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:20250626T233528Z
LOCATION:B302-B305
DTSTART;TZID=America/New_York:20241120T100000
DTEND;TZID=America/New_York:20241120T170000
UID:submissions.supercomputing.org_SC24_sess533_post205@linklings.com
SUMMARY:JUmPER: Performance Data Monitoring, Instrumentation and Visualiza
 tion for Jupyter Notebooks
DESCRIPTION:Elias Werner and Anton Rygin (Center for Scalable Data Analyti
 cs and Artificial Intelligence (ScaDS.AI) Dresden/Leipzig; Technical Unive
 rsity Dresden, Center for Interdisciplinary Digital Sciences (CIDS)); Andr
 eas Gocht-Zech (German Environment Agency, AI Lab); and Sebastian Döbel an
 d Matthias Lieber (Technical University Dresden, ZIH; Technical University
  Dresden, Center for Interdisciplinary Digital Sciences (CIDS))\n\nComputa
 tional performance, e.g. CPU or GPU utilization, is crucial for analyzing 
 machine learning (ML) applications and their resource-efficient deployment
 . However, the ML community often lacks accessible tools for holistic perf
 ormance engineering, especially during exploratory programming such as imp
 lemented by Jupyter. Therefore, we present JUmPER, a Jupyter kernel that s
 upports coarse-grained performance monitoring and fine-grained analysis ta
 sks of user code in Jupyter. \n\nJUmPER collects system metrics and stores
  them alongside executed user code. Additionally, code instrumentation can
  be enabled to collect performance events using Score-P. Built-in Jupyter 
 magic commands provide visualizations of the monitored performance data di
 rectly in Jupyter. In addition, JUmPER preserves the exploratory programmi
 ng experience by seamlessly integrating with Jupyter and reducing kernel r
 untime overhead through in-memory (pipe) communication and parallel marsha
 lling of Python's interpreter state for the Score-P execution.\n\nJUmPER t
 hus provides a low-hurdle infrastructure for performance engineering in Ju
 pyter and supports resource-efficient ML applications.\n\nRegistration Cat
 egory: Tech Program Reg Pass, Exhibits Reg Pass\n\nSession Chairs: Ayesha 
 Afzal (Friedrich-Alexander University, Erlangen-Nuremberg; Erlangen Nation
 al High Performance Computing Center); Sally Ellingson (University of Kent
 ucky); and Alan Sussman (University of Maryland)\n\n
END:VEVENT
END:VCALENDAR