BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20220812T074335Z
LOCATION:Darwin Room
DTSTART;TZID=Europe/Stockholm:20220628T170000
DTEND;TZID=Europe/Stockholm:20220628T173000
UID:submissions.pasc-conference.org_PASC22_sess151_msa248@linklings.com
SUMMARY:The Performance-Portability Trade-Off Challenge in Next Generation
  Particle-In-Cell Simulations
DESCRIPTION:Minisymposium\n\nThe Performance-Portability Trade-Off Challen
 ge in Next Generation Particle-In-Cell Simulations\n\nTan, Luedtke, Taufer
 , Albright\n\nModern systems are becoming increasingly more heterogeneous,
  making performance optimizations of individual Particle-In-Cell (PIC) cod
 es progressively more challenging. Leveraging performance portability fram
 eworks such as Kokkos allows for a single source code to run on all major 
 systems with good performance. However, mapping a single programming model
  to different hardware models naturally leads to inefficiencies and leaves
  room for improvements. In this talk we demonstrate the unique challenges 
 involved in preparing the Vector Particle-In-Cell (VPIC) code, a general p
 urpose particle-in-cell simulation code for modeling plasma phenomena, for
  operation at exascale, outlining important optimizations to make VPIC eff
 icient demonstrate the VPIC performance on different platforms (including 
 three of the top 10 Top500 systems). We discuss the work undertaken in ada
 pting VPIC to exploit Kokkos for code portability and highlight the enhanc
 ements to VPIC's modeling capabilities to achieve performance. We demonstr
 ate the VPIC performance on up to 17,100 GPUs and provide insights on the 
 impacts of VPIC for scientific discovery for the next generation exascale 
 supercomputers.\n\nDomain: Computer Science and Applied Mathematics, Physi
 cs
END:VEVENT
END:VCALENDAR