_WORLD.Get_size(); Master = 0; std::cout << "Hello World! I am " << my_rank << " of " << mpi_size << std::endl; } void MPI_finalize() {MPI::Finalize();} ================================== I believe that it is a terrible practice, if not a violation of the MPI standard to: (1) define one's own mpi.h, since this may confuse the preprocessor (or are there precedence rules in the ANSI C standard?) as to which to include first (especially if CC=mpicc or CXX=mpicxx), and (2) overload MPI_Init, even if one is using C++ and relying upon C++ name-mangling to avoid a symbol conflict. The user believes that the inability to overload MPI_Init in C++ means that MPI is not compliant with the C++ standard. Thanks, Jeff -- Jeff Hammond Argonne Leadership Computing Facility University of Chicago Computation Institute jhammond(a)alcf.anl.gov / (630) 252-5381 http://www.linkedin.com/in/jeffhammond https://wiki.alcf.anl.gov/old/index.php/User:Jhammond https://wiki-old.alcf.anl.gov/index.php/User:Jhammond

1 0

Re: [Mpi-forum] NOTE: Agenda change
by Marc Snir 04 Mar '12

04 Mar '12

313 is not on agenda. It is cleaning up the text on init and finalize that is quite confused (in fact, wrong) as it stands now. Could I start with this tomorrow? This has a strong connection to 310-311, On Feb 29, 2012, at 2:52 PM, Graham, Richard L. wrote: > Note that a large swap was made in the agenda on Monday - the hybrid-wg and the ft-wg plenary sessions were swapped to accommodate someones travel schedule. > > Rich > _______________________________________________ > mpi-forum mailing list > mpi-forum(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

