BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250822T115805Z
LOCATION:Campussaal - Plenary Room
DTSTART;TZID=Europe/Stockholm:20250616T102000
DTEND;TZID=Europe/Stockholm:20250616T105000
UID:submissions.pasc-conference.org_PASC25_sess149_pos133@linklings.com
SUMMARY:P12 - Enhancing Productivity and Performance Analysis on Euro HPC 
 Systems
DESCRIPTION:Michael Redenti and Nitin Shukla (CINECA)\n\nLarge-scale EuroH
 PC High-Performance Computing (HPC) systems, such as Leonardo and Lumi, pr
 esent significant challenges for developers. A key difficulty is adapting 
 their software to new architectures, accelerators and different compiler o
 ptions in order to fully leverage available resources. As a result, develo
 pers often spend a substantial amount of time manually running performance
  and scalability tests to ensure changes do not degrade performance or com
 promise portability across platforms. To streamline these repetitive, manu
 al steps that hinder productivity, we propose an automated framework for i
 ntegrating HPC performance testing and benchmarking into a Continuous Inte
 gration (CI) pipeline. By extending the ReFrame framework, a widely used H
 PC regression and benchmarking tool, our approach automates the monitoring
  of application performance. This enables both strong and weak scaling ben
 chmarks and performance portability tests across multiple architectures. T
 he system collects metrics, generates visual reports, and alerts developer
 s to any performance regressions. With these capabilities, HPC developers 
 can focus on scientific advancements rather than repetitive, time-consumin
 g testing and benchmarking. We demonstrate this workflow using the ECsim s
 pace-plasma physics application, illustrating how integrating DevOps pract
 ices and ReFrame-driven automation can streamline HPC software development
 .\n\nSession Chair: Chris Cantwell (Imperial College London)\n\n
END:VEVENT
END:VCALENDAR
