Search

In a decade, AI frontier research transitioned from the researcher's workstation to thousands of high-end hardware-accelerated compute nodes. This rapid evolution shows no signs of slowing down in the foreseeable future. While top cloud providers may be able to keep pace with this growth rate, obtaining and efficiently exploiting computing resources at that scale is a daunting challenge for universities and SMEs. This work introduces the Cross-Facility Federated Learning (XFFL) framework to bridge this compute divide, extending the opportunity to efficiently exploit multiple independent data centres for extreme-scale deep learning tasks to data scientists and domain experts. XFFL relies on hybrid workflow abstractions to decouple tasks from environment-specific technicalities, reducing complexity and enhancing reusability. In addition, Federated Learning (FL) algorithms eliminate the need to move large amounts of data between different facilities, reducing time-to-solution and preserving data privacy. The XFFL approach is empirically evaluated by training a full LLaMAv2 7B instance on two facilities of the EuroHPC JU, showing how the increased computing power completely compensates for the additional overhead introduced by two data centres. [ABSTRACT FROM AUTHOR]

• To improve data center performance, it must be efficiently cooled. • This paper undertakes a detailed examination of the most recent cooling optimization techniques. • Several methods are discussed like regular air conditioning, free cooling, and liquid cooling. • An assessment of overhead and underfloor air delivery technologies is presented. • The effects of baffles, variable airflow control, and server placement are investigated. In order to increase data centers' efficiency and performance, a proper cooling system should be applied. This article provides a comprehensive assessment which explores current cooling optimization technologies for data centers. Standard traditional technologies like air conditioning, free cooling, and liquid cooling are investigated. Their limitations as well as proposed improvements are thoroughly discussed. The article also delves into underfloor and overhead air supply technologies, presenting their advantages and drawbacks. The review introduces novel approaches at the rack level cooling, including baffles, variable airflow management, and optimal server placement. Key findings stress the efficacy of optimized airflow systems and innovative rack-level cooling, underlining their role in reducing energy consumption and enhancing overall performance. Notably, potential energy savings of up to 67.2 % compared to traditional methods are demonstrated. The adoption of advanced cooling technologies, such as direct and indirect natural cooling, liquid-cooling cold plates, submersion, heat pipe, and thermosiphon-based cooling, exhibits promising energy efficiency gains and reduced Power Usage Effectiveness (PUE) values. Research into content-specific efficiency enhancements, adaptable energy-saving tactics, and all-encompassing server power consumption models is necessary. To work toward future sustainable, energy-efficient data centers, researchers should investigate how CRACs and server fans interact, find ways to integrate thermal and power management, and how artificial intelligence can be used to cool servers in an eco-friendly way. [ABSTRACT FROM AUTHOR]

• Advantage of hot water cooling was proven by entropy and entranspy theory. • Thermodynamic model of a waste heat recovery-free cooling data center was established. • Heat transfer efficiency of waste heat recovery-free cooling data centers was proven by entranspy theory. • Performance indicators of waste heat recovery-free cooling data center was predicted. This study investigated the potential of integrating separate heat pipe and adsorption refrigeration systems as a data center efficient cooling system. Compared to water-cooled data centers, this approach demonstrates improved heat transfer efficiency and reduced energy consumption. The effectiveness of heat transfer in this approach is illustrated by applying exergy and entranspy theories. By conducting numerical simulations across five representative Chinese cities with different climates, both Power Usage Effectiveness (PUE) and key cooling system performance indicators are assessed. The analysis reveals that all cities can achieve a PUE below 1.095, realized a 65.87 % and 13.60 % reduction in cooling system energy consumption and entransy dissipation, respectively. Data centers using mechanical chiller cooling cause a high energy consumption. This paper indicates the benefits of integrating heat pipe and adsorption refrigeration systems in reducing energy consumption and optimizing overall data center performance. [ABSTRACT FROM AUTHOR]

The explosive adoption of IoT applications in different domains, such as healthcare, transportation, and smart home and industry, has led to the pervasive adoption of edge and cloud computing. Large-scale edge and cloud data centers, consisting of thousands of computing servers, are hungry-energy infrastructure exacerbating issues such as environmental carbon footprint and high electricity costs. Developing energy-efficient solutions for cloud infrastructure requires knowledge of the correlation between computing server resource utilization and power consumption. Power consumption modeling exhibits this relationship and is crucial for energy savings. In this paper, we propose PowerGen, a framework to generate server resources utilization and corresponding power consumption dataset. The proposed framework will aid academic researchers to formulate correlations between resources utilization and power consumption by using power prediction models, and evaluate energy-aware resource management approaches in an edge-cloud computing system. It will help edge and cloud administrators to evaluate the energy-efficiency of heterogenous severs architectures in a datacenter. We exemplify the applicability of the dataset, generated by our proposed framework, in power prediction modeling and energy-aware scheduling for green computing scenarios. [ABSTRACT FROM AUTHOR]

In the dynamic landscape of cloud computing, efficient resource allocation and workload prediction are paramount for optimizing infrastructure utilization, cost management, and overall service quality. As a result, an efficient resource management approach involves pooling available resources among many customers in a manner that considers their energy usage, SLAs, and forecasting accuracy all at once. Based on the performance analysis results, it is determined that integrating these techniques would result in the most efficient and adaptable cloud data centre. We proposed a Model based on Deep Convolutional Neural Networks (DCNN) and Long Short-Term Memory (LSTM) for handling SLAs in the cloud from both the perspective of consumers and service providers. Subsequently, we delve into the methodology of applying deep CNN and LSTM models to the problem of workload prediction in cloud environments. This methodology encompasses data preprocessing, model architecture, training parameters, and the choice of performance metrics. To anticipate CPU consumption from time series data and detect SLA violations, we suggested a DCNN-LSTM model. The accuracy prediction, energy usage, CPU use, and Service Level Agreement monitoring are all a part of this model. The proposed method is effective in helping cloud providers cut down on service violations and associated fines. Regarding the composite metric of Energy SLA Violation, which assesses the combined aspects of energy use and adherence to Service Level Agreements (SLAs), DCNN-LSTM surpasses ARIMA-LSTM, CNN, LSTM, and ARIMA 6.8%, 10.88%, 16.6%, and 22.4%, respectively. [ABSTRACT FROM AUTHOR]

Cloud computing technology provides access on demand to virtualized resources, services, and applications via a distributed network. In cloud data centers effective energy utilization is a critical concern in today's technology-driven world. Cloud data centers (CDC) are massive facilities that host and manage an enormous amount of data and computing resources. This article addresses the growing significance and energy-intensive nature of data centers. Due to the rapid growth of cloud computing, it offers on-demand access to resources globally and leads to substantial power consumption and carbon impact on the environment. They consume substantial amounts of energy, and optimizing their energy utilization is essential for reducing operational costs, minimizing environmental impact, and ensuring sustainable growth. To combat this, efficient energy-saving approaches using machine learning methods have been researched. ML methods hold great potential for enhancing energy efficiency in CDCs by analysing data, detecting patterns, and optimizing resource usage. The focus areas include CPU usage prediction, overload finding, underload estimation, selection, migration, and relocation of VMs to attain resources and improve energy utilization. The paper compares energy-saving results achieved by different machine learning techniques in data centers to minimize energy usage and meet service level agreements (SLA). The machine learning approaches reduce energy consumption from 1.6% to 88.5% compared to the benchmark approach mentioned, considering various settings and parameters. [ABSTRACT FROM AUTHOR]

