Mostra el registre d'ítem simple

dc.contributor.authorParés Pont, Ferran
dc.contributor.authorMegias Montsesinos, Pedro
dc.contributor.authorGarcía Gasulla, Dario
dc.contributor.authorGarcia Gasulla, Marta
dc.contributor.authorAyguadé Parra, Eduard
dc.contributor.authorLabarta Mancho, Jesús José
dc.contributor.otherUniversitat Politècnica de Catalunya. Doctorat en Intel·ligència Artificial
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
dc.contributor.otherBarcelona Supercomputing Center
dc.date.accessioned2021-07-14T11:12:41Z
dc.date.available2021-07-14T11:12:41Z
dc.date.issued2021-03
dc.identifier.citationParés, F. [et al.]. Size & shape matters: The need of HPC benchmarks of high resolution image training for deep learning. "Supercomputing frontiers and innovations", Març 2021, vol. 8, núm. 1, p. 28-44.
dc.identifier.issn2313-8734
dc.identifier.urihttp://hdl.handle.net/2117/349274
dc.description.abstractOne of the purposes of HPC benchmarks is to identify limitations and bottlenecks in hardware. This functionality is particularly influential when assessing performance on emerging tasks, the nature and requirements of which may not yet be fully understood. In this setting, a proper benchmark can steer the design of next generation hardware by properly identifying said requirements, and quicken the deployment of novel solutions. With the increasing popularity of deep learning workloads, benchmarks for this family of tasks have been gaining popularity. Particularly for image based tasks, which rely on the most well established family of deep learning models: Convolutional Neural Networks. Significantly, most benchmarks for CNN use low-resolution and fixed-shape (LR&FS) images. While this sort of inputs have been very successful for certain purposes, they are insufficient for some domains of special interest (e.g., medical image diagnosis or autonomous driving) where one requires higher resolutions and variable-shape (HR&VS) images to avoid loss of information and deformation. As of today, it is still unclear how does image resolution and shape variability affect the nature of the problem from a computational perspective. In this paper we assess the differences between training with LR&FS and HR&VS, as means to justify the importance of building benchmarks specific for the latter. Our results on three different HPC clusters show significant variations in time, resources and memory management, highlighting the differences between LR&FS and HR&VS image deep learning.
dc.format.extent17 p.
dc.language.isoeng
dc.rightsAttribution-NonCommercial 3.0 Spain
dc.rights.urihttp://creativecommons.org/licenses/by-nc/3.0/es/
dc.subjectÀrees temàtiques de la UPC::Informàtica::Arquitectura de computadors
dc.subjectÀrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
dc.subject.lcshMachine learning
dc.subject.lcshSupercomputers
dc.subject.lcshNeural networks (Computer science)
dc.subject.otherDeep learning
dc.subject.otherConvolutional neural networks
dc.subject.otherHigh-resolution images
dc.subject.otherVariable shape images
dc.subject.otherHPC benchmarks
dc.titleSize & shape matters: The need of HPC benchmarks of high resolution image training for deep learning
dc.typeArticle
dc.subject.lemacAprenentatge automàtic
dc.subject.lemacSupercomputadors
dc.subject.lemacXarxes neuronals (Informàtica)
dc.contributor.groupUniversitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
dc.identifier.doi10.14529/jsfi210103
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttps://superfri.org/superfri/article/view/365/383
dc.rights.accessOpen Access
local.identifier.drac31870050
dc.description.versionPostprint (published version)
local.citation.authorParés, F.; Megias, P.; García, D.; Garcia, M.; Ayguadé, E.; Labarta, J.
local.citation.publicationNameSupercomputing frontiers and innovations
local.citation.volume8
local.citation.number1
local.citation.startingPage28
local.citation.endingPage44


Fitxers d'aquest items

Thumbnail

Aquest ítem apareix a les col·leccions següents

Mostra el registre d'ítem simple