Cargando…
The CMS software performance at the start of data taking
The CMS software framework (CMSSW) is a complex project evolving very rapidly as the first LHC colliding beams approach. The computing requirements constrain performance in terms of CPU time, memory footprint and event size on disk to allow for planning and managing the computing infrastructure nece...
Autor principal: | |
---|---|
Lenguaje: | eng |
Publicado: |
2009
|
Materias: | |
Acceso en línea: | https://dx.doi.org/10.1109/NSSMIC.2008.4774926 http://cds.cern.ch/record/1270188 |
_version_ | 1780920191139446784 |
---|---|
author | Benelli, Gabriele |
author_facet | Benelli, Gabriele |
author_sort | Benelli, Gabriele |
collection | CERN |
description | The CMS software framework (CMSSW) is a complex project evolving very rapidly as the first LHC colliding beams approach. The computing requirements constrain performance in terms of CPU time, memory footprint and event size on disk to allow for planning and managing the computing infrastructure necessary to handle the needs of the experiment. A performance suite of tools has been developed to track all aspects of code performance, through the software release cycles, allowing for regression and guiding code development for optimization. In this talk, we describe the CMSSW performance suite tools used and present some sample performance results from the release integration process for the CMS software. |
id | cern-1270188 |
institution | Organización Europea para la Investigación Nuclear |
language | eng |
publishDate | 2009 |
record_format | invenio |
spelling | cern-12701882019-09-30T06:29:59Zdoi:10.1109/NSSMIC.2008.4774926http://cds.cern.ch/record/1270188engBenelli, GabrieleThe CMS software performance at the start of data takingComputing and ComputersThe CMS software framework (CMSSW) is a complex project evolving very rapidly as the first LHC colliding beams approach. The computing requirements constrain performance in terms of CPU time, memory footprint and event size on disk to allow for planning and managing the computing infrastructure necessary to handle the needs of the experiment. A performance suite of tools has been developed to track all aspects of code performance, through the software release cycles, allowing for regression and guiding code development for optimization. In this talk, we describe the CMSSW performance suite tools used and present some sample performance results from the release integration process for the CMS software.oai:cds.cern.ch:12701882009 |
spellingShingle | Computing and Computers Benelli, Gabriele The CMS software performance at the start of data taking |
title | The CMS software performance at the start of data taking |
title_full | The CMS software performance at the start of data taking |
title_fullStr | The CMS software performance at the start of data taking |
title_full_unstemmed | The CMS software performance at the start of data taking |
title_short | The CMS software performance at the start of data taking |
title_sort | cms software performance at the start of data taking |
topic | Computing and Computers |
url | https://dx.doi.org/10.1109/NSSMIC.2008.4774926 http://cds.cern.ch/record/1270188 |
work_keys_str_mv | AT benelligabriele thecmssoftwareperformanceatthestartofdatataking AT benelligabriele cmssoftwareperformanceatthestartofdatataking |