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:20260422T000711Z
LOCATION:501-502
DTSTART;TZID=America/Denver:20231112T110000
DTEND;TZID=America/Denver:20231112T111500
UID:submissions.supercomputing.org_SC23_sess416_ws_hppss101@linklings.com
SUMMARY:Maximizing Data Utility for HPC Python Workflow Execution
DESCRIPTION:Thanh Son Phung (University of Notre Dame), Ben Clifford (CQX 
 Limited), Kyle Chard (University of Chicago), and Douglas Thain (Universit
 y of Notre Dame)\n\nLarge-scale HPC workflows are increasingly implemented
  in dynamic languages such as Python, which allow for more rapid developme
 nt than traditional techniques. However, the cost of executing Python appl
 ications at scale is often dominated by the distribution of common dataset
 s and complex software dependencies. As the application scales up, data di
 stribution becomes a limiting factor that prevents scaling beyond a few hu
 ndred nodes. To address this problem, we present the integration of Parsl 
 (a Python-native parallel programming library) with TaskVine (a data-inten
 sive workflow execution engine). Instead of relying on a shared filesystem
  to provide data to tasks on demand, Parsl is able to express advance data
  needs to TaskVine, which then performs efficient data distribution at run
 time. This combination provides a performance speedup of 1.48x over the ty
 pical method of on-demand paging from the shared filesystem, while also pr
 oviding an average task speedup of 1.79x with 2048 tasks and 256 nodes.\n\
 nTag: Applications, Distributed Computing, Large Scale Systems, Programmin
 g Frameworks and System Software, Runtime Systems\n\nRegistration Category
 : Workshop Reg Pass\n\nSession Chairs: Sam Foreman (Argonne National Labor
 atory (ANL)); Daniel Margala (Lawrence Berkeley National Laboratory (LBNL)
 ); Pete Mendygral (Hewlett Packard Enterprise (HPE)); Laurie A. Stephey (L
 awrence Berkeley National Laboratory (LBNL), National Energy Research Scie
 ntific Computing Center (NERSC)); and Rollin Thomas (Lawrence Berkeley Nat
 ional Laboratory (LBNL), National Energy Research Scientific Computing Cen
 ter (NERSC))\n\n
END:VEVENT
END:VCALENDAR