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:20211207T055354Z
LOCATION:Second Floor Atrium
DTSTART;TZID=America/Chicago:20211117T083000
DTEND;TZID=America/Chicago:20211117T170000
UID:submissions.supercomputing.org_SC21_sess279_rpost171@linklings.com
SUMMARY:Accelerating Parallel Monte Carlo Simulations for Statistical Phys
 ics: Portability on Many-Core Processors
DESCRIPTION:Posters, Research Posters\n\nAccelerating Parallel Monte Carlo
  Simulations for Statistical Physics: Portability on Many-Core Processors\
 n\nSattar, Li\n\nMonte Carlo (MC) simulations are important tools for stud
 ying thermodynamics and materials properties at finite temperature. Scalin
 g up to larger systems allows for simulations comparable to experimental l
 ength scale. In this work, we focus on improving the weak scaling performa
 nce to simulate large magnetic spin systems up to millions of atoms. We pa
 rallelize the computation of physical observables within a single random w
 alker using checkerboard domain decomposition implemented with OpenMP and 
 compare the performance of the same code base on different many-core proce
 ssors. We analyze the effects of thread affinity and hyper-threading on th
 e runtime performance and achieve 14x/12x/6x speedups for Arm/Intel/AMD ar
 chitectures respectively for the largest system we report in this work (ab
 out 4.2 millions spins). This study is helpful in informing end-users on t
 he choice of the best multi-core architecture suited to their needs for si
 milar pattern of parallel applications.\n\nRegistration Category: Tech Pro
 gram Reg Pass, Exhibit Hall Only
END:VEVENT
END:VCALENDAR