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:20190719T085744Z
LOCATION:HG F 1
DTSTART;TZID=Europe/Stockholm:20190614T093000
DTEND;TZID=Europe/Stockholm:20190614T100000
UID:submissions.pasc-conference.org_PASC19_sess180_pap_jan120@linklings.co
 m
SUMMARY:Towards Continuous Benchmarking: An Automated Performance Evaluati
 on Framework for High Performance Software
DESCRIPTION:Paper\nComputer Science and Applied Mathematics\n\nTowards Con
 tinuous Benchmarking: An Automated Performance Evaluation Framework for Hi
 gh Performance Software\n\nAnzt, Chen, Cojean, Dongarra, Flegar...\n\nWe p
 resent an automated performance evaluation framework that enables an autom
 ated workflow for testing and performance evaluation of software libraries
 . Integrating this component into an ecosystem enables sustainable softwar
 e development as a community effort via a web application for interactivel
 y evaluating the performance of individual software components. The perfor
 mance evaluation tool is based exclusively on web technologies, which remo
 ves the burden of downloading performance data or installing additional so
 ftware. We employ this framework for the Ginkgo software ecosystem, but th
 e framework can be used with essentially any software project, including t
 he comparison between different software libraries. The Continuous Integra
 tion (CI) framework of Ginkgo is also extended to automatically run a benc
 hmark suite on predetermined HPC systems, store the state of the machine a
 nd the environment along with the compiled binaries, and collect results i
 n a publicly accessible performance data repository based on git. The Gink
 go performance explorer (GPE) can be used to retrieve the performance data
  from the repository, and visualizes it in a web browser. GPE also impleme
 nts an interface that allows users to write scripts, archived in a git rep
 ository, to extract particular data, compute particular metrics, and visua
 lize them in many different formats (as specified by the script). The comb
 ination of these approaches creates a workflow which enables performance r
 eproducibility and software sustainability of scientific software. In this
  paper, we present example scripts that extract and visualize performance 
 data for Ginkgo's SpMV kernels that allow to identify the optimal kernel f
 or specific problem characteristics.<br /><br />Full paper: https://doi.or
 g/10.1145/3324989.3325719
END:VEVENT
END:VCALENDAR