Connecting to the internet will increase our computing challenges as it becomes an integral part of our daily lives. Therefore, it is necessary to advance the service qualities of Internet of Things (IoT) applications because the data produced by all these devices will need to be processed quickly and sustainably. Previously, cloud data centers with large capacity interface IoT devices with support servers. While IoT devices proliferate and generate massive amounts of data, communicating between devices and the Cloud is becoming more complex and harder, resulting in high costs and inefficiencies. Fog computing emerges as an approach to address the growing demand for IoT solutions. In this article, an IoT-fog-cloud application's general framework is developed, followed by an algorithm for Energy efficiency through an integrated approach computation model. Fog-Enabled Smart Cities (FESC) are proposed to minimize service delay and response time by using a fog offloading policy for the fog-enabled IoTs. Also, we developed an analytical model evaluating the proposed framework's effectiveness in reducing the delay of IoT services. Comparing the proposed model and the Alternating Direction Method of Multipliers (ADMM-VS) algorithm, the proposed model performs significantly better. Thus, by optimizing response and processing times, fog-enabled smart grids determine whether computation will be performed autonomously or semi-autonomously on fog nodes or in the Cloud. [ABSTRACT FROM AUTHOR]

The technological advancements of Connected and Automated Vehicles (CAVs) are outpacing the current regulatory regime, potentially resulting in a disconnect between legislators, technology, and CAV stakeholders. Although many studies explore the regulatory requirements of operations of CAVs, studies on regulatory challenges specific to the cybersecurity of CAVs are also emerging and receiving lots of attention among researchers and practitioners. However, studies providing an up-to-date synthesis and analysis on CAVs regulatory requirements specific to cyber-risk reduction or mitigation are almost non-existent in the literature. This study aims to overcome this limitation by presenting a comprehensive overview of the role of key Intelligent Transportation Systems (ITS) stakeholders in CAV's cybersecurity. These stakeholders include road operators, service providers, automakers, consumers, repairers, and the general public. The outcome of this review is an in-depth synthesis of CAV-based ITS stakeholders by visualising their scope in developing a Cybersecurity Regulatory Framework (CRF). The study demonstrated the compliance requirements for ITS communication service providers, regulatory standards for CAVs automakers, policy readiness for CAVs customers and the general public who interact with CAVs, and the role of the CAVs Network Operator Centre in regulating CAVs data flow. Moreover, the study illuminates several critical pathways necessary in future for synthesizing and forecasting the legal landscape of CAV-based transportation systems to integrate the regulatory framework for CAV stakeholders. The paper's findings and conclusions would assist policymakers in developing a comprehensive CRF. CAV Cybersecurity Regulation, Stakeholder Roles, and Conceptualization of CAVs (traffic laws) breaches. [Display omitted] • The study offers a comprehensive overview of CAV Cybersecurity's Regulatory Framework. • Considers road operators, service providers, automakers, consumers, repairers and the public. • Emphasizes automaker regulatory requisites for cyber-safe CAV operation. • Proposes CAVs Network Operation Centre to regulate data flow. • Advocates eSafety Traffic Unit for regulating CAV operations. [ABSTRACT FROM AUTHOR]

In this paper, we utilized a eutectic gallium-indium liquid metal (LM) with high thermal conductivity to immerse carbon-based phase change material (PCM) capsules to acquire a PCM-LM composite. This produced a scalable, thermally stable, mechanically reliable, high thermally conductive, and leakage-free material. This composite was then applied as a heat mitigator for CPU cooling. Results demonstrate that clear improvements could be made compared to traditional pin-fin heat-sink cooling. Here the continuous full-load operating time could be prolonged by 414.3%. Moreover, the temperature difference between the idle and full-load operating states could be significantly decreased, and prevent damage from thermal stress. Significantly, a 23% energy-saving performance was attained because of the thermal buffering effect and high responsive ability of the composite. Guidelines for an effective match between the heat absorption of the composite and the heat generation from the CPU could also be established. Here the proposed PCM-LM composite can be applied to other PCM-based energy storage industries and efficient temperature control applications, contributing to carbon neutralization and global warming mitigation. [ABSTRACT FROM AUTHOR]

Changes to which age groups are recommended for mammography may have affected screening rates for all women. This spillover effect has not previously been shown empirically in a national sample. Using data from the Centers for Disease Control and Prevention's Behavioral Risk Factor Surveillance System from 2002 to 2018, authors tested whether screening trends at a national level changed after the 2009 U.S. Preventive Services Task Force guideline change. The authors also tested whether state-level screening trends for women aged 40–49 years and 75+ years were associated with screening trends for women aged 50–74 years. Analyses were conducted in March–December 2022. In a model predicting state-level trends in screening for women aged 50–74 years, authors find positive, statistically significant associations with screening trends for women aged 40–49 years (p =0.033) and for women aged 75+ years (p <0.001). Deimplementation is difficult and important for controlling healthcare spending and delivering high value care. However, states most successful at reducing mammography screening rates among those aged 40–49 years and 75+ years also had greater reductions in recommended screening among women aged 50–74 years. More work is needed to understand and mitigate the unintended consequences of deimplementation. [ABSTRACT FROM AUTHOR]

The limited available data regarding the prevalence of transthyretin amyloidosis, both for wild-type (ATTRwt) and hereditary form (ATTRv), is inferred from highly selected patients and subsequent extrapolations that limit the comprehension of the clinical disease impact. The Tuscan healthcare system in 2006 developed a web-based rare disease registry, to monitor and profile patients affected by rare diseases. Clinicians belonging to regional validated healthcare data centres can register patients at the diagnosis, with a rigorous approach and distinguishing the types of amyloidosis, i.e., ATTRwt versus ATTRv. Thanks to this data collection method, available from July 2006 and extended with electronic therapy plans related to a diagnosis since May 2017, we analysed prevalence and incidence of ATTR and its subtypes. On November 30th 2022, ATTRwt prevalence in Tuscany is 90.3 per 1,000,000 persons and ATTRv prevalence is 9.5 per 1,000,000 persons, whereas the annual incidence ranges from 14.4 to 26.7 per 1,000,000 persons and from 0.8 to 2.7 per 1,000,000 persons, respectively. The male gender is predominant in both forms. All except one patient showed evidence of cardiomyopathy. This epidemiological data requires attention, not only to increase the effort for the clinical management and earlier diagnosis, but also to underline the need for the disease-specific treatments. • The exact prevalence transthyretin amyloidosis (ATTR) is still largely unknown. • Available data so far are mostly drawn from insurance claims or registries. • Prevalence of ATTRwt and ATTRv is 90.3 and 9.5 per million based on data from a Tuscany registry of rare diseases. • ATTR is increasingly being recognized and is less uncommon than previously believed. [ABSTRACT FROM AUTHOR]

