BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240116T191701Z
LOCATION:DEF Concourse
DTSTART;TZID=America/Denver:20231114T100000
DTEND;TZID=America/Denver:20231114T170000
UID:submissions.supercomputing.org_SC23_sess291_rpost127@linklings.com
SUMMARY:MPI Performance Analysis in Vlasiator:  Unraveling Communication B
 ottlenecks
DESCRIPTION:Posters, Research Posters\n\nJennifer Faj, Jeremy J. Williams,
  and Ivy B. Peng (KTH Royal Institute of Technology, Sweden); Urs Ganse, M
 arkus Battarbee, Yann Pfau-Kempf, Leo Kotipalo, and Minna Palmroth (Univer
 sity of Helsinki); and Stefano Markidis (KTH Royal Institute of Technology
 , Sweden)\n\nVlasiator is a popular and powerful massively parallel code f
 or accurate magnetospheric and solar wind plasma simulations. This work pr
 ovides an in-depth analysis of Vlasiator, focusing on MPI performance usin
 g the Integrated Performance Monitoring (IPM) tool. We show that MPI non-b
 locking point-to-point communication accounts for most of the communicatio
 n time. The communication topology shows a large number of MPI messages ex
 changing data in a six-dimensional grid. We also show that relatively larg
 e messages are used in MPI communication, reaching up to 256MB. As a commu
 nication-bound application, we found that using OpenMP in Vlasiator is cri
 tical for eliminating intra-node communication. Our results provide import
 ant insights for optimizing Vlasiator for the upcoming Exascale machines.\
 n\nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pass
END:VEVENT
END:VCALENDAR