BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20181221T160730Z
LOCATION:C140/142
DTSTART;TZID=America/Chicago:20181114T110000
DTEND;TZID=America/Chicago:20181114T113000
UID:submissions.supercomputing.org_SC18_sess188_pap463@linklings.com
SUMMARY:Framework for Scalable Intra-Node Collective Operations Using Shar
 ed Memory
DESCRIPTION:Paper\nArchitectures, MPI, Networks, Performance, Programming 
 Systems, State of the Practice, Tech Program Reg Pass\n\nFramework for Sca
 lable Intra-Node Collective Operations Using Shared Memory\n\nJain, Kaleem
 , Balmana, Langer, Durnov...\n\nCollective operations are used in MPI prog
 rams to express common communication patterns, collective computations, or
  synchronizations. In many collectives, such as barrier or allreduce, the 
 intra-node component of the collective is in the critical path, as the int
 er-node communication cannot start until the intra-node component has been
  executed. Thus, with increasing number of core counts in each node, intra
 -node optimizations that leverage the intra-node shared memory become incr
 easingly important.\n\nIn this paper, we focus on the performance benefit 
 of optimizing intra-node collectives using shared memory. We optimize seve
 ral collectives using the primitives in broadcast and reduce as building b
 locks for other collectives. A comparison of our implementation on top of 
 MPICH shows significant performance speedups with respect to the original 
 MPICH implementation, MVAPICH, and OpenMPI, among others.
URL:https://sc18.supercomputing.org/presentation/?id=pap463&sess=sess188
END:VEVENT
END:VCALENDAR