New Algorithms for Balancing Energy Consumption and Performance in Computational Clusters

keywords: Heterogeneous resources, computational cluster, common queue model, real-time load
In this paper, we propose new real-time measurement-based scheduling algorithms to achieve a trade-off between the energy efficiency and the performance capability of computational clusters. An investigation is performed using a specific scenario of computing clusters with realistic parameters. Numerical results show that a trade-off between the performance and the energy efficiency can be controlled by the proposed algorithms.
mathematics subject classification 2000: 37M05, 97R30
reference: Vol. 36, 2017, No. 2, pp. 307–330