Investigating power capping toward energy‐efficient scientific applications

Large-scale Internet services require a computing infrastructure that can beappropriately described as a warehouse-sized computing system. The cost ofbuilding datacenter facilities capable of delivering a given power capacity tosuch a computer can rival the recurring energy consumption costs themselves.Therefore, there are strong economic incentives to operate facilities as closeas possible to maximum capacity, so that the non-recurring facility costs canbe best amortized. That is difficult to achieve in practice because ofuncertainties in equipment power ratings and because power consumption tends tovary significantly with the actual computing activity. Effective powerprovisioning strategies are needed to determine how much computing equipmentcan be safely and efficiently hosted within a given power budget.In this paper we present the aggregate power usage characteristics of largecollections of servers (up to 15 thousand) for different classes ofapplications over a period of approximately six months. Those observationsallow us to evaluate opportunities for maximizing the use of the deployed powercapacity of datacenters, and assess the risks of over-subscribing it. We findthat even in well-tuned applications there is a noticeable gap (7 - 16%)between achieved and theoretical aggregate peak power usage at the clusterlevel (thousands of servers). The gap grows to almost 40% in wholedatacenters. This headroom can be used to deploy additional compute equipmentwithin the same power budget with minimal risk of exceeding it. We use ourmodeling framework to estimate the potential of power management schemes toreduce peak power and energy usage. We find that the opportunities for powerand energy savings are significant, but greater at the cluster-level (thousandsof servers) than at the rack-level (tens). Finally we argue that systems needto be power efficient across the activity range, and not only at peakperformance levels.

The power management capabilities of hybrid energy storage systems offer several advantages for EVs, including improved performance, longer battery life, and reduced cost. HES technologies which combine Li-ion cells with ultra-capacitors (UC), are developing as just a possible alternative to the constraints of conventional ESS for conserving electricity. Lithium-ion batteries have a limited cycle life, meaning that they gradually degrade with each charge-discharge cycle. This can reduce the overall lifetime of the ESS and require more frequent replacements, adding to the cost and environmental impact of EVs. To circumvent those constraints, mixed battery packs combining battery technology as well as superconductors are being developed. To improve the design and control of hybrid energy storage systems for electric vehicles (EVs), a bi-level multi-objective framework has been proposed. A Bi-level multi-objective design and control framework, incorporating NSGA-II and fuzzy logic control, to obtain an optimal sized hybrid energy storage system and corresponding power management strategy. This dynamic and multi-objective architecture may achieve optimum solution hybrids battery bank as well as a matching significant power scheme one at a moment through integrating non-dominated sorting genetic and fuzzy logic control methods. To evaluate the effectiveness of the proposed bi-level multi-objective framework, Pareto optimal solutions of different hybrid energy storage systems are obtained. This approach can lead to significant improvements in the performance and efficiency of hybrid energy storage systems, making them more suitable for EVs and other applications that require high-performance energy storage solutions. Index Terms—Batteries, ultra capacitors, recursive imprecise input, non – linear and non optimisation, and e – mobility

The University of Tennessee (UTK) and University of Texas at El Paso (UTEP) partnership supported the three main thrusts of the SUPER project---performance, energy, and resilience. The UTK-UTEP effort thus helped advance the main goal of SUPER, which was to ensure that DOE's computational scientists can successfully exploit the emerging generation of high performance computing (HPC) systems. This goal is being met by providing application scientists with strategies and tools to productively maximize performance, conserve energy, and attain resilience. The primary vehicle through which UTK provided performance measurement support to SUPER and the larger HPC community is the Performance Application Programming Interface (PAPI). PAPI is an ongoing project that provides a consistent interface and methodology for collecting hardware performance information from various hardware and software components, including most major CPUs, GPUs and accelerators, interconnects, I/O systems, and power interfaces, as well as virtual cloud environments. The PAPI software is widely used for performance modeling of scientific and engineering applications---for example, the HOMME (High Order Methods Modeling Environment) climate code, and the GAMESS and NWChem computational chemistry codes---on DOE supercomputers. PAPI is widely deployed as middleware for use by higher-level profiling, tracing, and sampling tools (e.g., CrayPat, HPCToolkit, Scalasca, Score-P, TAU, Vampir, PerfExpert), making it the de facto standard for hardware counter analysis. PAPI has established itself as fundamental software infrastructure in every application domain (spanning academia, government, and industry), where improving performance can be mission critical. Ultimately, as more application scientists migrate their applications to HPC platforms, they will benefit from the extended capabilities this grant brought to PAPI to analyze and optimize performance in these environments, whether they use PAPI directly, or via third-party performance tools. Capabilities added to PAPI through this grant include support for new architectures such as the lastest GPU and Xeon Phi accelerators, and advanced power measurement and management features. Another important topic for the UTK team was providing support for a rich ecosystem of different fault management strategies in the context of parallel computing. Our long term efforts have been oriented toward proposing flexible strategies and providing building boxes that application developers can use to build the most efficient fault management technique for their application. These efforts span across the entire software spectrum, from theoretical models of existing strategies to easily assess their performance, to algorithmic modifications to take advantage of specific mathematical properties for data redundancy and to extensions to widely used programming paradigms to empower the application developers to deal with all types of faults. We have also continued our tight collaborations with users to help them adopt these technologies to ensure their application always deliver meaningful scientific data. Large supercomputer systems are becoming more and more power and energy constrained, and future systems and applications running on them will need to be optimized to run under power caps and/or minimize energy consumption. The UTEP team contributed to the SUPER energy thrust by developing power modeling methodologies and investigating power management strategies. Scalability modeling results showed that some applications can scale better with respect to an increasing power budget than with respect to only the number of processors. Power management, in particular shifting power to processors on the critical path of an application execution, can reduce perturbation due to system noise and other sources of runtime variability, which are growing problems on large-scale power-constrained computer systems.