Deploying data centers in the waters near coastal cities can greatly reduce the distance between data and users, not only eliminating the need to occupy land resources, but also saving energy consumption, which is a completely green and sustainable data center solution. Under the background of resource conservation and green development, this paper puts forward the project of undersea data center heat dissipation. Through the establishment of multiple linear regression model, the influence of tidal and seasonal changes on the heat dissipation of undersea data center is analyzed. Finally, it is concluded that the seasonal changes have the greatest impact on the heat dissipation of the undersea data center container. In the future, the placement strategy of data center and the methods to improve the heat dissipation efficiency will be further discussed. [ABSTRACT FROM AUTHOR]

Personalised Integrated Healthcare Monitoring System (PIHMS) has received major attention from the healthcare industry. With large population and strong growth of Indonesian middle class, PIHMS is certainly a big market for major healthcare providers. Although Indonesia has strong growth of middle-class, ironically, it is not supported with adequate physician ratio. Indonesian Ministry of Health (MoH) reported in 2020 the ratio of physicians per 1.000 patients is 0,62, far below WHO standard. As consequence, many Indonesian patients prefer to go abroad to have better healthcare services. The article supports the government's initiative to develop PIHM design by implementing IoT devices and blockchain networks. The use of PIHMS is essential to reach large patients that spread in remote areas, that suits to Indonesian condition. The article proposes PIHM based on blockchain solution (MedDApp) that enables to utilise IoT devices to carry out remote monitoring and medical consultation. The article examines the use of blockchain architecture to support developing data interchange between patients, healthcare providers and third parties such insurances and ministry of health (MoH). Data gatherings from respective stakeholders from Data and Information Centre division (Pusdatin), MoH and state-own hospitals to support developing smart health care architecture (MedDApp). The MedDApp design is expected to provide secure access between patients with healthcare professionals and third parties (national insurance companies and researchers in MoH). The outcome of the article is expected to be used as a reference for major healthcare providers to provide better services for their large patients. [ABSTRACT FROM AUTHOR]

• The novel two-path thermal energy storage with fin-enhanced PCM is designed and manufactured. • The thermal performance and temperature profile are experimentally investigated. • Heat transfer rate and accumulated thermal energy are fitted by test data. • Thermal environment during emergency cooling is analyzed with different inlet condition of CRAC. Emergency cooling system is necessary to guarantee high security for data center. Thermal energy storage is one of the feasible methods for emergency cooling with quick response ability to supply higher thermal safety. To simplify the traditional thermal energy storage system and improve its heat transfer efficiency, the novel two-path latent thermal energy storage unit developed for emergency cooling is experimentally studied. The effects of inlet temperature, flow rate on thermal performance of LTES, including heat transfer rate, accumulated energy, effectiveness and overall heat transfer coefficient, are investigated during both charging and discharging processes. Results from the experimental analysis show that the released energy of system could reach up to 2.9 MJ with the average heat transfer rate 1575W-2286 W. The overall heat transfer coefficient ranges from about 297–660 W/K when the flow rate of HTF varied from 12 to 16 L/min. The experimental LTES is then studied integrated with CRAC for emergency cooling. For air flow rate higher than 300 m3/h, and thermal load lower than 1800 W, the LTES meets 15 min of emergency cooling. The inlet and outlet air temperature of CRAC are successfully controlled under 40.5 °C and 23.9 °C. The newly designed LTES shows good thermal performance, and provides more flexibility and adaptability for real application in data center. [ABSTRACT FROM AUTHOR]

Predictive maintenance is gradually replacing conventional preventive maintenance, through an informed decision-making process for a fleet operator to proactively monitor the health status of the sea-based equipment and machinery. As a result, the fleet operator will only need to ask for on-time repair and maintenance services instead of periodic maintenance, which saves both time and cost significantly. While moving into this new maritime business model, which is also known as Service-as-a-Product (SaaP), both fleet operators and maintenance, repair, and overhaul (MRO) companies could also build a stronger collaboration in the maritime industries. With the advent of digital transformation towards Industry 4.0 (I4.0), the Industrial Internet-of-Things (IIoT) renders the massive collection of operational and machinery process data from the vessel equipment through sensorization, where these collected real-time information are useful for advanced analytics to predict equipment failure and to avoid unplanned downtime. Unlike other industries, there are a few challenges in the maritime industries when developing a centralized smart vessel equipment monitoring platform. One of the key challenges, is the lack of a feasible data management system, requires the centralized host must be able to handle and manage the telemetry IoT data that is transmitted over the satellite communication, by complying with all cybersecurity considerations and regulations in maritime sector. In this paper, we present a scalable, hybrid cloud-based data management framework that can connect to multiple edge systems where each system is deployed on a physical vessel, to establish a scalable SaaP business model. Hence, the ship-to-shore sensorized data can be processed and monitored via a real-time visualization dashboard through the centralized cloud-based platform. We argue that our approach, would be a turn-key solution that can be implemented for mostly all types of marine equipment and machinery, and thus to further improve the prediction tools to support advanced decision-making techniques, such as optimal time repair of vessel equipment. [ABSTRACT FROM AUTHOR]

In data center (DC) environments, machine learning algorithms play an important role in resource management to increase efficiency by means of proper predictive monitoring workload trends and adjusting jobs accordingly. In this paper, we propose a system to predict the CPU usage of virtual machines (VMs) of a DC. Our proposal performs clustering of VMs based on their historical information (i.e., time series) by evaluating several traditional ML algorithms using common statistical features of VM time series, which facilitates grouping VMs with similar behaviors and establishing clusters based on these features. Then, training of representative models is performed to finally choose the one with the lowest mean error per cluster. The simulation results show that by performing clustering and training the model with representative time series, it is indeed possible to obtain a low mean error while reducing the local training time per individual VM. [ABSTRACT FROM AUTHOR]

With the development of the information and communication technology (ICT) industry, the energy consumption of data centers is continuously increasing. High-temperature polymer electrolyte fuel cells (HT-PEFC) have a high operating temperature of 120 °C or higher; thus, the heat generated from fuel cells stack is effectively used as a heat source for a absorption refrigerator (AR) to generate cooling. The combined cooling and power (CCP) system is proposed to satisfy the energy demand of data centers that require both power and cooling. The CCP system comprises an HT-PEFC and a double-effect AR that recovers the wasted heat of the fuel cell stack. As a result of analyzing the electrical and cooling efficiency of the CCP system according to the fuel cell operating conditions, the overall efficiency increased to 95%, which was significantly higher than the existing system. Based on the simulation of the developed model, coefficient of performance (COP) and cooling capacity depending on the temperature changes of chilled water and cooling water were calculated within the stack temperature range of 150–180 °C, and the COP changed from 1.1 to 1.58 depending on the conditions. The developed CCP system model can be used to plan strategies for flexible fuel cell operation according to the power and cooling demands required in the data center. • The combined cooling and power (CCP) system to meet energy demand of data centers. • The CCP system comprises an HT-PEFC and a double-effect absorption refrigerator. • The electric and cooling efficiency of the proposed CCP system were analyzed. [ABSTRACT FROM AUTHOR]

