MAGPIE: System-level Evaluation of Manycore Systems with Emerging Memory Technologies

Thibaud Delobelle; Pierre-Yves Péneau; Abdoulaye Gamatié; Florent Bruguier; Sophiane Senni; Gilles Sassatelli; Lionel Torres

MAGPIE: System-level Evaluation of Manycore Systems with Emerging Memory Technologies

Acte de colloque

Open Access

MAGPIE: System-level Evaluation of Manycore Systems with Emerging Memory Technologies

Thibaud Delobelle, Pierre-Yves Péneau, Abdoulaye Gamatié, Florent Bruguier, Sophiane Senni, Gilles Sassatelli et Lionel Torres

2nd International Workshop on Emerging Memory Solutions co-located with DATE

EMS: Emerging Memory Solutions (Lausanne, Switzerland, 31/03/2017)

31/03/2017

Résumé

Energy efficiency

Non volatile memory

Design space exploration

Simulation flow

We present a holistic design evaluation tool, named MAGPIE explorer, which aims at manycore systems integrating emerging NVM technologies for energy-efficiency. MAGPIE stands for Manycore Architecture enerGy and Performance evaluatIon Environment. It is built upon three mature and popular tools: the gem5 full-system simulator, the McPAT and NVSim power/energy and area estimation tools for CMOS, SOI and NVM technologies. MAGPIE explorer enables a designer to specify input design specifications and run a seamless evaluation flow that automatically produces results including application outputs, performance numbers, area, power and energy consumption.

Fichiers et liens (1)

url

Find in HALAfficher

Indicateurs

1 Consultations de la notice

Détails

Titre: MAGPIE: System-level Evaluation of Manycore Systems with Emerging Memory Technologies
Créateurs - sans rôle: Thibaud Delobelle - ADAptive Computing
Pierre-Yves Péneau - ADAptive Computing
Abdoulaye Gamatié - ADAptive Computing
Florent Bruguier - ADAptive Computing
Sophiane Senni - ADAptive Computing
Gilles Sassatelli - ADAptive Computing
Lionel Torres - ADAptive Computing
Détails de publication: 2nd International Workshop on Emerging Memory Solutions co-located with DATE
Colloque: EMS: Emerging Memory Solutions (Lausanne, Switzerland, 31/03/2017)
Identifiants: 9937093209311
Unité académique: Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier - LIRMM
Langue: English
Type de ressource: Conference proceeding
Champs locaux: lirmm-01467328