Power Management Policies Research Articles

Applying Pay-Burst-Only-Once Principle for Periodic Power Management in Hard Real-Time Pipelined Multiprocessor Systems

Pipelined computing is a promising paradigm for embedded system design. Designing a power management policy to reduce the power consumption of a pipelined system with nondeterministic workload is, however, nontrivial. In this article, we study the problem of energy minimization for coarse-grained pipelined systems under hard real-time constraints and propose new approaches based on an inverse use of the pay-burst-only-once principle. We formulate the problem by means of the resource demands of individual pipeline stages and propose two new approaches, a quadratic programming-based approach and fast heuristic, to solve the problem. In the quadratic programming approach, the problem is transformed into a standard quadratic programming with box constraint and then solved by a standard quadratic programming solver. Observing the problem is NP-hard, the fast heuristic is designed to solve the problem more efficiently. Our approach is scalable with respect to the numbers of pipeline stages. Simulation results using real-life applications are presented to demonstrate the effectiveness of our methods.

Lifetime Bounds of Wi-Fi Enabled Sensor Nodes

Wireless sensor and actuator networks are often deployed in areas that already have a wireless infrastructure. Coexisting in this infrastructure is a challenge, as the used communication techniques interfere. Instead of fighting to fit in, one could try to tap into the existing infrastructure. This solves the coexistence problem, does not require new infrastructure and makes the integration of systems trivial. To determine the feasibility of such a solution, our work focuses on analysing energy consumption and power management policies of Wi-Fi enabled nodes based on recent hardware. Our analysis shows that for some tasks, Wi-Fi nodes are becoming a viable option and it may not be necessary to retain the current practice of deploying and maintaining specialised wireless sensor and actuator networks.

Message Passing-Aware Power Management on Many-Core Systems

Dynamic frequency and voltage scaling (DVFS) techniques have been widely used for meeting energy constraints. Single-chip many-core systems bring new challenges owing to the large number of operating points and the shift to message passing from shared memory communication. DVFS, however, has been mostly studied on single-chip systems with one or few cores, without considering the impact of the communication among cores. This paper evaluates the impact of voltage and frequency scaling on the performance and power of many-core systems with message passing (MP) based communication, and proposes a power management policy that leverages the communication pattern information to efficiently traverse the search space for finding the optimal voltage and frequency operating point. We conduct experiments on a 48-core Intel Single-Chip Cloud Computer (SCC), as our target many-core platform. The paper first introduces the runtime monitoring infrastructure and the application suite we have designed for an in-depth evaluation of the SCC. We then quantify the effects of frequency perturbations on performance and energy efficiency. Experimental results show that runtime communication patterns lead to significant differences in power/performance tradeoffs in many-core systems with MP-based communication. We show that the proposed power management policy achieves up to the 70% energy-delayproduct (EDP) improvements compared to existing DVFS policies, while meeting the performance constraints.

SEATS: smart energy-aware task scheduling in real-time cloud computing

Mitigating energy consumption in Clouds reduces operational costs for providers. Power management policies which aim to reduce total energy consumed in data-centers pose challenges in both hardware technologies and resource management policies. We introduce an optimal utilization level of a host to execute a certain number of instructions to minimize energy consumption of the host. We also propose a virtual machine (VM) scheduling algorithm based on the unsurpassed utilization level to come up with optimal energy consumption while meeting a given QoS. In other words, our proposed algorithm regulates allocated computing resources of VMs on a host which results in reaching an optimal energy level in the host. The simulation results show that our proposed method not only reduces total energy consumption of a Cloud by 60 %, but also has a profound impact on turnaround times of real-time tasks by 94 %. It also increases the acceptance rate of arrival tasks by 96 %. Moreover, it takes a leading part in accepting lengthy tasks which have short deadlines.

Experimental Demonstration of a Tiered Power Management System for Economic Operation of Grid-Tied Microgrids

Power management of grid-tied microgrids including distributed generations (DGs) and storage devices can be challenging due to the intermittent and uncontrollable nature of many types of DGs, load variations in time, different grid power tariffs, availability of different options to balance the electric supply and demand, and other parameters. In order to operate a microgrid efficiently, the management system should accomplish two tasks. 1) It needs to be adaptive and optimize the microgrid's performance by defining long-term (daily-based) directives or control strategies. 2) The management system should be able to operate and control the microgrid in real time and satisfy all operational constraints. To address the above-mentioned tasks, a comprehensive power management system that includes two control layers is developed in this paper. Concept for the proposed power management policy was demonstrated on an experimental microgrid system composed of lead-acid batteries, photovoltaic (PV) system, 3-kW peak load, and a utility connection.

