BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250626T233527Z
LOCATION:B302-B305
DTSTART;TZID=America/New_York:20241120T100000
DTEND;TZID=America/New_York:20241120T170000
UID:submissions.supercomputing.org_SC24_sess533_post210@linklings.com
SUMMARY:Lagrangian Particle-Tracking in GPU-Enabled Extreme Scale Turbulen
 ce Simulations
DESCRIPTION:Rohini Uma-Vaideswaran and P.K. Yeung (Georgia Institute of Te
 chnology) and Kiran Ravikumar (Science and Technology Corporation, NASA Am
 es Research Center)\n\nMany practical turbulent flow phenomena are natural
 ly studied using a Lagrangian approach that treats the fluid medium as a c
 ollection of infinitesimal fluid particles. We present a GPU-accelerated a
 lgorithm for tracking particles in direct numerical simulations of isotrop
 ic turbulence, scaling up to 32768^3 using the world's first exascale comp
 uter (Frontier). Cubic spline interpolation is used to compute the particl
 e velocity as the particles wander among sub-domains held by different par
 allel processes. We use a programming model that minimizes host-device dat
 a transfer by leveraging\nmemory parity between the CPU and GPU, reduces c
 ommunication costs through a local decomposition for the particles, and us
 es OpenMP offloading on the GPU to accelerate the computation of cubic spl
 ine coefficients. The result is an algorithm shown to attain good weak sca
 ling and strong scaling at problem sizes close to the capacity supported b
 y the machine, at a cost nearly independent of the particle count.\n\nRegi
 stration Category: Tech Program Reg Pass, Exhibits Reg Pass\n\nSession Cha
 irs: Ayesha Afzal (Friedrich-Alexander University, Erlangen-Nuremberg; Erl
 angen National High Performance Computing Center); Sally Ellingson (Univer
 sity of Kentucky); and Alan Sussman (University of Maryland)\n\n
END:VEVENT
END:VCALENDAR