In conventional EV powertrain, Battery is the sole energy storage. This makes higher transient current stress on battery during operation, which shall reduce the life of the battery considerably. The size of the battery may required to be over rated to take care heavy transient demands with respect to its C rating. The regenerative braking capability of Electric power train is also gets limited due to the charging transient limitations in Battery. This eventually results in poor EV power train designs and vehicle performance. A hybrid energy storage system with Battery as primary energy source and Ultra capacitor as transient power manager is proposed as a viable solution. The combination shall provide better utilization of regenerative power, thus increase overall efficiency and range. The secondary storage can absorb all transients during regeneration and supply transients when required, during motoring. This can reduce the transient stress on battery increase the life. Cost effective Power train solution is also a major challenge in practical EV system design, and the proposed architecture is without dedicated DCDC converter for battery. The control logic can be formulated so as to meet the transient supplies of the load by ultra-capacitor and indirectly reduce that from battery. This provides low capacity DC-DC converter design, which could be appropriately overloaded for transient demands, thus reducing the overall size also. This can reduce the overall cost of EV power train. This paper discusses Hybrid Energy Storage System configuration for Battery Electric Vehicles and its Power Management strategy. The ultracapacitor is connected to DC bus using bidirectional converter in current control mode. Power split mechanism is implemented such that, drive current transient requirements, is met by the ultracapacitor depending on its voltage condition and remaining current requirement is met by battery.

The design of low-power, energy-efficient communication modules was a pivotal challenge in the development of wearable health monitoring devices, where prolonged battery life and reliable data transmission are critical for continuous, real-time monitoring. This chapter explores the latest advancements in communication technologies, power management strategies, and energy-efficient hardware aimed at optimizing wearable devices for healthcare applications. Key topics discussed include AI-based power management techniques, hybrid communication architectures, energy-efficient antennas, and lightweight security mechanisms. Special focus was given to the integration of Artificial Intelligence (AI) for dynamic power optimization and adaptive data transmission, enhancing both performance and battery longevity. The impact of hybrid communication systems and seamless handoff protocols was also analyzed, highlighting their potential in reducing energy consumption while ensuring uninterrupted service.the challenges posed by security protocols and their trade-offs in energy efficiency are examined, with an emphasis on achieving a balance between safety and power usage. This chapter provides a comprehensive overview of the current research and future directions in low-power communication systems, with a view to enabling more efficient, secure, and sustainable wearable health monitoring technologies.

There is a growing gap between the peak speed of parallel computing systems and the actual delivered performance for scientific applications. In general this gap is caused by inadequate architectural support for the requirements of modern scientific applications, as commercial applications and the much larger market they represent, have driven the evolution of computer architectures. This gap has raised the importance of developing better benchmarking methodologies to characterize and to understand the performance requirements of scientific applications, to communicate them efficiently to influence the design of future computer architectures. This improved understanding of the performance behavior of scientific applications will allow improved performance predictions, development of adequate benchmarks for identification of hardware and application features that work well or poorly together, and a more systematic performance evaluation in procurement situations. The Berkeley Institute for Performance Studies has developed a three-level approach to evaluating the design of high end machines and the software that runs on them: (1) A suite of representative applications; (2) A set of application kernels; and (3) Benchmarks to measure key system parameters. The three levels yield different type of information, all of which are useful in evaluating systems, and enable NSF and DOE centers to select computer architectures more suited for scientific applications. The analysis will further allow the centers to engage vendors in discussion of strategies to alleviate the present architectural bottlenecks using quantitative information. These may include small hardware changes or larger ones that may be out interest to non-scientific workloads. Providing quantitative models to the vendors allows them to assess the benefits of technology alternatives using their own internal cost-models in the broader marketplace, ideally facilitating the development of future computer architectures more suited for scientific computations. The three levels also come with vastly different investments: the benchmarking efforts require significant rewriting to effectively use a given architecture, which is much more difficult on full applications than on smaller benchmarks.