With the development of Global Navigation Satellite System (GNSS), the detection of precipitable water vapor (PWV) using the GNSS atmospheric sounding technique becomes a research interest in GNSS meteorology. In the conversion of zenith tropospheric delay (ZTD) to PWV, the weighted mean temperature (Tm) plays a crucial role. Generally, the Tm estimated by the linear regression models based on surface temperature (Ts) cannot meet the requirement for global use, and the accuracy of Tm derived from the empirical models is limited. In this study, a new Tm model, named GGTm-Ts model, was developed using the global geodetic observing system (GGOS) atmosphere Tm data and European Centre for Medium-Range Weather Forecasts (ECMWF) data from 2011 to 2015. Resting upon a global 2.5°*2° grid of coefficients of Tm-Ts linear function, the new model can provide Tm at any site in two modes, one for the case with measured Ts provided, i.e., the accurate mode, the other for the case that Ts provided by a subroutine, i.e., the normal mode. The performance of GGTm-Ts model was assessed against the Bevis formula, GPT2w and GPT2wh model using different data sources in 2016-the GGOS atmosphere and radiosonde data. The results show that the GGTm-Ts model in accurate mode achieves best performance with an improvement of 46.9 %/15.3 %, 37.8 %/19.5 % and 34.4 %/14.2 % over other three models in the GGOS atmosphere/radiosonde comparison. For the normal mode, the GGTm-Ts model outperforms the GPT2w model and achieves equivalence results with the GPT2wh model. Moreover, the impact of Tm on GNSS-PWV was analyzed to validate the performance of the GGTm-Ts model. [ABSTRACT FROM AUTHOR]

The line graphs of some well-known graphs, such as the generalized hypercube and the crossed cube, have been adopted for the logic graphs of data center networks (DCNs). Conditional bijective connection networks (conditional BC networks) are a class of networks which have been proved to include hypercubes, crossed cubes, locally twisted cubes and Möbius cubes, etc. Hence, the researches on the line graphs of conditional BC networks can be applied to DCNs. Edge independent spanning trees (EISTs) on a graph have received extensive attention because of their applications in reliable communication, fault-tolerant broadcasting and secure message distribution. Since the line graph of an n -dimensional conditional BC network, denoted as L (X C n) , is (2 n − 2) -regular, whether there exist 2 n − 2 EISTs on L (X C n) is an open question. In this paper, we first propose a parallel algorithm to construct 2 n − 2 EISTs rooted at an arbitrary node on L (X C n) , where n ≥ 2. Then, we prove the correctness of our algorithm. Finally, we give a simulation result of our algorithm. [ABSTRACT FROM AUTHOR]

The energy consumption of data centers (DCs) is on a sharp upward trend in recent years. DCs are playing an increasingly important role in demand response (DR) programs. However, the reassignment of computing tasks among DCs leads to different energy demands of different DCs. Given that the investment cost of energy storage is high, this work proposes a shared energy storage business model for the DC cluster (DCC) to improve economic benefits and promote renewable energy accommodation. Besides, an internal energy balance mechanism is set up to make full use of the complementary energy consumption characteristics of different DCs. Considering the renewable energy uncertainty, an optimization model is proposed based on the chance-constrained goal programming (CCGP). Finally, simulation results prove that the proposed energy storage business model has a positive effect on improving the economic benefits of the DCC. It also proves that for a DCC adopting the proposed internal energy balance mechanism, its total renting power can be effectively reduced and renewable energy consumption can be greatly promoted. [ABSTRACT FROM AUTHOR]

The world's energy consumption shows an increasing trend. Unfortunately, it is still dominated by the use of fossil energy. This condition results in concerns that an energy crisis will occur due to running out of non-renewable energy and an environmental crisis of global warming. One of the technology sectors that consumes a lot of energy is computer servers and data centers. This technology sector is growing rapidly due to the increasingly complex computing and larger data. One of the main causes of inefficiency in this technology is a decrease in performance due to an increase in temperature. Therefore, effective and efficient technology is needed to overcome the increasing operating temperature of computer servers and data centers. One of the most prominent cooling technologies to solve this problem is immersion colling. This method has developed in various types with their respective advantages and disadvantages according to application needs. Therefore, review literature is needed to comprehensively discuss the development of immersion cooling technology from the past until now and projected developments in the future. In more detail, this paper comprehensively compiles the latest findings of immersion cooling technology which includes an overview of the cooling system, history, implementation, construction type, and advantages. [ABSTRACT FROM AUTHOR]

Graphic processing units (GPUs) are becoming gradually popular in large-scale data centers and cloud computing. Sharing a GPU across different applications is an important feature in these platforms to improve resource utilization and system productivity. However, in situations where GPUs are competitively shared, some challenges arise. In this paper, concurrent kernel execution and interference detection are investigated using Extreme Gradient Boosting, Convolutional Neural Network, Gated Recurrent Unit, and Conditional Generative Adversarial Network learning techniques on Tesla P100 and RTX-2080 GPU architectures. The experimental results obtained by applying the deep learning models on the dataset, which consists of four GPU resources: blocks per grid, threads per block, number of registers, and shared memory, show that the Extreme Gradient Boosting and Multi-Channel Convolutional Neural Network models have the promising capability of concurrent kernel execution classification and interference detection. In comparison to the existing work, the proposed models investigated using real and synthesized datasets show clear outperformance and generalization capability in terms of recall and precision. [ABSTRACT FROM AUTHOR]

• OS of larynx cancer patients treated at Odense DK, Manchester UK and Liverpool AUS is driven by tumour volume and PS. • Federated learning can be used to create survival models without patient-sensitive data leaving the individual institutions. • The baseline hazards of the three institutions are similar, indicating the GTV volume and PS explain the cohort differences. • Smoking during treatment has the same risk profile as a ten year older patient not smoking. Federated learning has the potential to perfrom analysis on decentralised data; however, there are some obstacles to survival analyses as there is a risk of data leakage. This study demonstrates how to perform a stratified Cox regression survival analysis specifically designed to avoid data leakage using federated learning on larynx cancer patients from centres in three different countries. Data were obtained from 1821 larynx cancer patients treated with radiotherapy in three centres. Tumour volume was available for all 786 of the included patients. Parameter selection among eleven clinical and radiotherapy parameters were performed using best subset selection and cross-validation through the federated learning system, AusCAT. After parameter selection, β regression coefficients were estimated using bootstrap. Calibration plots were generated at 2 and 5-years survival, and inner and outer risk groups' Kaplan-Meier curves were compared to the Cox model prediction. The best performing Cox model included log(GTV), performance status, age, smoking, haemoglobin and N-classification; however, the simplest model with similar statistical prediction power included log(GTV) and performance status only. The Harrell C-indices for the simplest model were for Odense, Christie and Liverpool 0.75[0.71–0.78], 0.65[0.59–0.71], and 0.69[0.59–0.77], respectively. The values are slightly higher for the full model with C-index 0.77[0.74–0.80], 0.67[0.62–0.73] and 0.71[0.61–0.80], respectively. Smoking during treatment has the same hazard as a ten-years older nonsmoking patient. Without any patient-specific data leaving the hospitals, a stratified Cox regression model based on data from centres in three countries was developed without data leakage risks. The overall survival model is primarily driven by tumour volume and performance status. [ABSTRACT FROM AUTHOR]

