• Energy Research

Abstract Computing systems have been focused on performance improvements, driven by the demand of user applications in past few decades, particularly from 1990 to 2010. However, due to their ever-increasing energy demand which causes large energy bills and CO 2 emissions, over the past six years the focus has shifted towards energy-performance aware. The average energy consumption of servers is increasing continuously; and several researchers suggest, if this trend continues further, the cost of energy consumed by a server during its lifetime will exceed the hardware costs. The energy consumption problem is even greater for large-scale infrastructures, such as clusters, grids and clouds, which consist of several thousand heterogeneous servers. Efforts are continuously made to minimize the energy consumption of these systems, but the interest of people in computational services and popularity of smart devices make it a difficult task. In this paper, we discuss the energy consumption of ICT equipment, and present a taxonomy of energy and performance efficient techniques for large computing systems covering clusters, grids and clouds (datacenters). We discuss both energy and performance efficiency, which makes this survey different from those already published in the literature. Key research papers are surveyed and mapped onto taxonomies to characterise and identify outstanding and key issues for further research. We discuss several state-of-the-art resource management techniques, reported in the literature, that claim significant improvement in the energy efficiency and performance of ICT equipment and large-scale computing systems such as datacenters, and identify a few open challenges.

AbstractThis work implements the recently developednth state Markovian jumping particle swarm optimisation (PSO) algorithm with local search (NS-MJPSOloc) awareness method to address the economic/environmental dispatch (EED) problem. The proposed approach, known as the Non-dominated Sorting Multi-objective PSO with Local Best (NS-MJPSOloc), aims to enhance the performance of the PSO algorithm in multi-objective optimisation problems. This is achieved by redefining the concept of best local candidates within the search space of multi-objective optimisation. The NS-MJPSOlocalgorithm uses an evolutionary factor-based mechanism to identify the optimum compromise solution, a Markov chain state jumping technique to control the Pareto-optimal set size, and a neighbourhood’s topology (such as a ring or a star) to determine its size. Economic dispatch refers to the systematic allocation of available power resources in order to fulfill all relevant limitations and effectively meet the demand for electricity at the lowest possible operating cost. As a result of heightened public consciousness regarding environmental pollution and the implementation of clean air amendments, nations worldwide have compelled utilities to adapt their operational practises in order to comply with environmental regulations. The (NS-MJPSOloc) approach has been utilised for resolving the EED problem, including cost and emission objectives that are not commensurable. The findings illustrate the efficacy of the suggested (NS-MJPSOloc) approach in producing a collection of Pareto-optimal solutions that are evenly dispersed within a single iteration. The comparison of several approaches reveals the higher performance of the suggested (NS-MJPSOloc) in terms of the diversity of the Pareto-optimal solutions achieved. In addition, a measure of solution quality based on Pareto optimality has been incorporated. The findings validate the effectiveness of the proposed (NS-MJPSOloc) approach in addressing the multi-objective EED issue and generating a trade-off solution that is both optimal and of high quality. We observed that our approach can reduce$$\sim $$∼6.4% of fuel costs and$$\sim $$∼9.1% of computational time in comparison to the classical PSO technique. Furthermore, our method can reduce$$\sim $$∼9.4% of the emissions measured in tons per hour as compared to the PSO approach.

