[Mpi3-hybridpm] [MPI Forum] #208: MPI3 Hybrid Programming: multiple endpoints per collective, option 1

MPI Forum mpi-22 at lists.mpi-forum.org
Wed Jan 6 07:48:08 CST 2010


#208: MPI3 Hybrid Programming: multiple endpoints per collective, option 1
----------------------------------------------------------+-----------------
                    Reporter:  dougmill@…                 |                       Owner:  dougmill@…             
                        Type:  Enhancements to standard   |                      Status:  new                    
                    Priority:  Not ready / author rework  |                   Milestone:  2010/01/19 Atlanta, USA
                     Version:  MPI 3.0                    |                    Keywords:                         
              Implementation:  Unnecessary                |           Author_bill_gropp:  0                      
          Author_rich_graham:  0                          |           Author_adam_moody:  0                      
      Author_torsten_hoefler:  0                          |        Author_dick_treumann:  0                      
Author_jesper_larsson_traeff:  0                          |       Author_george_bosilca:  0                      
           Author_david_solt:  0                          |   Author_bronis_de_supinski:  0                      
        Author_rajeev_thakur:  0                          |         Author_jeff_squyres:  0                      
    Author_alexander_supalov:  0                          |    Author_rolf_rabenseifner:  0                      
----------------------------------------------------------+-----------------

Comment(by dougmill@…):

 Perhaps "active group" needs to be more flexible - arbitrary groups of
 endpoints can spontaneously form and dissolve. This would probably lean
 towards a join/leave semantics on top of the attach/detach, but adding
 some parameters to join that indicate some sort of group identity. In this
 way, endpoints that truly wish to do communication independently of each
 other are not penalized by barriers when they "leave".

-- 
Ticket URL: <https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/208#comment:3>
MPI Forum <https://svn.mpi-forum.org/>
MPI Forum




More information about the mpiwg-hybridpm mailing list