Traditional privacy-preserving technologies have been unable to provide adequate protection and are vulnerable to background knowledge. MapReduce is parallel distributed computing model which has the advantages of good scalability and high fault tolerance. In this paper, we design a differential privacy protection based on MapReduce for the security problems faced in the distributed environment. Firstly, we propose a random forest algorithm DPMRRF that satisfies differential privacy under the MapReduce framework. Secondly, we use the MapReduce framework in the Hadoop platform to make the classification results satisfy differential privacy by adding random noise values to the leaf nodes. Finally, through the design of the feature selection scheme combined with the index mechanism, three different privacy budget allocation methods are adopted, namely, equal distribution, equal distribution and equal distribution. Experiments show that the algorithm has better classification accuracy while ensuring data availability. The proposed algorithm effectively improves the classification accuracy and operating efficiency, and greatly reduces the amount of computation. [ABSTRACT FROM AUTHOR]

Aim: According to the current resuscitation guidelines, the use of mechanical chest compression devices could be considered under special circumstances like transport with ongoing resuscitation or long-term resuscitation. The aim of this study was to investigate whether survival is improved using mechanical devices under such circumstances.Methods: Out-of-hospital cardiac arrests from all high-quality data centres of the German Resuscitation Registry from 2007 to 2020 were investigated. The use of mechanical devices was compared separately for transport with ongoing resuscitation, prolonged resuscitation (>45 min), and resuscitation with fibrinolytic agents applied. Baseline characteristics, 30-day survival/discharged alive, and neurological function at discharge were analysed descriptively; and 30-day survival/discharged alive was additionally analysed using multivariate logistic regression.Results: Overall, patients who were treated with a mechanical device tended to be younger and were significantly more likely to have a witnessed cardiac arrest and a shockable initial rhythm. During the study period, 4,851 patients were transported to hospital with ongoing resuscitation (devices used in 44.2%). The 30-day survival was equal (odds ratio, OR: 1.13, 95%-CI: 0.79-1.60). In 3,920 cases, a resuscitation duration > 45 min was documented (9.5% with device). When a device was used, 30-day survival was significantly increased (OR 2.33, 95%-CI: 1.30-4.15). Fibrinolytic agents were used in 2,106 patients (22.2% with device). Here, 30-day survival was significantly worse with a device (OR: 0.52, 95%-CI: 0.30-0.91).Conclusion: Mechanical devices are not associated with better survival when used during transport, but rescuer safety could still be an important argument for their use. Devices are associated with better survival in prolonged resuscitation, but worse survival when a fibrinolytic was used. [ABSTRACT FROM AUTHOR]

The ever growing Internet of Things (IoT) paradigm aims to put people's incorporation into the new phase of connectivity and sensor technology. While, IoT has the potential to deliver new value-added services in order to make life easier and healthier for people, there are still several issues to be addressed in order to harness the widespread dissemination and adoption of the IoT paradigm, considering its potential benefits. In this context, considering an IoT ecosystem holistically i.e. end-to-end which includes the proprietary Operational Technology (OT) comprising of software and hardware to monitor, detect and control the equipment through sensors and actuators and the general Information Technology (IT) which comprises of the data center infrastructure catering to the backend needs of IoT such as compute, storage and network, one major challenge is converging OT with IT. Such an OT/IT convergence has the potential to explore many problems that exist in the end-to-end IoT ecosystem. One such problem is related to drawing actionable insights through operational analytics. The data center paradigm which concerns this research is the Software-Defined Data Center (SDDC) which is touted as the preferred data center paradigm for IoT. While most of the research happening on IoT today deals with the data from the IoT applications and sensors, this research focuses on the plethora of the metadata contained in the end-to-end IoT ecosystem. This research demonstrates the modelling of the end-to-end IoT ecosystem using semantic-based approach such as ontology. We arrive at a cohesive unified ontology unifying the OT and the IT constituents of the IoT fabric with SDDC together with rich context aware attributes embedded in the unified ontology, thereby addressing the problem of OT/IT convergence as well as laying the foundation to explore many solutions to problems pertinent in such a converged OT/IT ecosystem. [ABSTRACT FROM AUTHOR]

The rapid growth in the cloud data center needs a dynamic resource provision to maintain the Quality of Services parameters. To guarantee it, Virtual Machine Migration as part of VM Consolidation has a significant role. Efficient VM migration requires knowledge of the host's future utilization in advance. Because of the high variation in cloud resource usage and dynamic workloads, predicting host utilization using utilization history is challenging. This paper proposes a Support Vector Regression-based methodology to predict a host's future utilization using multiple resource's utilization history. A Hybrid Kernel function that includes radial basis function and polynomial kernel function has been proposed and then trains the Support Vector Machine using multiple-resource utilization history. Compared to the existing approaches: multiple linear regression-based prediction, Euclidean distance, and Absolute Summation based regression, the proposed method performs better in terms of root mean square error, mean square error, mean absolute percentage error, mean absolute error, and R 2 . The result section concludes that on evaluating error percent, the prediction error is 16% for the proposed approach and predicts host utilization with 7%, 64%, and 67% more accuracy than MRHOD, MDRHU-ED, MDRHU-AS approaches, respectively. [ABSTRACT FROM AUTHOR]

A graph G is two-disjoint-cycle-cover [ r 1 , r 2 ] -pancyclic if for any integer ℓ satisfying r 1 ≤ ℓ ≤ r 2 , there exist two vertex-disjoint cycles C 1 and C 2 in G such that | V (C 1) | = ℓ and | V (C 2) | = | V (G) | − ℓ. The k -dimensional data center network DCell with n port switches and order t k , n , denoted by D k , n , plays an important role in big data cloud computing. In this paper, we study the structural properties of D k , n and show that D k , n is two-disjoint-cycle-cover [ 3 , ⌊ t k , n 2 ⌋ ] -pancyclic for n ≥ 7 and n + k ≥ 11 , that is, for any integer ℓ satisfying 3 ≤ ℓ ≤ ⌊ t k , n 2 ⌋ , there exist two vertex-disjoint cycles C 1 and C 2 such that | V (C 1) | = ℓ and | V (C 2) | = t k , n − ℓ in D k , n. This paper generalizes the results about the Hamiltonicity in (Wang et al., 2015 [25]) and the pancyclicity in (Hao and Tian, 2021 [11]). • It is the first study to consider two-disjoint-cycle-cover pancyclicity in data center networks. • The two-disjoint-cycle-cover pancyclicity of data center network DCell is shown. • This result extends and generalizes the Hamiltonicity and bipancyclicity of DCell. [ABSTRACT FROM AUTHOR]

