Predictive Thermal Management for Chip Multiprocessors Using Co-designed Virtual Machines

Abstract : The sustained push for performance, transistor count, and instruction level parallelism has reached a point where chip level power density issues are at the forefront of design constraints. Many high performance computing platforms are integrating several homogeneous or heterogeneous processing cores on the same die to fit small form factors. Due to design limitations of using expensive cooling solutions, complex chip multiprocessors require an architectural solution to mitigate thermal problems. Many of the proposed systems deploy DVFS to address thermal emergencies, either within an operating system or hardware. These techniques have certain limitations in terms of response lag, scalability, cost or being reactive. In this paper, we present an alternative thermal management system to address these limitations, based on co-designed virtual machines concept. The proposed scheme delivers localized and preemptive response to thermal events, adapts well to multi-core and multi-threading environment, while delivering maximum performance under thermal stress.
Type de document :
Communication dans un congrès
André Seznec and Joel Emer and Mike O'Boyle and Margaret Martonosi and Theo Ungerer. HiPEAC 2009 - High Performance and Embedded Architectures and Compilers, Jan 2009, Paphos, Cyprus. Springer, 2009, 〈10.1007/978-3-540-92990-1_22〉
Liste complète des métadonnées

https://hal.inria.fr/inria-00446303
Contributeur : Ist Rennes <>
Soumis le : mardi 12 janvier 2010 - 14:23:03
Dernière modification le : lundi 20 juin 2016 - 14:10:32

Identifiants

Collections

Citation

Omer Khan, Sandip Kundu. Predictive Thermal Management for Chip Multiprocessors Using Co-designed Virtual Machines. André Seznec and Joel Emer and Mike O'Boyle and Margaret Martonosi and Theo Ungerer. HiPEAC 2009 - High Performance and Embedded Architectures and Compilers, Jan 2009, Paphos, Cyprus. Springer, 2009, 〈10.1007/978-3-540-92990-1_22〉. 〈inria-00446303〉

Partager

Métriques

Consultations de la notice

27