Temperature-aware adaptive power management: An analytical approach for joint processor and cooling mode selection

This paper presents a temperature-aware adaptive power management policy for a Markovian single-processor system equipped with a cooling fan working under variable ambient temperature. The method tries to minimize the jitter of the system performance, subject to a given power consumption constraint. The adaptation is performed through simultaneous dynamic processor speed selection (to control power of the system processor) and dynamic fan mode selection (to control power of its cooling fan). First, we provide an analytical method for calculation of the system power consumption and performance under some processor and fan speed setups. All the inter-effects among the processor dynamic power, leakage power, temperature, and the system stochastic properties are taken into account in the calculations to reach more accurate analysis. Then, to adapt the system power consumption with respect to the variations of the ambient temperature, we have used a model predictive control (MPC) approach beside some optimizations based on the analysis to appropriately change the processor and fan speed setups. Simulation results show the efficacy of the proposed adaptive method with an emphasis on the preference of the considered backlog-dependent mode selection comparing to some traditional static mode selection methods.

Real-Time Power Sensors for Intelligent Power Management and Beyond

This article presents a low-overhead on-chip power sensor design that enables cross-layer power management policy used in future computing systems.

SleepScale

Power consumption in data centers has been growing significantly in recent years. To reduce power, servers are beingequipped with increasingly sophisticated power management mechanisms. Different mechanisms offer dramatically different trade-offs between power savings and performance penalties. Considering the complexity, variety, and temporallyvarying nature of the applications hosted in a typical data center, intelligently determining which power management policy to use and when is a complicated task. In this paper we analyze a system model featuring both performance scaling and low-power states. We reveal the interplay between performance scaling and low-power states via intensive simulation and analytic verification. Based on the observations, we present SleepScale, a runtime power management tool designed to efficiently exploit existing power control mechanisms. At run time, SleepScale characterizes power consumption and quality-of-service (QoS) for each lowpower state and frequency setting, and selects the best policy for a given QoS constraint. We evaluate SleepScale using workload traces from data centers and achieve significant power savings relative to conventional power management strategies

Energy adaptation in multi-core software routers

We consider a new generation of Commercial-Off-The-Shelf (COTS) Software Routers (SRs), able to effectively exploit multi-core/CPU hardware platforms. This choice was mainly driven from the fact that SRs are one of the few platforms including power management features, and that can be easily customized. Our main objective is to evaluate and to model the impact of power saving mechanisms, generally included in today’s COTS processors, on the SR networking performance and behavior. To this purpose, we separately characterize the roles of both hardware and software layers through a large set of internal and external experimental measurements, obtained with a heterogeneous set of hardware platforms and SR setups. Starting from this detailed measure analysis, we propose a simple model, able to represent the SR performance with a high accuracy level in terms of packet throughput and related power consumption. The proposed model is thought to estimate how SRs behave under different network conditions and power management configurations. In fact, the model has been effectively applied inside a “green optimization” policy, which aims at minimizing power consumption, while maintaining a certain SR performance target. Finally, in order to provide an experimental evaluation of the proposed ideas, we apply such power management policies to a new generation software router platform, and we test their performance under real traffic traces.

Online learning of timeout policies for dynamic power management

Dynamic power management (DPM) refers to strategies which selectively change the operational states of a device during runtime to reduce the power consumption based on the past usage pattern, the current workload, and the given performance constraint. The power management problem becomes more challenging when the workload exhibits nonstationary behavior which may degrade the performance of any single or static DPM policy. This article presents a reinforcement learning (RL)-based DPM technique for optimal selection of timeout values in the different device states. Each timeout period determines how long the device will remain in a particular state before the transition decision is taken. The timeout selection is based on workload estimates derived from a Multilayer Artificial Neural Network (ML-ANN) and an objective function given by weighted performance and power parameters. Our DPM approach is further able to adapt the power-performance weights online to meet user-specified power and performance constraints, respectively. We have completely implemented our DPM algorithm on our embedded traffic surveillance platform and performed long-term experiments using real traffic data to demonstrate the effectiveness of the DPM. Our results show that the proposed learning algorithm not only adequately explores the power-performance trade-off with nonstationary workload but can also successfully perform online adjustment of the trade-off parameter in order to meet the user-specified constraint.

