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:20260522T150122Z
LOCATION:D173
DTSTART;TZID=America/Chicago:20181111T112000
DTEND;TZID=America/Chicago:20181111T114500
UID:submissions.supercomputing.org_SC18_sess163_ws_works104@linklings.com
SUMMARY:LOS: Level Order Sampling for Task Graph Scheduling on Heterogeneo
 us Resources
DESCRIPTION:Carl Witt (Humboldt University of Berlin); Sam Wheating (Unive
 rsity of Victoria, British Columbia); and Ulf Leser (Humboldt University o
 f Berlin)\n\nList scheduling is an approach to task graph scheduling that 
 has been shown to work well for scheduling tasks with data dependencies on
  heterogeneous resources. Key to the performance of a list scheduling heur
 istic is its method to prioritize the tasks, and various ranking schemes h
 ave been proposed in the literature. We propose a method that combines mul
 tiple random rankings instead of a using a deterministic ranking scheme.\n
 \nWe introduce L-Orders, which are a subset of all topological orders of a
  directed acyclic graph. L-Orders can be used to explore targeted regions 
 of the space of all topological orders.  Using the observation that the ma
 kespans in one such region are often approximately normal distributed, we 
 estimate the expected time to solution improvement in certain regions of t
 he search space. We combine targeted search and improvement time estimatio
 ns into a time budgeted search algorithm that balances exploration and exp
 loitation of the search space.  In 40,500 experiments, our schedules are 5
 % shorter on average and up to 40% shorter in extreme cases than schedules
  produced by HEFT.\n\nTag: Reproducibility, Scientific Computing, Scientif
 ic Workflows, Workflows, HPC, Data Intensive\n\nRegistration Category: Wor
 kshop Reg Pass\n\n
END:VEVENT
END:VCALENDAR