Today's fluid simulations deal with complex geometries and numerical data on an extreme scale. As computation approaches the exascale, it will no longer be possible to write and store the full-sized data set. In situ data analysis and scientific visualisation provide feasible solutions to the analysis of complex large scaled CFD simulations. To bring pre- and postprocessing to the exascale we must consider modifications to data structure and memory layout, and address latency and error resiliency. In this respect, a particular challenge is the exascale data processing for the sparse geometry lattice-Boltzmann code HemeLB, intended for hemodynamic simulations. In this paper, we assess the needs and challenges of HemeLB users and sketch a co-design infrastructure and system architecture for pre- and post-processing the simulation data. To enable in situ data visualisation and analysis during a running simulation, post-processing needs to work on a reduced subset of the original data. Particular choices of data structure and visualisation techniques need to be co-designed with the application scientists in order to achieve efficient and interactive data processing and analysis. In this work, we focus on the hierarchical data structure and suitable visualisation techniques which provide possible solutions to interactive in situ data processing at exascale. Architectural challenges and road-maps will be presented as the major focus of this paper. We sketch a software architecture which integrates pre- and post-processing techniques that can provide in situ analysis and ultimately computational steering to HemeLB.

CubeSats, a revolutionary class of small satellites, have significantly democratized access to space by offering compact, cost-effective, and highly versatile platforms for scientific, commercial, and educational applications. These miniaturized spacecrafts have become indispensable tools for conducting Earth observation, atmospheric studies, deep-space exploration, and technology demonstrations. However, the efficient management of power systems remains a critical challenge due to the constraints imposed by their small size, limited surface area for solar energy harvesting, and the harsh conditions of space. This paper explores innovative approaches to optimizing the energy systems of CubeSats, focusing on advancements in solar energy harvesting, energy storage technologies, and power management strategies. It highlights the potential of high- efficiency photovoltaic materials, such as multi-junction and perovskite solar cells, to enhance energy capture by utilizing broader portions of the solar spectrum. The study also examines energy storage technologies, including lithium-ion, lithium-polymer, and emerging solid-state batteries, comparing their performance, reliability, and applicability in space missions. Furthermore, hybrid energy storage systems, which combine different storage technologies like batteries and supercapacitors, are analyzed for their ability to dynamically balance energy supply and demand, ensuring stability and efficiency under variable conditions. The research delves into advanced optimization techniques, such as Maximum Power Point Tracking (MPPT) algorithms, which adaptively maximize energy extraction from solar panels, and thermal management strategies that mitigate efficiency losses due to overheating. Load scheduling, depth-of-discharge control, and real-time diagnostics are discussed as critical strategies to extend battery lifespan and ensure continuous operation. Mathematical models and experimental data are integrated to validate these techniques and provide actionable insights for the design and operation of CubeSat energy systems. By addressing the unique challenges of CubeSat power management, this study contributes to the development of robust, sustainable, and high-performance energy solutions. These advancements not only enhance CubeSat mission capabilities and longevity but also pave the way for more ambitious applications, including interplanetary exploration and multi- satellite constellations. The findings of this research are intended to benefit academia, space agencies, and industries, fostering innovation in small satellite technology and expanding the possibilities for cost-effective and impactful space exploration.

This paper describes an object‐oriented interface for the memory management of sparse discrete mathematical operators in numerical scientific applications. The basic data structure we propose is intended to implement and cope with mathematical vectors in sparse format. This structure allows an effective implementation of more complex data structures, such as matrix and matrix‐like objects, to represent the discrete version of differential operators in numerical application codes. A research example concerning adaptive wavelet algorithms illustrates the possible applications.Copyright © 2002 John Wiley & Sons, Ltd.