Les centres de données fournissent les bases de l'informatique en nuage, mais nécessitent de grandes quantités d'électricité pour leur fonctionnement. Les approches qui promettent de réduire la consommation d'énergie en minimisant le temps d'exécution, par exemple en utilisant différentes techniques de planification et de gestion des ressources, sont discutées dans la littérature. Cet article résume certaines des techniques de planification les plus importantes dans les nuages axées sur la consommation d'énergie, couvrant la planification au niveau des machines virtuelles, des hôtes et des tâches, l'approche la plus prometteuse étant la planification au niveau des tâches, avec des économies d'énergie grâce au filtrage de la charge, à la consolidation, au débit CPU adapté ou au contrôle de la puissance de l'hôte. Nous explorons l'utilisation des algorithmes de monotonie de débit (RM) et de remblayage pour l'ordonnancement des tâches en temps réel dans un environnement cloud, car la RM est la technique d'ordonnancement à priorité fixe la plus simple, et donc le choix pour les systèmes modernes en temps réel, et les utilisations antérieures de la RM dans l'ordonnancement des tâches ont démontré une efficacité énergétique avec des résultats optimaux. Nous considérons spécifiquement la planification des tâches en fonction des délais pour les nuages en temps réel qui, à notre connaissance, n'ont pas été utilisés auparavant. La MP avec remblayage est évaluée expérimentalement et les résultats montrent que, par rapport aux algorithmes classiques, toutes les tâches ont été planifiées avec une consommation d'énergie minimale (5,5% – 29,3%), sur des ressources minimales (3,9% – 25,2% de moins) tandis que la majorité respectaient leurs délais (93,21% – 94,7%). L'approche peut garantir un logiciel en tant que service (SaaS) axé sur les délais dans le cloud si le taux d'arrivée, c'est-à-dire le temps de transfert du réseau, peut être estimé à l'avance. Nous avons par la suite fourni une extension de l'approche proposée à l'équilibrage de charge basé sur les tâches pour une utilisation des ressources presque équilibrée et une efficacité énergétique d'environ 1,0 % à 1,6 %. Los centros de datos proporcionan las bases para la computación en la nube, pero requieren grandes cantidades de electricidad para su funcionamiento. Los enfoques que prometen reducir el uso de energía al minimizar el tiempo de ejecución, por ejemplo, utilizando diferentes técnicas de programación y gestión de recursos, se discuten en la literatura. Este documento resume algunas de las técnicas de programación más importantes en las nubes que se centran en el consumo de energía, cubriendo la programación a nivel de máquina virtual, a nivel de host y a nivel de tarea, donde el enfoque más prometedor es la programación a nivel de tarea, con ahorros de energía por medio de filtrado de carga, consolidación, rendimiento de CPU adaptado o control de potencia del host. Exploramos el uso de los algoritmos de tasa monótona (RM) y de relleno para la programación de tareas en tiempo real en el entorno de la nube porque la RM es la técnica de programación de prioridad fija más simple y, por lo tanto, la elección para los sistemas modernos en tiempo real, y los usos anteriores de la RM en la programación de tareas han demostrado eficiencia energética con resultados óptimos. Consideramos específicamente la programación de tareas basada en plazos para nubes en tiempo real que, a nuestro leal saber y entender, no se han empleado anteriormente. La RM con relleno se evalúa experimentalmente y los resultados muestran que, en comparación con los algoritmos clásicos, todas las tareas se programaron con un consumo mínimo de energía (5.5% – 29.3%), con recursos mínimos (3.9% – 25.2% menos) mientras que la mayoría cumplía con sus plazos (93.21% – 94.7%). El enfoque puede garantizar un software como servicio (SaaS) orientado a plazos en la nube si la tasa de llegada, es decir, el tiempo de transferencia de la red, se puede estimar por adelantado. Posteriormente, proporcionamos una extensión del enfoque propuesto para el equilibrio de carga basado en tareas para una utilización de recursos casi equilibrada y una eficiencia energética de aproximadamente 1.0% a 1.6%. Datacentres provide the foundations for cloud computing, but require large amounts of electricity for their operation. Approaches that promise to reduce power use by minimizing execution time, for example using different scheduling and resource management techniques, are discussed in the literature. This paper summarizes some of the most important scheduling techniques in clouds focusing on power consumption, covering VM-level, host-level and task-level scheduling where the most promising approach is task level scheduling, with energy savings by means of load filtering, consolidation, adapted CPU throughput, or host power control. We explore use of the rate monotonic (RM) and backfilling algorithms for real-time task scheduling in cloud environment because RM is the simplest fixed priority scheduling technique, and thus the choice for modern real-time systems, and prior uses of RM in task scheduling have demonstrated power efficiency with optimal results. We specifically consider deadline-based tasks scheduling for real-time clouds which, to the best of our knowledge, has not been employed previously. RM with backfilling is experimentally evaluated and results show that, compared to the classical algorithms, all tasks were scheduled with minimum power consumption (5.5% – 29.3%), on minimum resources (3.9% – 25.2% less) while majority were meeting their deadlines (93.21% – 94.7%). The approach can guarantee deadline oriented Software as a Service (SaaS) in cloud if arrival rate i.e. network transfer time can be estimated in advance. We subsequently provided an extension of the proposed approach to task-based load balancing for almost balanced resource utilization and approximately 1.0% to 1.6% energy efficiency. توفر مراكز البيانات الأسس للحوسبة السحابية، ولكنها تتطلب كميات كبيرة من الكهرباء لتشغيلها. تتم مناقشة الأساليب التي تعد بتقليل استخدام الطاقة من خلال تقليل وقت التنفيذ، على سبيل المثال باستخدام تقنيات الجدولة وإدارة الموارد المختلفة، في الأدبيات. تلخص هذه الورقة بعضًا من أهم تقنيات الجدولة في السحب التي تركز على استهلاك الطاقة، وتغطي الجدولة على مستوى الجهاز الافتراضي ومستوى المضيف ومستوى المهام حيث يكون النهج الواعد هو جدولة مستوى المهمة، مع توفير الطاقة عن طريق تصفية الحمل أو الدمج أو إنتاجية وحدة المعالجة المركزية المعدلة أو التحكم في طاقة المضيف. نستكشف استخدام خوارزميات المعدل الرتيب (RM) والردم لجدولة المهام في الوقت الفعلي في البيئة السحابية لأن RM هي أبسط تقنية جدولة ذات أولوية ثابتة، وبالتالي فإن اختيار الأنظمة الحديثة في الوقت الفعلي، والاستخدامات السابقة لـ RM في جدولة المهام أثبتت كفاءة الطاقة مع النتائج المثلى. نحن نأخذ في الاعتبار على وجه التحديد جدولة المهام القائمة على الموعد النهائي للسحابات في الوقت الفعلي والتي، على حد علمنا، لم يتم توظيفها سابقًا. يتم تقييم إدارة المخاطر مع الردم تجريبيًا وتظهر النتائج أنه مقارنة بالخوارزميات الكلاسيكية، تمت جدولة جميع المهام مع الحد الأدنى من استهلاك الطاقة (5.5 ٪ – 29.3 ٪)، على الحد الأدنى من الموارد (3.9 ٪ – 25.2 ٪ أقل) في حين أن الغالبية كانت تفي بالمواعيد النهائية (93.21 ٪ – 94.7 ٪). يمكن أن يضمن النهج البرامج الموجهة نحو الموعد النهائي كخدمة (SaaS) في السحابة إذا كان من الممكن تقدير معدل الوصول، أي وقت نقل الشبكة مسبقًا. قدمنا لاحقًا امتدادًا للنهج المقترح لموازنة الحمل القائم على المهام من أجل استخدام الموارد بشكل متوازن تقريبًا ونحو 1.0 ٪ إلى 1.6 ٪ من كفاءة الطاقة.

