SaaS for Energy Efficient Utilization of HPC Resources of Linear Algebra Calculations

Hrachya Astsatryan (1) , Wahi Narsisian (1) , Georges da Costa (2, 3)

1 Institute for Informatics and Automation Problems
2 IRIT-SEPIA - Système d’exploitation, systèmes répartis, de l’intergiciel à l’architecture
3 UT3 - Université Toulouse III - Paul Sabatier

The most important factor of High performance computing (HPC) systems nowadays is to limit or decrease the power consumption while preserving a high utilization. And with the availability of alternative energy, which powers such systems, there is a need to maximize the usage of alternative energy over brown power. For now, the usage of alternative energy is varying in time due to different factors such as sunny days, the wind, etc. and it is crucial to have an energy-aware algorithm to maximize the usage of this energy. In this paper a SaaS service is presented to optimize a usage of alternative energy, to reduce the power consumption and to preserve a best possible percentage of resource utilization.

Mots clés: Linear algebra; HPC; Energy efficient utilization; DVFS; Cloud service

Domaines: Réseaux et télécommunications [cs.NI]; Système d'exploitation [cs.OS]; Architectures Matérielles [cs.AR]; Systèmes embarqués

Hrachya Astsatryan, Wahi Narsisian, Georges da Costa. SaaS for Energy Efficient Utilization of HPC Resources of Linear Algebra Calculations. Scalable Computing : Practice and Experience, 2017, 18 (2), pp.145-150. ⟨10.12694/scpe.v18i2.1286⟩. ⟨hal-01887823⟩