• A hybrid-integrated 400 GbE ROSA using silica-based AWG and InP-based PD arrays is fabricated, which performs satisfying coupling efficiency between AWG and PD arrays, small signal O/E responses, eye diagrams and sensitivities. The fabricated ROSA has promising prospects to support higher transmission data rate by simply replacing the PD arrays and TIA arrays with higher-speed types. • In this work, a total reflection surface at the end facet of the AWG is used to realize a 90 degrees shift in the optical path, and realize a convenient hybrid integration method. This structure is designed and optimized using FDTD. • Eight-channel O-band LWDM AWG with an improved Rowland circle is designed and fabricated, which performs flat-top spectrum, satisfying uniformity and polarization insensitivity. A silica-based hybrid-integrated receiver optical subassembly (ROSA) for 400 Gbps Ethernet (400GbE) is demonstrated, which consists of an 8-channel silica-based arrayed waveguide grating (AWG), two 4-channel InP-based surface-illuminated photodetector (PD) arrays, two 4-channel transimpedance amplifier (TIA) arrays, and a high-frequency flexible printed circuit (FPC). All lanes of the ROSA exhibit responsivities of over 0.539 A/W, 3-dB bandwidths of greater than 30.86 GHz, and receiver sensitivities of less than −11.7 dBm. The ROSA has achieved a transmission rate of 8 × 50 Gbps for data center interconnect. [ABSTRACT FROM AUTHOR]

In this paper, the dynamic performance of Proton Exchange Membrane Fuel Cell-based Combined Cooling Heating and Power (PEMFC-based CCHP) system combined with a data center real-time load is studied. A novel synergistic control method is proposed to achieve optimization of energy flow between system units under fluctuating load. First, the off-design working condition models of PEMFC-based CCHP system are constructed. The delay in the changes of data center temperature is considered because of the thermal inertia under dynamic conditions. Second, the dynamic performance of PEMFC-based CCHP system to step changes of data center load is analyzed. Third, to further reveal the energy flow between units, the units output under varying load conditions are compared under the constraints of power and refrigeration supply. Finally, the dynamic performance of PEMFC-based CCHP system under the traditional control method of following electric load are revealed. A novel synergistic control method for the PEMFC-based CCHP system is proposed by refining the traditional method to optimize the energy flow between units. The result presents that the average energy efficiency reaches 78.29 %. Compared with the following electric load method, the energy efficiency is increased by 6.02 %. The response time of synergistic control method is 27.0 ms. Improved energy efficiency and fast response have been achieved. • The off-design working condition model of PEMFC-based CCHP system is constructed. • The step response characteristic of PEMFC-based CCHP system is analyzed. • Full working condition performances of PEMFC-based CCHP system are studied. • A novel synergistic control method for the PEMFC-based CCHP system is proposed. [ABSTRACT FROM AUTHOR]

Utilizing the thermal stratification of reservoirs to obtain cold water for cooling green data centers (GDCs) is a new mode of energy conservation and emission reduction. However, global warming is expected to alter this phenomenon in deep reservoirs and thus may affect the digitalization process. While the frequency and intensity of extreme climate (EC) events are increasing under a changing climate, the impacts of these highly destructive events on thermal stratification and the related cascading effects on GDC cold-water supplies are still unclear. A two-lateral-dimensional thermohydrodynamic model was established to determine the characteristics of reservoir thermal stratification changes and its potential water environment impacts based on the heat extreme experienced by the Jinshuitan Reservoir, a large, pumped storage power station (PSPS) with a GDC in southeastern China. Fourteen different PSPS inlet/outlet schemes under 2 climate scenarios were designed, and two new indexes, namely, the risk index of cold-water shortage and the index of thermal stratification, were proposed to explore the impact of extreme climates on GDC water extraction. The results showed that compared to the present climate conditions, the reservoir experienced stronger thermal stratification with a 20.6 % average increase in thermal stability, a longer thermal stratification duration with a transitional trend from seasonal to permanent stratification, and a thinner cold-water layer with a 15-m average decrease. Changes in the thermal stratification of reservoirs under EC conditions will benefit some species, such as diatoms (except for early July and the period from mid-August to early October), green algae and cyanobacteria and the migration of warm-water fish, but adversely impact dissolved oxygen, diatoms and cold-water fish. Furthermore, the risk of cold-water shortages under EC conditions was consistently greater than that under average climate (AC) conditions (traditional assessment) when the elevation of pumped storage intake/outlet was less than 145 m. According to the traditional assessment method, the intake/outlet elevation of pumped storage reservoirs resulted in a 60.27 % increase in the risk of cold-water shortages during climate extremes. Furthermore, the demand for green cold-water withdrawal and optimal pumped storage power generation benefits were met when the PSPS intake/outlet elevation was 145 m. Beyond solving this specific case, this study elucidates the importance of linking the thermal stratification and data center of cold-water withdrawal to ECs. These findings provide new insights for increasing climate change resilience. [Display omitted] [ABSTRACT FROM AUTHOR]

• Establishing an energy-aware data center loading model considering three scheduling methods. • Analysing the impact of the coupling relationship among the three scheduling methods. • Proposing a coordinated optimization model for the electricity consumption costs of data centers. • An incentive profit-sharing mechanism based on Nash Bargaining theory is designed. • Contribution ratios of each data center in the coordination for cost reduction are considered. As digital technology continues to advance rapidly, the substantial operational electricity costs associated with the extensive deployment of globally distributed internet data centers have become a significant concern. Addressing this challenging issue is crucial for the future expansion of internet data centers. In this paper, a coordinated strategy is proposed that leverages the spatial–temporal flexibility of internet data centers and their interaction with the power system. Firstly, an energy-aware internet data center loading model is introduced that incorporates multiple coupled scheduling methods to maximize the utilization of internet data centers' flexibility potential. Secondly, the proposed internet data centers loading model is integrated into a coordinated optimization model aimed at achieving higher profits through cost reduction in response to varying electricity prices across multiple regions. To facilitate the coordinated operation of geo-graphically internet data centers, an incentive profit-sharing mechanism based on Nash Bargaining theory is designed to fairly distribute coordination profits based on the contribution ratio of each data center in coordination. Through comprehensive case studies utilizing real-world datasets, the results demonstrate that the internet data center loading model, considering multiple coupled scheduling methods, can enhance internet data centers' flexibility potential, resulting in a significant 11.73% reduction in total cost, with the coupling relationship accounting for 5.05% of this reduction. Furthermore, the proposed profit-sharing mechanism effectively fosters coordination and enthusiasm among geographically distributed internet data centers. [ABSTRACT FROM AUTHOR]