Cloud computing is the de facto platform for deploying resource- and data-intensive real-time applications due to the collaboration of large scale resources operating in cross-administrative domains. For example, real-time systems are generated by smart devices (e.g., sensors in smart homes that monitor surroundings in real-time, security cameras that produce video streams in real-time, cloud gaming, social media streams, etc.). Such low-end devices form a microgrid which has low computational and storage capacity and hence offload data unto the cloud for processing. Cloud computing still lacks mature time-oriented scheduling and resource allocation strategies which thoroughly deliberate stringent QoS. Traditional approaches are sufficient only when applications have real-time and data constraints, and cloud storage resources are located with computational resources where the data are locally available for task execution. Such approaches mainly focus on resource provision and latency, and are prone to missing deadlines during tasks execution due to the urgency of the tasks and limited user budget constraints. The timing and data requirements exacerbate the efficient task scheduling and resource allocation problems. To cope with the aforementioned gaps, we propose a time- and cost-efficient resource allocation strategy for smart systems that periodically offload computational and data-intensive load to the cloud. The proposed strategy minimizes the data files transfer overhead to computing resources by selecting appropriate pairs of computing and storage resources. The celebrated results show the effectiveness of the proposed technique in terms of resource selection and tasks processing within time and budget constraints when compared with the other counterparts.

