BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//juliacon-2026//talk//DAY8TS
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-juliacon-2026-DAY8TS@pretalx.com
DTSTART;TZID=CET:20260812T110000
DTEND;TZID=CET:20260812T113000
DESCRIPTION:In this talk\, we present techniques for implementing efficient
  GPU-based random sampling\nalgorithms\, including rejection sampling and 
 multi-stage sampling pipelines using\nKernelAbstractions.jl. We discuss st
 rategies to avoid costly synchronizations between host\nand device\, manag
 e divergent execution flows\, and schedule heterogeneous\nworkloads entire
 ly on device. Particular attention is given to structuring rejection\nsamp
 ling and sequential transport algorithms in ways that preserve GPU occupan
 cy while\nmaintaining statistical correctness.\n\nTo demonstrate the appli
 cability of these techniques\, we present results from two Julia\npackages
  developed for plasma and high-energy physics applications. The first\,\nQ
 EDevents.jl\, focuses on Monte-Carlo event generation for high-multiplicit
 y\nscattering processes in quantum electrodynamics\, building on the gener
 ic sampling\nframework RejectionSamplers.jl and the QuantumElectrodynamics
 .jl ecosystem. The second\,\nPhotonTransport.jl implements a Monte-Carlo p
 hoton transport code for warm-dense matter.\nTogether\, these case studies
  illustrate how Julia enables both the expressive\nimplementation of paral
 lel sampling algorithms and near-peak accelerator performance\,\nwhile mai
 ntaining portability across hardware backends.
DTSTAMP:20260428T150904Z
LOCATION:Room 5
SUMMARY:Performance-Portable Random Sampling in Julia: Event Generation and
  Particle Transport on GPUs - Uwe Hernandez Acosta
URL:https://pretalx.com/juliacon-2026/talk/DAY8TS/
END:VEVENT
END:VCALENDAR