A Simple Model for Estimating Power Consumption of a Multicore Server System

Balancing the performance and the energy consumption of the servers is one of the important issues in large-scale computing infrastructure such as data centers. Measuring or accurately estimating power consumption of a server is one of the most fundamental and enabling technologies for enhancing energy efficiency of a server because how the server consumes the supplied power is essential for constructing a power management policy. For the purpose, power models for server systems have been extensively studied. However, most of existing works are too complex to be used real-time, because gathering the data for estimating the power consumption causes much overhead. In this paper, we propose a simple power model for a multicore server. Our model is simple enough to gather only four parameters: operating frequency, the number of active cores, the number of cache accesses and the number of the last level cache misses. We show our model is simple but relatively accurate by experiments that show the model has over 90% accuracy.

Stochastic modeling of dynamic power management policies in server farms with setup times and server failures

SUMMARYServer farms generally consume an enormous amount of energy, which not only increases the running cost but also simultaneously enhances their greenhouse gas emissions. One way to improve the energy efficiency in server farms is dynamical powering on/off servers. However, this method suffers from many setup time to turn the servers back on, which would have a negative impact on a job's response time and waste yet additional energy. This situation is further exacerbated by the unreliability of the servers, which has become the norm in today's server farms. In this paper, we investigate the impact of dynamical powering on/off servers on energy and performance in a typical server farm environment. Prior work has analyzed similar models for a single server, but no analytical results are known for multiservers. We mainly use the matrix geometric method to analyze this model, and system performance measures are explicitly developed in terms of computable forms. An energy‐performance trade‐off model is derived to determine the optimal management policy for the server farms. Finally, we discuss some extensions of our model to show its robustness and to point out avenues for future research. Numerical examples are provided at several points throughout the paper to illustrate the correctness of our analysis results and to validate the optimization approach. Copyright © 2014 John Wiley & Sons, Ltd.

Equi-Noise: A Statistical Model That Combines Embedded Memory Failures and Channel Noise

This paper exploits the predominance of embedded memories in current and emerging wireless transceivers as a means to save power via channel state aware voltage scaling. The paper presents a statistical model that captures errors in embedded memories due to voltage over-scaling and maps the errors to a Gaussian distribution that represents a combination of communication channel noise and hardware noise. Designers can use the proposed model to investigate different power management policies, that capture the performance of the system as a function of both channel and hardware dynamics, thus creating a much richer design space of power, performance and reliability. A case study of a DVB receiver is presented and the validity of the proposed model is confirmed by simulations.

Cooperative autonomous control for active power sharing in multi-terminal VSC-HVDC

Economical autonomous control for sharing active power among MTDC systems based on the availability of active power or on a power management policy is proposed in this paper. Power sharing among MTDC systems involves a priority or sequential procedural problem because of the use of the conventional droop strategy. On the other hand, the use of predefined or constant power sharing fails to identify the available power that can be shared when it is not consumed by another inverter. The proposed strategy addresses these issues through a variety of options. The test system presented in this paper consists of three simulated VSC terminals based on a detailed switching VSC model with two AC voltage levels. The MTDC system has been simulated in a PSCAD/EMTDC environment. The simulation results show a significant decrease in operating costs as well as the ability to provide effective protection with respect to the problem of overloading.

Energy saving in data processing and communication systems.

The power management of ICT systems, that is, data processing (Dp) and telecommunication (Tlc) systems, is becoming a relevant problem in economical terms. Dp systems totalize millions of servers and associated subsystems (processors, monitors, storage devices, etc.) all over the world that need to be electrically powered. Dp systems are also used in the government of Tlc systems, which, besides requiring Dp electrical power, also require Tlc-specific power, both for mobile networks (with their cell-phone towers and associated subsystems: base stations, subscriber stations, switching nodes, etc.) and for wired networks (with their routers, gateways, switches, etc.). ICT research is thus expected to investigate into methods to reduce Dp- and Tlc-specific power consumption. However, saving power may turn into waste of performance, in other words, into waste of ICT quality of service (QoS). This paper investigates the Dp and Tlc power management policies that look at compromises between power saving and QoS.

Power Management of Server Farms

The power management of server farms (Sf) is becoming a relevant problem in economical terms. Server farms totalize millions of servers all over the world that need to be electrically powered. Research is thus expected to investigate into methods to reduce Sf power consumption. However, saving power may turn into waste of performance (high response_times), in other words, into waste of Sf Quality of Service (QoS). By use of a Sfmodel, this paper investigates Sf power management strategies that look at compromises between power-saving and QoS. Various optimizing Sf power management policies are studied combined with the effects of job queueing disciplines. The (policy, discipline) pairs, or strategies, that optimize the Sf power consumption (minimum absorbed Watts), the Sf performance (minimum response_time), and the Sf performance-per-Watt (minimum response_time-per-Watt) are identified.By use of the model, the work the server-manager has to do to direct hisSf is greatly simplified, since the universe of all possible (strategies he needs to choose from is drastically reduced to a very small set of most significant strategies.