Introduction: The substantial energy usage in the cloud computing system is a significant drawback both for the cloud providers along with the cloud service users. In order to decrease the amount of energy used, it is necessary to implement virtualization techniques. Objectives: this paper employs a GA to optimize the migration of VMs and reduce energy usage in a cloud environment. Methods: The VM consolidation method effectively handles cloud resources, meeting the needs of both cloud consumers and suppliers. Moreover, it helps in enhancing the efficiency of servers while simultaneously decreasing the excessive energy usage in data canters. Nevertheless, the unnecessary activities of the VM consolidation technique result in inadequate VM selection and improper VM assignment, causing low performance, QoS, and violations of SLAs. Data center management struggles with energy consumption. VM migration and placement works well for this. Data centers require energy-saving solutions without affecting other parameters. Results: The performance of the proposed method has been evaluated utilizing factors like CPU usage, memory usage, network speed, power usage, and SLA breaches. Conclusions: The comparative analysis of the proposed approach with existing methods highlights its effectiveness and trustworthiness.

As a non-invasive detection method, Electrical Tomography (ET) can reach a much higher imaging speed than γ-ray or X-ray CT, etc. To archive an even higher speed, there are several factors which should be considered thoroughly, such as excitation/measurement strategy, data acquisition system design, communication protocol, reconstruction algorithm, imaging software, etc. To apply the wooden bucket theory to this situation, the slowest section determines the system's actual speed. And imaging software becomes the lowest board in some cases. Based on Tianjin University Electrical Tomography (TJU-ET) hardware, this paper introduced optimized multithread program architecture, standard BLAS (Basic Linear Algebra Subprograms) and LAPACK (Linear Algebra PACKage) numerical library, optimized OpenGL rendering engine and many other improvements into imaging software. As a result, the brand new software can fully releases TJU-ET system's power, which means it can perform 750fps on TJU dual plane 12-electrode ECT system.

The objective of this paper and our current research is to optimize energy usage in a humanoid robot during diverse tasks such as basic walking by dynamically controlling individual joint stiffness. In the current work we analyze individual and total usage of current, voltage and power in a NAO V4 humanoid robot joints during short walks around a circle at different speeds and under varying control of joint stiffness. We perform experimental studies to understand the main factors affecting power consumption and energy usage and look at ways to improve overall energy usage. We describe experiments and corresponding results. We discuss the state of advancement of our research.

I. ABSTRACT High performance in the age of lightning-fast computing and energy-efficient electrical devices depends on the creation of optimised arithmetic units. The efficiency of adders, which are crucial components of the majority of digital systems, has a direct effect on the system's overall speed and power usage. In this study, a low-power, high-speed parallel prefix adder (PPA) architecture leveraging Cadence nclaunch is designed and implemented. In terms of gates, power, and area, prefixes topologies such as Brent-Kung and Kogge-Stone are examined and contrasted. The suggested architecture uses effective gate-level implementations and logic-level optimisations to lower power consumption without sacrificing speed. To confirm functionality and performance, the design is synthesised and simulated using the cadence tool. When compared to traditional adder architectures, the findings show a notable gain in both power efficiency and area, demonstrating the usefulness of the suggested integration strategy in contemporary low-power, high-speed VLSI systems. Keywords: PPA, EDP, SoC, FPGA.

Cloud infrastructures increasingly include a heterogeneous mix of components in terms of performance, power, and energy usage. As the size of cloud infrastructures grows, power consumption becomes a significant constraint. We use Apache Mesos and Apache Aurora, which provide massive scalability to web-scale applications, to demonstrate how a policy driven approach involving bin-packing workloads according to their power profiles, instead of the default allocation by Mesos and Aurora, can effectively reduce the peak-power and energy usage as well as the node utilization, when workloads are co-scheduled. Our experimental results show reductions of 11% in peak power, 86% for total energy usage, and an increase in utilization of 148% for memory and 8% CPU for the different policies.

Several MANET designs looked into renewable resource efficient protocols. The amount of energy dissipated by MANET nodes is an important coordinating component. In monitoring applications, power usage is ensured, and battery depletions are reduced to avoid frequent replacement. MANET's primary goal is to relay data by utilising energy-efficient routing techniques and extending the network's lifespan. In this research, we provide an Energy Efficient Protocol on Remaining Energy, a new paradigm for MANET multicasting. Our idea is based on a tone system that integrates solutions across the Medium Access Control (MAC) layer to deliver more efficiency and better performance. To provide increased stability and lower energy consumption, the protocol includes a new construction mechanism for mobile nodes that uses a clustering strategy based on distance and remaining energy. We also propose modifying the conventional multicast flow by introducing unicast links between clusters. In addition, we describe a strategy based on a busy tone system to eliminate excessive control overhead produced by control packets in error recovery in our model. The simulation results are displayed in NS2, and they indicate improvements in reliability, packet delivery ratio, energy usage, and throughput.