Deliverable D5.2 – Performance and power consumption evaluation of compute node - LIRMM - Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier Accéder directement au contenu
Rapport (Rapport De Recherche) Année : 2020

Deliverable D5.2 – Performance and power consumption evaluation of compute node

Philippe Naudin
  • Fonction : Auteur
  • PersonId : 1045646
Guillaume Devic
Abdoulaye Gamatié
Gilles Sassatelli

Résumé

The CONTINUUM project aims to define a new energy-efficient compute node model, which will benefit from a suitable combination of efficient compilation techniques, emerging memory, and communication technologies together with heterogeneous cores. This deliverable deals with the comprehensive evaluation of the architecture prototypes defined in the project [2]. Various heterogeneous multicore designs are developed and prototyped on FPGA. These designs rely on cost-effective and inherently ultra-low-power cores commercialized by Cortus SA. Microarchitecture-level design considerations, e.g. floating-point and out-of-order computing capabilities, are taken into account for exploring candidate solutions. In addition, a tailored and flexible multi-task programming model is defined for the proposed architecture paradigm. We analyze the behavior of various application programs on available core configurations. This provides valuable insights on the best architecture setups that match program characteristics, so as to enable increased energy-efficiency.
Fichier principal
Vignette du fichier
D5.2.pdf (2.97 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

lirmm-03168364 , version 1 (12-03-2021)

Identifiants

  • HAL Id : lirmm-03168364 , version 1

Citer

Stefano Bernabovi, Michael Chapman, Philippe Naudin, Guillaume Devic, Abdoulaye Gamatié, et al.. Deliverable D5.2 – Performance and power consumption evaluation of compute node. [Research Report] Cortus S.A.S; LIRMM (UM, CNRS). 2020. ⟨lirmm-03168364⟩
52 Consultations
63 Téléchargements

Partager

Gmail Facebook X LinkedIn More