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:20240116T191702Z
LOCATION:DEF Concourse
DTSTART;TZID=America/Denver:20231114T100000
DTEND;TZID=America/Denver:20231114T170000
UID:submissions.supercomputing.org_SC23_sess291_rpost220@linklings.com
SUMMARY:Modeling Parallel Programs Using Large Language Models
DESCRIPTION:Posters, Research Posters\n\nDaniel Nichols (University of Mar
 yland); Aniruddha Marathe, Harshitha Menon, and Todd Gamblin (Lawrence Liv
 ermore National Laboratory (LLNL)); and Abhinav Bhatele (University of Mar
 yland)\n\nIn the past year a large number of large language model (LLM) ba
 sed tools for software development have been released.  These tools have t
 he capability to assist developers with many of the difficulties that aris
 e from the ever-growing complexity in the software stack.  As we enter the
  exascale era, with a diverse set of emerging hardware and programming par
 adigms, developing, optimizing, and maintaining parallel software is becom
 ing burdensome for developers.  While LLM-based coding tools have been ins
 trumental in revolutionizing software development, mainstream models are n
 ot designed, trained, or tested on High Performance Computing (HPC) proble
 ms.  We present a LLM fine-tuned on HPC data and demonstrate its effective
 ness in HPC code generation, OpenMP parallelization, and performance model
 ing.\n\nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pass
END:VEVENT
END:VCALENDAR