Multi-processor systems consist of more than one processor and are mostly used for computationally intensive applications. Real-time systems are those systems that require completing execution of tasks within a pre-defined deadline. Traditionally, multiprocessor systems are given attention in periodic models, where tasks are executed at regular intervals of time. Gradually, as maturity in a multiprocessor design had increased; their usage has become very common for real-time systems to execute both periodic and aperiodic tasks. As the priority of an aperiodic task is usually but not essentially greater than the priority of a periodic task, they must be completed within the deadline. There is a lot of research works on multiprocessor systems with scheduling of periodic tasks, but the task scheduling is relatively remained unexplored for a mixed workload of both periodic and aperiodic tasks. Moreover, higher energy consumption is another main issue in multiprocessor systems. Although it could be reduced by using the energy-aware scheduling technique, the response time of aperiodic tasks still increases. In the literature, various techniques were suggested to decrease the energy consumption of these systems. However, the study on reducing the response time of aperiodic tasks is limited. In this paper, we propose a scheduling technique that: 1) executes aperiodic tasks at full speed and migrates periodic tasks to other processors if their deadline is earlier than aperiodic tasks-reduces the response time and 2) executes aperiodic tasks with lower speed by identifying appropriate processor speed without affecting the response time-reduces energy consumption. Through simulations, we demonstrate the efficiency of the proposed algorithm and we show that our algorithm also outperforms the well-known total bandwidth server algorithm.

In many production clouds, with the notable exception of Google, aggregation-based VM placement policies are used to provision datacenter resources energy and performance efficiently. However, if VMs with similar workloads are placed onto the same machines, they might suffer from contention, particularly, if they are competing for similar resources. High levels of resource contention may degrade VMs performance, and, therefore, could potentially increase users' costs and infrastructure's energy consumption. Furthermore, segregation-based methods result in stranded resources and, therefore, less economics. The recent industrial interest in segregating workloads opens new directions for research. In this paper, we demonstrate how aggregation and segregation-based VM placement policies lead to variabilities in energy efficiency, workload performance, and users' costs. We, then, propose various approaches to aggregation-based placement and migration. We investigate through a number of experiments, using Microsoft Azure and Google's workload traces for more than twelve thousand hosts and a million VMs, the impact of placement decisions on energy, performance, and costs. Our extensive simulations and empirical evaluation demonstrate that, for certain workloads, aggregation-based allocation and consolidation is ~9.61% more energy and ~20.0% more performance efficient than segregation-based policies. Moreover, various aggregation metrics, such as runtimes and workload types, offer variations in energy consumption and performance, therefore, users' costs.<br>

Researchers are turning to nanofluids in PV/T hybrid systems for enhanced efficiency due to nanoparticle dispersion, improving thermal and optical properties over conventional fluids. Three different concentrations of formulated soybean oil based MXene nanofluids are considered 0.025, 0.075 and 0.125 wt.%. Maximum specific heat capacity nanofluids ( $c_{pNF}$ ) augmentation is 24.49% at 0.125 wt.% loading of Ti3C2 in the base oil. The calculation of the $c_{pNF}$ based on the temperature and nanoflakes concentration is very expensive and time-consuming as it should be calculated via the practical test investigation. This study employs a long short-term memory (LSTM) as an efficient machine learning method to extract the surrogate model for calculating the $c_{pNF}$ based on the temperature and nanoflakes concentration. In addition, a couple of other machines learning methods, including support vector regression (SVR), group method of data handling (GMDH), and multi-layer perceptron (MLP), are developed to prove the higher efficiency of the recently proposed LSTM model in the calculation of the $c_{pNF}$ . In addition, the Bayesian optimization technique is employed to calculate the optimal hyperparameters of the developed SVR, GMDH, MLP and LSTM to reach the highest efficiency of the system in predicting the $c_{pNF}$ based on temperature and nanoflakes concentration. Notably, 95% of the recorded data via differential scanning calorimetry (DSC) is used for training machine learning techniques. In comparison, 5% is used for testing and validation purposes of the developed algorithm. The newly proposed optimized SVR, GMDH, MLP, and LSTM are modelled in MATLAB software. The results show that the newly proposed optimized LSTM model can reduce the mean square error in calculating the $c_{pNF}$ by 99%, 99% and 91% compared with SVM, GMDH and MLP, respectively. The proposed methodology can be used to calculate other thermophysical properties of nanofluids.