Publicação

Removing inefficiencies from scientific code: the study of the Higgs boson couplings to top quarks

Ver documento

Detalhes bibliográficos
Resumo:This paper presents a set of methods and techniques to remove inefficiencies in a data analysis application used in searches by the ATLAS Experiment at the Large Hadron Collider. Profiling scientific code helped to pinpoint design and runtime inefficiencies, the former due to coding and data structure design. The data analysis code used by groups doing searches in the ATLAS Experiment contributed to clearly identify some of these inefficiencies and to give suggestions on how to prevent and overcome those common situations in scientific code to improve the efficient use of available computational resources in a parallel homogeneous platform.
Autores principais:Pereira, André Martins
Outros Autores:Onofre, A.; Proença, Alberto José
Assunto:Scientific computing High performance computing Code efficiency ATLAS Experiment Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática Ciências Naturais::Ciências Físicas
Ano:2014
País:Portugal
Tipo de documento:comunicação em conferência
Tipo de acesso:acesso aberto
Instituição associada:Universidade do Minho
Idioma:inglês
Origem:RepositóriUM - Universidade do Minho
Descrição
Resumo:This paper presents a set of methods and techniques to remove inefficiencies in a data analysis application used in searches by the ATLAS Experiment at the Large Hadron Collider. Profiling scientific code helped to pinpoint design and runtime inefficiencies, the former due to coding and data structure design. The data analysis code used by groups doing searches in the ATLAS Experiment contributed to clearly identify some of these inefficiencies and to give suggestions on how to prevent and overcome those common situations in scientific code to improve the efficient use of available computational resources in a parallel homogeneous platform.