1 0

EuroMPI 2012: Call for Papers
by Rolf Rabenseifner 03 Mar '12

03 Mar '12

Dear HLRS User or member of my course-invitation-list, as member of the program committee, I'm sending you the CFP for the EuroMPI 2012. The deadline for full paper and poster abstract submissions is May 5th, 2012. Best regards Rolf Rabenseifner ------------------------------------------------------------ [Our apologies if you receive multiple copies of this CFP] ------------------------------------------------------------ EuroMPI 2012 CALL FOR PAPERS Recent Advances in Message Passing Interface. 19th European MPI Users' Group Meeting Austrian Academy of Sciences, Vienna, Austria, September 23-26, 2012 www.eurompi2012.org BACKGROUND AND TOPICS EuroMPI is the preeminent meeting for users, developers and researchers to interact and discuss new developments and applications of message-passing parallel computing, in particular in and related to the Message Passing Interface (MPI). The annual meeting has a long, rich tradition, and the 19th European MPI Users' Group Meeting will again be a lively forum for discussion of everything related to usage and implementation of MPI and other parallel programming interfaces. Traditionally, the meeting has focused on the efficient implementation of aspects of MPI, typically on high-performance computing platforms, benchmarking and tools for MPI, short-comings and extensions of MPI, parallel I/O and fault tolerance, as well as parallel applications using MPI. The meeting is open towards other topics, in particular application experience and alternative interfaces for high-performance heterogeneous, hybrid, distributed memory systems. Topics of interest include, but are not limited to: * MPI implementation issues and improvements * Extensions to and shortcomings of MPI * Tools and environments for MPI * Hybrid and heterogeneous programming with MPI and other interfaces * Relation of MPI to alternative interfaces for hybrid/heterogeneous distributed memory systems * Interaction between message-passing software and hardware, in particular new high performance architectures * Fault tolerance in message-passing implementations and systems * Performance evaluation for MPI and MPI based applications * Automatic performance tuning of MPI applications and implementations * Verification of message passing applications and protocols * Applications using message-passing, in particular in Computational Science and Scientific Computing * Non-standard message-passing applications * Parallel algorithms in the message-passing paradigm * Algorithms using the message-passing paradigm The meeting will feature contributed talks on the selected, peer-reviewed papers, invited expert talks covering upcoming and future issues, a vendor session where selected vendors will present their new developments in hybrid and heterogeneous cluster and high-performance architectures, a poster session, and a tutorial day. As in 2011, EuroMPI 2012 will include the special session IMUDI dedicated to improving MPI user and developer interaction. The special session will have the same deadlines as the conference, but its own CFP and feature its own specialist program committee that will review all dedicated IMUDI submissions. This session should attract strongly application focused papers showing potential or problems with MPI and current MPI practice. IMUDI papers may be shorter and may center on problem analysis rather than solution, and will thus contribute to intensify discussion between MPI developers and application programmers. Other, special sessions may be organized, and proposals are solicited. Proceedings for EuroMPI 2012 will be published in Springer Lecture Notes in Computer Science (LNCS). A Special Issue of the Springer "Computing" journal is planned for a selection of the best papers of the conference. SUBMISSION INFORMATION Contributors are invited to submit a full paper as a PDF document not exceeding 10 pages in English. The title page should contain an abstract of at most 100 words and five specific, topical keywords. The paper must be formatted according to the Springer LNCS guidelines. The usage of LaTeX for preparation of the contribution as well as the submission in camera ready format is strongly recommended. Style files can be found with Springer New work that is not yet mature for a full paper, short observations, and similar brief announcements are invited for the poster session. Contributions to the poster session should be submitted in the form of a two page abstract. All contributions will be fully peer reviewed by the program committee. Papers shall be submitted electronically via Easychair, see http://www.easychair.org/conferences/?conf=eurompi2012 Papers with a very strong application focus concerned with use of MPI and possible obstacles should be considered for submission to the IMUDI special session, see http://press.mcs.anl.gov/imudi SCHEDULE, IMPORTANT DATES * Submission of full papers and poster abstracts: May 5th, 2012 * Author Notification: June 11th, 2012 * Camera Ready papers due: July 6th, 2012 * Tutorial(s): September 23rd, 2012 * Conference: September 24-26th, 2012 GENERAL CHAIR Jack Dongarra ORGANIZERS, PC CHAIRS Siegfried Benkner, University of Vienna, sigi(a)par.univie.ac.at Jesper Larsson Traeff, Vienna University of Technology, traff(a)par.tuwien.ac.at PROGRAM COMMITTEE Pavan Balaji, Argonne National Laboratory, USA Siegfried Benkner, University of Vienna, Austria Gil Bloch, Mellanox Technologies George Bosilca, University of Tennessee, Knoxville, USA Ron Brightwell, Sandia National Laboratories, Albuquerque, USA Darius Buntinas, Argonne National Laboratory, USA Franck Cappello, INRIA, France and University of Illinois at Urbana-Champaign, USA Gilles Civario, Irish Centre for High-End Computing Yiannis Cotronis, University of Athens, Greece Jim Cownie, Intel, UK Anthony Danalis, University of Tennessee, Knoxville, USA Bronis R. de Supinski, Lawrence Livermore National Laboratory, USA Luiz DeRose, Cray, USA Edgar Gabriel, University of Houston, USA Brice Goglin, INRIA, France David Goodell, Argonne National Laboratory, USA Ganesh Gopalakrishnan, University of Utah, USA Richard Graham, Oak Ridge National Laboratory, USA William Gropp, University of Illinois at Urbana-Champaign, USA Thomas Herault, University of Tennessee, Knoxville Torsten Hoefler, University of Illinois at Urbana-Champaign, USA Yutaka Ishikawa, Universty of Tokyo, Japan Michael Kagan, Mellanox Technologies Dries Kimpe, Argonne National Laboratory, USA Jesus Labarta, Technical University of Catalonia, Barcelona Supercomputing Center, Spain Dong Li, Oak Ridge National Laboratory, USA Ewing Rusty Lusk, Argonne National Laboratory, USA Amith Rajith Mamidala, IBM, USA Satoshi Matsuoka, Tokyo Institute of Technology, Japan Guillaume Mercier, INRIA, France Bernd Mohr, Juelich Supercomputing Centre (JSC), Germany Matthias Mueller, Center for Information Services and HPC, TU Dresden, Germany Rolf Rabenseifner, High Performance Computing Center Stuttgart (HLRS), Germany Thomas Rauber, University of Bayreuth, Germany Rolf Riesen, IBM, Ireland Robert Ross, Argonne National Laboratory, USA Peter Sanders Karlsruhe Institute of Technology, Germany Mitsuhisa Sato, University of Tsukuba, Japan Saba Sehrish, Northwestern University, USA Christian Siebert, University of Aachen, Germany Stephen Siegel, University of Delaware, USA Anna Sikora, Autonomous University of Barcelona, Spain Jeff Squyres, Cisco, USA Shinji Sumimoto, Fujitsu, Japan Rajeev Thakur, Argonne National Laboratory, USA Vinod Tipparaju, AMD, USA Carsten Trinitis, Technical University of Munich, Germany Denis Trystram, Grenoble Institute of Technology, France Jesper Larsson Traeff, Vienna University of Technology, Austria Keith Underwood, Intel, USA Robert Van De Geijn, The University of Texas at Austin, USA Alan Wagner, University of British Columbia, Canada Roland Wismueller, University of Siegen, Germany Xin Yuan, Florida State University, USA CONFERENCE VENUE EuroMPI 2012 will be held in Vienna, Austria, at the Austrian Academy of Sciences. Vienna is easily accessible by plane and train, and in addition to the conference has a highly attractive cultural heritage. CONFERENCE FEES The conference will hopefully be well supported by sponsors, which will help directly towards keeping the fees low. As far as possible, special student fees will be offered.

1 0