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:20260202T201259Z
LOCATION:Second Floor Atrium
DTSTART;TZID=America/Chicago:20251121T080000
DTEND;TZID=America/Chicago:20251121T120000
UID:submissions.supercomputing.org_SC25_sess620_post171@linklings.com
SUMMARY:Inference-as-a-Service Prototype at NERSC
DESCRIPTION:Colin Thomas (University of Notre Dame); Po-Han Huang (Georgia
  Institute of Technology); Hilary Utaegbulam (University of Rochester); Jo
 hannes Blaschke (ESnet; Lawrence Berkeley National Laboratory (LBNL)); Bru
 no Coimbra (Fermi National Laboratory); Pengfei Ding, Xiangyang Ju, and An
 drew Naylor (ESnet; Lawrence Berkeley National Laboratory (LBNL)); and Mic
 hael Wang (Fermi National Laboratory)\n\nThe increasing scale and complexi
 ty of scientific experiments has led to a growing need for efficient and s
 calable machine learning model inference serving systems. High-energy phys
 ics experiments and simulations of complex climate models involve petabyte
 s of data and massive amounts of computational resources to produce accura
 te results. Thus, scientists are increasingly turning to utilize ML techni
 ques to analyze and interpret the vast amount of data generated by these e
 xperiments. \n\nHowever, the deployment of ML models in scientific applica
 tions poses significant challenges. Traditional approaches to deploying ML
  models by individual users with local resources or small clusters often s
 uffer from long startup costs and inefficient resource utilization. To add
 ress this challenge, we present a prototyped system that provides on-deman
 d inference serving capabilities for multiple scientific ML models. Our sy
 stem is deployed across the NERSC Perlmutter supercomputer and the NERSC K
 8s cluster, enabling on-demand scalability.\n\nTag: Research & ACM SRC Pos
 ters\n\nRegistration Category: Technical Program Reg Pass\n\nSession Chair
 s: Kento Sato (RIKEN Center for Computational Science (R-CCS)); Anja Gerbe
 s (Georg-August-Universität Göttingen); and Chris Schlipalius (Pawsey Supe
 rcomputing Research Centre; Commonwealth Scientific and Industrial Researc
 h Organisation (CSIRO), Australia)\n\n
END:VEVENT
END:VCALENDAR