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:20181221T160726Z
LOCATION:D166
DTSTART;TZID=America/Chicago:20181111T143000
DTEND;TZID=America/Chicago:20181111T150000
UID:submissions.supercomputing.org_SC18_sess174_ws_exampi105@linklings.com
SUMMARY:A Novel Approach to Supporting Communicators for In-Switch Process
 ing of MPI Collectives
DESCRIPTION:Workshop\nExascale, MPI, Networks, System Software, Workshop R
 eg Pass\n\nA Novel Approach to Supporting Communicators for In-Switch Proc
 essing of MPI Collectives\n\nStern, Herbordt\n\nMPI collective operations 
 can often be performance killers in HPC applications, and we seek to solve
  this bottleneck by offloading them to switch hardware. We've seen from pr
 evious works including our own that moving collectives into the network of
 fers significant performance benefits. However, there has been little adva
 ncement in providing support for sub-communicator collectives in an effici
 ent manner. Using FPGAs, which provide the ability to couple communication
  and computation, we have designed an in-switch hardware accelerator to im
 plement support for MPI communicators and full offload of MPI collectives.
  With our design, preliminary results show that we can achieve up to a 10x
  speedup over conventional clusters for short message collectives over irr
 egular intra-communicators.
URL:https://sc18.supercomputing.org/presentation/?id=ws_exampi105&sess=ses
 s174
END:VEVENT
END:VCALENDAR