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:20260202T201804Z
LOCATION:241
DTSTART;TZID=America/Chicago:20251116T110000
DTEND;TZID=America/Chicago:20251116T113000
UID:submissions.supercomputing.org_SC25_sess209_ws_fgai109@linklings.com
SUMMARY:Agentic AI vs ML-based Autotuning: A Comparative Study for Loop Re
 ordering Optimization
DESCRIPTION:Miguel Romero Rosas (University of Delaware), Khaled Ibrahim (
 Lawrence Berkeley National Laboratory (LBNL)), and Rudolf Eigenmann (Unive
 rsity of Delaware)\n\nHigh Performance Computing (HPC)  applications rely 
 heavily on code optimizations to achieve good performance on modern CPU an
 d GPU architectures. Traditional Machine Learning autotuning approaches ha
 ve demonstrated success in exploring high-dimensional spaces, but they oft
 en require expensive compile-run evaluations and lack adaptability for lar
 ge HPC applications. The recent advances in Large Language Models (LLMs) a
 nd Agentic AI systems raise intriguing questions about the potential of th
 ese approaches to address specific optimization methodologies. This work a
 ims to answer an essential question for the HPC community: "How Agentic AI
  Systems Compare to Traditional ML Autotuning Techniques?"\nTo address thi
 s question, we present a comparative analysis between a traditional ML-bas
 ed optimization approach and an Agentic AI system, evaluating their respec
 tive capabilities and limitations for loop-level optimization. In addition
 , we introduced a new Agentic AI system named LoopGen-AI using three diffe
 rent Large Language Models: GPT-4.1, Claude 4.0, and Gemini 2.5.\n\nRecord
 ing: Livestreamed, Recorded\n\nRegistration Category: Technical Program Re
 g Pass, Workshop Reg Pass\n\nSession Chairs: Charlie Catlett (Argonne Nati
 onal Laboratory (ANL), University of Chicago); Javier Aula-Blasco (Barcelo
 na Supercomputing Center (BSC)); Valerie Taylor (Argonne National Laborato
 ry (ANL), University of Chicago); Laura Morselli (CINECA); and Rio Yokota 
 (Institute of Science Tokyo)\n\n
END:VEVENT
END:VCALENDAR