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:20260202T201806Z
LOCATION:274
DTSTART;TZID=America/Chicago:20251117T105000
DTEND;TZID=America/Chicago:20251117T111000
UID:submissions.supercomputing.org_SC25_sess215_ws_ai4s105@linklings.com
SUMMARY:Guiding Application Users via Estimation of Computational Resource
 s for Massively Parallel Chemistry Computations
DESCRIPTION:Tanzila Tabassum (Louisiana State University), Omer Subasi and
  Ajay Panyala (Pacific Northwest National Laboratory (PNNL)), Epiya Ebiapi
 a and Gerald Baumgartner (Louisiana State University), Erdal Mutlu (Pacifi
 c Northwest National Laboratory (PNNL)), P. Saday Sadayappan (University o
 f Utah), and Karol Kowalski (Pacific Northwest National Laboratory (PNNL))
 \n\nWe develop several machine learning (ML)-based methods to estimate res
 ources required for massively-parallel chemistry computations, e.g., coupl
 ed-cluster methods, to guide application users before they run expensive s
 imulations on supercomputers. By estimating computational resources, our M
 L-based methods predict optimal runtime parameters (number of nodes, tile 
 sizes, etc.). With these predictions, we answer users' questions such as i
 ) what is the minimum execution time for a given problem size?, ii) what a
 re the number of nodes and tiles sizes to achieve this minimum execution t
 ime?, and iii) how about a supercomputer for which the number of past appl
 ication runs that an ML model can be trained by is limited? Our work offer
 s several ML models trained by the simulations of a coupled-cluster method
  run on Frontier, Aurora and Perlmutter supercomputers. We devise two stra
 tegies based on active and generative learning. By inquiring about costs b
 eforehand, users can save significant amount of expenses.\n\nRecording: Li
 vestreamed, Recorded\n\nRegistration Category: Technical Program Reg Pass,
  Workshop Reg Pass\n\nSession Chairs: Gokcen Kestor (Barcelona Supercomput
 ing Center (BSC); University of California, Merced); Dong Li (University o
 f California, Merced); and Murali Emani (Argonne National Laboratory (ANL)
 )\n\n
END:VEVENT
END:VCALENDAR