An Empirical Method for Processing I/O Traces to Analyze the Performance of DL Applications

Edixon Parraga*, Betzabeth Leon, Sandra Mendez, Dolores Rexachs, Remo Suppi, Emilio Luque

*Autor corresponent d’aquest treball

Producció científica: Capítol de llibreCapítolRecercaAvaluat per experts

1 Citació (Scopus)
1 Descàrregues (Pure)

Resum

The exponential growth of data handled by Deep Learning (DL) applications has led to an unprecedented demand for computational resources, necessitating their execution on High Performance Computing (HPC) systems. However, understanding and optimizing Input/Output (I/O) of the DL applications can be challenging due to the complexity and scale of DL workloads and the heterogeneous nature of I/O operations. This paper addresses this issue by proposing an I/O traces processing method that simplifies the generation of reports on global I/O patterns and performance to aid in I/O performance analysis. Our approach focuses on understanding the temporal and spatial distributions of I/O operations and related with the behavior at I/O system level. The proposed method enables us to synthesize and extract key information from the reports generated by tools such as Darshan tool and the seff command. These reports offer a detailed view of I/O performance, providing a set of metrics that deepen our understanding of the I/O behavior of DL applications.
Títol traduït de la contribucióUn método empírico para procesar trazas de E/S para analizar el rendimiento de aplicaciones DL
Idioma originalAnglès
Títol de la publicacióCloud Computing, Big Data and Emerging Topics - 12th Conference, JCC-BD and ET 2024, Revised Selected Papers
EditorsMarcelo Naiouf, Laura De Giusti, Franco Chichizola, Leandro Libutti
EditorSpringer Science and Business Media Deutschland GmbH
Pàgines74-90
Nombre de pàgines17
ISBN (electrònic)978-3-031-70807-7
ISBN (imprès)978-3-031-70807-7, 9783031708060
DOIs
Estat de la publicacióPublicada - 11 d’oct. 2024

Sèrie de publicacions

NomCommunications in Computer and Information Science
Volum2189 CCIS

Fingerprint

Navegar pels temes de recerca de 'An Empirical Method for Processing I/O Traces to Analyze the Performance of DL Applications'. Junts formen un fingerprint únic.

Com citar-ho