• Developing a new water free-cooling system and its control schemes models. • Control performance quantification of free-cooling systems in three operating modes. • Evaluating energy performance and ecological benefit of economizers in worldwide climates. • Quantifying effect of space air temperature increase on energy efficiency. Water-side economizer (WSE) is an effective free-cooling solution aiming to enhance the natural cooling duration and in turn reduce the cooling-energy consumption of data centers. However, the system's performance is constrained by the limited understanding of its control mechanisms and the influence of local climatic conditions on its free-cooling capabilities. To overcome these limitations, a thermodynamic model of the WSE system with control strategies is provided and its adaptability performance in diverse worldwide locations is systematically evaluated. Modelica-based models of chiller plants and control strategies are firstly developed, and then the energy consumption and ecological benefit of the WSE system are investigated via large-scale simulations in worldwide climates. The performance indicators and application potential of the system are demonstrated through visualization maps. The analysis reveals significant energy savings potential of the WSE system, with the capacity to reduce annual cooling energy consumption of data centers by 4.4–33.7 % in hot to marine zones, and even by a remarkable 62.6 % in plateaus and cold zones. Furthermore, it is observed that every 1 °C increase in outdoor air temperature can reduce energy efficiency ratio by 1.3–8.7 % while simultaneously increase P U E by 0.3–1.0 %. The ecological benefit results revealed that the WSE system can effectively harness larger quantities of stable free-cooling resources in mixed and cool climate zones, exceeding its capabilities in hot and warm regions. These results offer valuable insights for the impact of climate on the energy and ecological benefits of WSE system for data center, which can be helpful to guide the design or retrofit of such cooling systems in specific locations, presenting an immense opportunity to maximize cooling-energy savings. [ABSTRACT FROM AUTHOR]

As the key component of uninterruptible power supply, the anomaly detection of valve-regulated lead-acid (VRLA) batteries is the reliable guarantee for the operation of internet data centers. Although deep neural networks (DNNs) have been investigated in battery anomaly detection, the key problem of feature learning under non-ideal conditions, i.e., data scarcity and class-imbalance, is not addressed very well. In this article, a novel DNN called multi-source self-supervised domain adaptation network (MSSSDAN) is proposed for VRLA battery anomaly detection of data center under non-ideal conditions. Firstly, multi-dimensional collaborative self-supervised learning (MDCSSL) is proposed to address the class-imbalanced problem by performing auxiliary feature learning task considering the temporal dependency of different channels. Secondly, a novel transfer learning (TL) framework, multi-source adaptive transfer learning (MSATL) is developed to solve the domain shift problem by adaptively minimizing the distribution discrepancy of different domains. Finally, the outperformance of MSSSDAN is verified on the VRLA battery dataset collected from Tencent data center. The testing results demonstrate that MSSSDAN can effectively solve the problem of battery anomaly detection under non-ideal conditions. • Multi-source self-supervised domain adaptation network (MSSSDAN) is proposed for non-ideal VRLA battery anomaly detection. • Multi-dimensional collaborative self-supervised learning is proposed to address the class-imbalance problem. • Multi-source adaptive transfer learning framework is proposed for solving the problem of data scarcity and domain shift. • The experimental results prove the effectiveness of MSSSDAN for VRLA battery anomaly detection. [ABSTRACT FROM AUTHOR]

• The differences and similarities between pump- and gravity-driven modes are studied. • The mechanism of the flood limit for the pump-driven mode is illustrated. • A new correlation based on the experimental results has been proposed. With the development of big data and cloud computing, data centers have grown rapidly worldwide. A data center is a special building with high energy consumption. The energy consumption of traditional computer room air conditioning is assumed to be 45 % of the total energy consumption of the data center. Currently, the use of a loop thermosiphon as a cooling system is one of the most effective methods. Although loop thermosiphons have already been widely used, the boiling mechanism inside the evaporator needs to be studied further, particularly for pump-driven loop thermosiphons. In this study, we investigated the heat transfer and flow characteristics of a pump-driven loop thermosiphon. The inside tube diameter is 8.05 mm, the heat flux changes from 6317 to 28085 W/m2, and the mass flux changes from 437 to 1311 kg/m2s. First, the differences and similarities between pump- and gravity-driven modes are studied. Second, the matching relationship between the heat flux and mass flow rate is studied, and the mechanism of the flood limit for the pump-driven mode is illustrated. Finally, we compare different boiling correlations and propose a new correlation based on our experimental results. [ABSTRACT FROM AUTHOR]

Regional data center clusters have flourished in recent years to serve customers in a major city with low latency. The optimal coordination of data centers in a regional cluster has become a pressing issue because of their rising energy consumptions. In this paper, an online algorithm based on Lyapunov optimization framework is developed for the combined computing workload and energy coordination of data centers in a regional cluster. The proposed online algorithm is prediction-free and easy to implement. We prove that the workload queues and battery energy level will be within their physical limits, though their related time-coupling constraints are not considered explicitly in the proposed algorithm. The previous online algorithms do not have such a guarantee. A theoretical upper bound on the optimality gap between the online and offline results is derived to provide a performance guarantee for the proposed algorithm. To enable distributed implementation, an accelerated distributed coordination algorithm is developed based on the alternating direction method of multipliers (ADMM) with iteration truncation and follow-up well-designed adjustments, whereby a nearly optimal solution is attained with much enhanced computational efficiency. Case studies show that the proposed algorithm reduces the operational costs and saves computation time compared to online benchmarks. • The combined computing workload and energy coordination of data centers is studied. • A prediction-free online coordination algorithm is built with proven properties. • An accelerated ADMM algorithm is developed for distributed implementation. • The proposed coordination method greatly reduces the total cost of data centers. • The computation time is largely reduced compared to the traditional ADMM. [ABSTRACT FROM AUTHOR]

Due to the electrification and digitization of things such as public services, communication systems, and transport systems, data centers have become part of the critical infrastructure of today's society. The data centers as a resource-intensive industry, perturbates the energy and spatial systems where established. Currently, municipalities as the main entities responsible for local resource management, often lack the power and/or determination to plan spatial and energy aspects cohesively. Consequently, municipalities encounter resource management issues when planning for data center locations; high resource demand (water, electricity, land) and high resource waste (excess heat). This research aims to support urban planners, municipal energy companies, politicians, and other relevant decision-makers, in developing municipal energy and spatial systems that account for data centers' resource demand as well as using their resources produced in the form of excess heat for the community's benefit. This study compares energy and spatial aspects in relevant data center locations in Stockholm and Luleå municipalities from an urban planning perspective. The results suggest that municipalities' energy and spatial systems can increase their adaptive capacity towards data center establishment by considering certain resource planning determinants: electricity and water infrastructure availability, heat producers and users' diversity, land use diversity, and property fragmentation. • Luleå and Stockholm municipalities approach planning data center locations in diverse ways, partly due to their climate or energy availability differences, which influence local resource planning and management. • Stockholm municipality has a higher adaptive capacity than Luleå municipality, being able to adapt its energy and spatial system to integrate the data center industry and taking the opportunity to build resiliency, by using the data centers' excess heat for the local community. • An increased collaboration between diverse interest groups could allow a more cohesive spatial and energy planning. • This cohesive planning might set the critical conditions (electricity and water infrastructure availability, heat producers and users' diversity, land use diversity, property fragmentation) to facilitate a more effective integration of such resource-intensive industries within a municipality. [ABSTRACT FROM AUTHOR]

