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:B313
DTSTART;TZID=America/New_York:20241117T104000
DTEND;TZID=America/New_York:20241117T111000
UID:submissions.supercomputing.org_SC24_sess732_ws_canopie103@linklings.co
 m
SUMMARY:PULSE: Using Mixed-Quality Models for Reducing Serverless Keep-Ali
 ve Cost
DESCRIPTION:Kausalya Sankaranarayanan, Rohan Basu Roy, and Devesh Tiwari (
 Northeastern University)\n\nThis paper addresses a key challenge with usin
 g serverless computing for machine learning (ML) inference, which is cold 
 starts that occur during initial invocations and container inactivity. Fix
 ed keep-alive policies, like the commonly adopted 10-minute strategy, have
  been implemented by cloud providers to alleviate cold start issues. Howev
 er, the substantial size of ML models poses a significant hurdle, leading 
 to elevated keep-alive costs and potential strain on system resources. In 
 response to these challenges, we introduce PULSE, a dynamic 10-minute keep
 -alive mechanism that employs ML model variants to optimize the balance be
 tween keep-alive costs, accuracy, and service time while avoiding peaks in
  keep-alive memory consumption. Our evaluation, using real-world serverles
 s workloads and commonly used machine learning models, demonstrates reduce
 d keep-alive costs compared to the fixed policy. Additionally, we observe 
 that integrating PULSE improves the performance of existing state-of-the-a
 rt serverless function warm-up strategies.\n\nTag: Cloud Computing, Middle
 ware and System Software, State of the Practice\n\nRegistration Category: 
 Workshop Reg Pass\n\nSession Chairs: Richard Shane Canon (Lawrence Berkele
 y National Laboratory (LBNL)), Alberto Madonna (Swiss National Supercomput
 ing Centre (CSCS)), Claudia Misale (IBM), and Andrew Younge (Sandia Nation
 al Laboratories)\n\n
END:VEVENT
END:VCALENDAR