In the last years, the development of broadband chirped-pulse Fourier transform microwave spectrometers has revolutionized the field of rotational spectroscopy. Currently, it is possible to experimentally obtain a large quantity of spectra that would be difficult to analyze manually due to two main reasons. First, recent instruments allow obtaining a considerable amount of data in very short times, and second, it is possible to analyze complex mixtures of molecules that all contribute to the density of the spectra. AUTOFIT is a spectral assignment software application that was developed in 2013 to support and facilitate the analysis. Notwithstanding the benefits AUTOFIT brings in terms of automation of the analysis of the accumulated data, it still does not guarantee a good performance in terms of execution time because it leverages the computing power of a single computing machine. To cater to this requirement, we developed a parallel version of AUTOFIT, called HS-AUTOFIT, capable of running on high-performance computing (HPC) clusters to shorten the time to explore and analyze spectral big data. In this paper, we report some tests conducted on a real HPC cluster aimed at providing a quantitative assessment of HS-AUTOFIT’s scaling capabilities in a multi-node computing context. The collected results demonstrate the benefits of the proposed approach in terms of a significant reduction in computing time.

Testing the scalability of the HS-AUTOFIT tool in a high-performance computing environment / Di Modica G.; Evangelisti L.; Foschini L.; Maris A.; Melandri S.. - In: ELECTRONICS. - ISSN 2079-9292. - ELETTRONICO. - 10:18(2021), pp. 2251.1-2251.15. [10.3390/electronics10182251]

Testing the scalability of the HS-AUTOFIT tool in a high-performance computing environment

Di Modica G.
Primo
;
Evangelisti L.
Secondo
;
Foschini L.;Maris A.
Penultimo
;
Melandri S.
Ultimo
2021

Abstract

In the last years, the development of broadband chirped-pulse Fourier transform microwave spectrometers has revolutionized the field of rotational spectroscopy. Currently, it is possible to experimentally obtain a large quantity of spectra that would be difficult to analyze manually due to two main reasons. First, recent instruments allow obtaining a considerable amount of data in very short times, and second, it is possible to analyze complex mixtures of molecules that all contribute to the density of the spectra. AUTOFIT is a spectral assignment software application that was developed in 2013 to support and facilitate the analysis. Notwithstanding the benefits AUTOFIT brings in terms of automation of the analysis of the accumulated data, it still does not guarantee a good performance in terms of execution time because it leverages the computing power of a single computing machine. To cater to this requirement, we developed a parallel version of AUTOFIT, called HS-AUTOFIT, capable of running on high-performance computing (HPC) clusters to shorten the time to explore and analyze spectral big data. In this paper, we report some tests conducted on a real HPC cluster aimed at providing a quantitative assessment of HS-AUTOFIT’s scaling capabilities in a multi-node computing context. The collected results demonstrate the benefits of the proposed approach in terms of a significant reduction in computing time.
2021
Testing the scalability of the HS-AUTOFIT tool in a high-performance computing environment / Di Modica G.; Evangelisti L.; Foschini L.; Maris A.; Melandri S.. - In: ELECTRONICS. - ISSN 2079-9292. - ELETTRONICO. - 10:18(2021), pp. 2251.1-2251.15. [10.3390/electronics10182251]
Di Modica G.; Evangelisti L.; Foschini L.; Maris A.; Melandri S.
File in questo prodotto:
File Dimensione Formato  
electronics-10-02251-v2.pdf

accesso aperto

Tipo: Versione (PDF) editoriale
Licenza: Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY)
Dimensione 508.16 kB
Formato Adobe PDF
508.16 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/861976
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? 3
social impact