Autonomic Power Aware Cloud Resources Orchestration Architecture for Web Applications

Endless resources provisioning illusion is the mainstay for cloud computing paradigm. However, the unpredictable volatility nature involving web applications workload demand would highly hinder cloud computing platforms performance, furthermore, expose cloud resources for possible devastation. Accordingly, this work proposes autonomic power aware SLA-oriented cloud resources orchestration two-tier architecture. Despite complexity and uncertainties of the workload fluctuations, the proposed architecture geared for leveraging cloud system resources utilization, ensuring explicit guarantees on web applications’ responsiveness obligations, meanwhile achieving power consumption minimization objectives. The proposed architecture consolidates heuristic methodologies along with control theory approaches in a resource orchestration hierarchical structure. Firstly, an autonomic global controller is presented. The proposed global controller exploits heuristic methodology for mapping virtual machines (VMs) to the appreciate cloud resources in accordance to heuristic multidimensional objectives based placement strategy. Secondly, a proactive fuzzy-logic based local controller is proposed. The proposed local controller aimed at in confronting workloads' sustainable fluctuations via proactive amendment for the placement and provisioning schedules. Furthermore, the proposed local controller oriented towards maintaining active power management policy especially during transient peak of usage, thereby mitigating overall costs, and extending resources capacity and performance capabilities. Simulation results and comparisons demonstrate that the proposed architecture significantly surpasses previous approaches in terms of total energy consumption, furthermore maintaining web applications SLAs objectives despite dynamic workload scenarios.

POWER-TRUCTOR: An Integrated Tool Flow for Formal Verification and Coverage of Architectural Power Intent

With the growing complexity and gradually shrinking power requirements in the system-on-chip designs, sophisticated global power management policies (which orchestrate the switching between power states of multiple power domains) are commonplace. Recent research has paved some novel ways to verify the sophisticated on-chip architectural power management decisions and analyze the verification coverage. However, one of the primary challenges in verifying such power management architectures stems from the mixed implementation of such strategies, where the local power controllers are in hardware and the global power management is implemented in software/firmware. There has been lack of effort to build a unified and automated framework for power intent verification and coverage analysis for generic power management logics. This paper tries to develop an end-to-end automated framework enabled by a tool named POWER-TRUCTOR for power intent validation.

Evaluating reliability of WSN with sleep/wake-up interfering nodes

A wireless sensor network (WSN) (singular and plural of acronyms are spelled the same) is a distributed system composed of autonomous sensor nodes wireless connected and randomly scattered into a geographical area to cooperatively monitor physical or environmental conditions. Adequate techniques and strategies are required to manage a WSN so that it works properly, observing specific quantities and metrics to evaluate the WSN operational conditions. Among them, one of the most important is the reliability. Considering a WSN as a system composed of sensor nodes the system reliability approach can be applied, thus expressing the WSN reliability in terms of its nodes' reliability. More specifically, since often standby power management policies are applied at node level and interferences among nodes may arise, a WSN can be considered as a dynamic system. In this article we therefore consider the WSN reliability evaluation problem from the dynamic system reliability perspective. Static–structural interactions are specified by the WSN topology. Sleep/wake-up standby policies and interferences due to wireless communications can be instead considered as dynamic aspects. Thus, in order to represent and to evaluate the WSN reliability, we use dynamic reliability block diagrams and Petri nets. The proposed technique allows to overcome the limits of Markov models when considering non-linear discharge processes, since they cannot adequately represent the aging processes. In order to demonstrate the effectiveness of the technique, we investigate some specific WSN network topologies, providing guidelines for their representation and evaluation.

Adaptive workload driven dynamic power management for high performance computing clusters

With the scale expansion of high performance computer systems, efficient power management has developed into an important issue. To strive to balance power consumption and performance, this paper proposes an adaptive workload-driven dynamic power management policy for homogeneous clusters, which dynamically adjusts the power mode of computing nodes according to workload variation. The proposed policy combines the pre-wakeup method and the feedback mechanism to reduce performance degradation due to the wakeup delay. The experimental results demonstrate that, as compared with two existing timeout policies, adaptive workload-driven dynamic power management effectively reduced the performance loss with a slight increase in power consumption.