• Energy Research
• EU close

Datacenters have become an important part of todays computing infrastructure. Recent studies have shown the increasing importance of thermal considerations to achieve effective resource management. In this paper, we study thermal-aware scheduling for homogeneous high-performance computing (HPC) datacenters under a thermal model that captures both spatial and temporal correlations of the temperature evolution. We propose an online scheduling heuristic to minimize the makespan for a set of HPC applications subject to a thermal constraint. The heuristic leverages the novel notion of thermal-aware load to perform both job assignment and thermal management. To respect the temperature constraint, which is governed by a complex spatio-temporal thermal correlation, dynamic voltage and frequency scaling (DVFS) is used to regulate the job executions during runtime while dynamically balancing the loads of the servers to improve makespan. Extensive simulations are conducted based on an experimentally validated datacenter configuration and realistic parameter settings. The results show improved performance of the proposed heuristic compared to existing solutions in the literature, and demonstrate the importance of both spatial and temporal considerations. In contrast to some scheduling problems, where DVFS introduces performanceenergy tradeoffs, our findings reveal the benefit of applying DVFS with both performance and energy gains in the context of spatio-temporal thermal-aware scheduling. Thermal model capturing both spatial and temporal temperature correlations in datacenters.Formulation of a spatio-temporal thermal-aware scheduling problem for HPC applications.Scheduling heuristic using thermal-aware load for job assignment and thermal management.Simulations to show the effectiveness of heuristic under a wide range of parameters.

The energy consumption of a computing system depends not only on its architecture, but also on its usage. This paper describes the Energy Consumption Library libec, a modular library of sensors and power estimators, which do not depend on wattmeter to measure the power dissipated by a machine and/or the applications that it executes, etc. In addition, four use cases are used to demonstrate some of the library's capabilities.

Green IT has recently appeared as a mandatory approach to take into account energy efficiency in information technology. This article investigates the Green IT area and its opportunities for innovation. This chapter analyses the main motivations for Green IT, and proposes a definition of Green IT including social, environmental and economic concerns. Beyond simply listing areas of possible innovation, this paper studies the virtuous circle that appears in Green IT: while Green IT has its own motivations, the resulting research feeds into other research fields. Innovation in this particular sector paves the way for further innovation by means of original research not foreseen initially.

This paper describes the CoolEmAll project and its approach for modeling and simulating energy-efficient and thermal-aware data centers. The aim of the project was to address energy-thermal efficiency of data centers by combining the optimization of IT, cooling and workload management. This paper provides a complete data center model considering the workload profiles, the applications profiling, the power model and a cooling model. Different energy efficiency metrics are proposed and various resource management and scheduling policies are presented. The proposed strategies are validated through simulation at different levels of a data center.

Heterogeneous servers are becoming prevalent in many high-performance computing environments, including clusters and datacenters. In this paper, we consider multi-objective scheduling for heterogeneous server systems to optimize simultaneously the application performance, energy consumption and thermal imbalance. First, a greedy online framework is presented to allow the scheduling decisions to be made based on any well-defined cost function. To tackle the possibly conflicting objectives, we propose a fuzzy-based priority approach for exploring the tradeoffs of two or more objectives at the same time. Moreover, we present a heuristic algorithm for the static placement of physical machines in order to reduce the maximum temperature at the server outlets. Extensive simulations based on an emerging class of high-density server system have demonstrated the effectiveness of our proposed approach and heuristics in optimizing multiple objectives while achieving better thermal balance.