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:20250626T234541Z
LOCATION:B309
DTSTART;TZID=America/New_York:20241119T160000
DTEND;TZID=America/New_York:20241119T163000
UID:submissions.supercomputing.org_SC24_sess401_pap605@linklings.com
SUMMARY:Static Generation of Efficient OpenMP Offload Data Mappings
DESCRIPTION:Luke Marzen, Akash Dutta, and Ali Jannesari (Iowa State Univer
 sity)\n\nIncreasing heterogeneity in HPC architectures and compiler advanc
 ements have led to OpenMP being frequently used to enable computations on 
 heterogeneous devices. However, the efficient movement of data on heteroge
 neous computing platforms is crucial for achieving high utilization. Progr
 ammers must explicitly map data between the host and connected accelerator
  devices to achieve efficient data movement. Ensuring efficient data trans
 fer requires programmers to reason about complex data flow. This can be a 
 laborious and error-prone process since the programmer must keep a mental 
 model of data validity and lifetime spanning multiple data environments. W
 e present a static analysis tool, OMPDart (OpenMP Data Reduction Tool), fo
 r OpenMP programs that models data dependencies between host and device re
 gions and applies source code transformations to achieve efficient data tr
 ansfer. Our evaluations on nine HPC benchmarks demonstrate that OMPDart is
  capable of generating effective data mapping constructs that substantiall
 y reduce data transfer between host and device.\n\nTag: Accelerators, Comp
 ilers, Heterogeneous Computing, Performance Evaluation and/or Optimization
  Tools\n\nRegistration Category: Tech Program Reg Pass\n\nSession Chair: S
 ascha Hunold (Technical University of Vienna)\n\n
END:VEVENT
END:VCALENDAR