• Study feasibility of a Grid-Connected PV power supply system in data center's cooling system. • With storage batteries, utilization of PV power increased by up to 27.64 % in the discussed days. • Optimized number of battery groups and PV panels were given to reduce carbon emission. • Find threshold of battery's unit price for the adoption of it to maximize economy performance. • Provide guidance for design and operation of such system with q IT between 500 and 1100 W/m2. The cooling system of a data center accounts for a significant part of its energy consumption, and the adoption of solar energy can reduce its power demand from the grid. This paper investigated the optimal configuration of a grid-connected PV power supply system to a data center's centralized water-cooling system. Firstly, mathematical models for photovoltaic panels and storage batteries were established. Then, two operating strategies were proposed, respectively, for two systems with and without storage batteries, and the supply–demand matching performances were studied. With storage batteries, mismatch problem between electricity generated by photovoltaic panels and electricity consumed by the water-cooling system can be significantly improved. Utilization ratio of electricity generated by photovoltaic panels was increased by up to 27.64 % in the discussed typical days. Annual utilization of electricity generated by photovoltaic panels can also be significantly increased, especially when heat dissipation density is small. Lastly, the optimal configurations were discussed. To reduce carbon emission, number of battery groups were recommended for different heat dissipation density and number of photovoltaic panels. Full life-cycle carbon reduction ranged from 1.77 to 3.71 tCO 2 /m2 and carbon emissions reduced from 14.23 to 62.14 % as compared with the traditional system. As for the economic performance, the recommended number of photovoltaic panels are 798, 1330 and 1589 for heat dissipation density being 500, 800 and 1100 W/m2, respectively. While the recommended number of batteries was 0 if unit price of batteries is higher than 11.1 ten thousand yuan/group. [ABSTRACT FROM AUTHOR]

To augment situational awareness capabilities for high-frequency disturbances, the modern power system necessitates the implementation of high-sampling rates for real-time Synchro-Waveform (SW) measurements. Nevertheless, the nonlinear nature of data collected from SW measurement devices poses significant challenges to the communication and storage capacities of data centers. To effectively tackle this issue, this paper proposes a Multi-stage Hybrid Coding (MHC) method for the lossless compression of SW data. The MHC method utilizes a well-structured two-stage approach to achieve efficient compression. In the first stage, a variable frame-based method is designed to reduce redundant information and achieve initial compression using multiple periodic recursive compression. Subsequently, the dictionary-based method, specifically the Lempel–Ziv–Markov chain algorithm, is seamlessly integrated to further compress the SW data. Experimental and numerical analyses are conducted using both simulated and real-world source data, with real-time performance validation in SW measurement units. The obtained results convincingly demonstrate that the proposed method enables online compression of both event and non-event SW data, achieving an impressive reduction of over 73.4% in data space, while maintaining a data loss ratio lower than 0.4% for high-density SW data. • Multiple periodic recursive method is proposed to eliminate repetitive information. • A multi-stage hybrid coding method is implemented for the lossless compression. • A series of synchro-waveform events are collected to evaluate the performance. [ABSTRACT FROM AUTHOR]

Highly dense and integrated data centers face key challenges of realizing efficient cooling and improved energy efficiency. To overcome these challenges, this study experimentally investigated the flow and heat transfer characteristics of single-phase immersion liquid cooling (SPILC) systems. The influence of two opposite coolant flow directions on the SPILC performance was examined. Furthermore, a correlation mechanism between coolant thermal properties and SPILC performance was established, along with a control chart for regulating the working conditions of SPILC systems. The results indicate that compared to a pro-gravity flow, an anti-gravity flow scheme reduces the chip case temperature and thermal resistance by 33.8% and 55.6%, respectively, while decreasing the power usage effectiveness (PUE) by 1.4%. Using coolant with the lowest viscosity reduces the chip case temperature and thermal resistance by 9.3% and 10.5%, respectively, while decreasing the PUE by 0.4%. Moreover, the cooling water temperature has a greater impact on the performance of SPILC systems than the volume flow rate of coolants. Additionally, this paper provides control charts for the cooling water temperature and coolant flow rate to improve the PUE while ensuring the safe operation of SPILC systems, with the highest chip temperature and total electricity consumption as indicators. • Immersion cooling performance of system-level data center servers is experimentally studied. • PUE of the anti-gravity flow decreases by 1.4% compared to the pro-gravity flow. • Correlation between coolant properties and immersion cooling performance is established. • Energy-efficient control charts for cooling water and immersion coolant are provided. [ABSTRACT FROM AUTHOR]

• Evaluate the impact of different cdfs on drought magnitude. • Global-scale analysis identifies 23 parametric cdfs as best-fit for streamflow values. • Streamflow simulation accuracy and reference period length influences cdf selection. • Opting for best-fitting parametric cdf mitigates overestimating drought magnitudes. Streamflow droughts pose significant challenges to water resource management and environmental sustainability. The accurate characterization of streamflow drought magnitude is essential for effective mitigation, assessment and adaptation strategies. In this study, we explore the implications of choosing different cumulative probability distribution functions (cdfs) on the determination of streamflow drought magnitude, as assessed by the z-score values of the standardized streamflow index (SSI) and return periods. Our investigation encompasses a comprehensive global-scale analysis, with the aim of identifying the most suitable cdfs for characterizing streamflow droughts. To accomplish this objective, we incorporate global simulated streamflow values from the model WaterGAP 2.2d and observed streamflow values from 280 Global Runoff Data Centre (GRDC) stations for the period 1981 – 2010. We evaluate the performance of 118 parametric cdfs for each of the 12 calendar monthly streamflow values at each model grid cell and GRDC station to determine the best fitting parametric cdf. Moreover, we go beyond the conventional scope of such studies by further assessing the best-fitting cdfs alongside non-parametric cdfs such as the Empirical cumulative distribution function (ecdf) and the cumulative distribution function derived from the Kernel density estimation function (kcdf). Among our findings, we establish a subset of 23 parametric cdfs that exhibit the best fit for the global streamflow values concerning all 12 calendar monthly streamflow values at individual GRDC stations or grid cells within the specified period. Furthermore, we observe that the accuracy of streamflow simulation and the length of the reference period directly influence the selection of the most suitable parametric cdf. Our comparative analysis reaffirms the existing understanding that opting for the best-fitting parametric cdf, rather than a non-parametric alternative, mitigates the risk of overestimating drought magnitudes. However, we emphasize the crucial importance of considering the intended purpose, approach-specific sensitivities, and uncertainties when choosing the most suitable parametric cdf for any calendar monthly streamflow values. [ABSTRACT FROM AUTHOR]

• Calibrated and rigorously validated model of silica gel-water adsorption chiller. • Screening of more than 40 adsorbents for application under data center conditions. • Review and evaluation of adsorption heat exchanger designs. • Cooling power of adsorption chiller can be increased by 59 % if MIL-100(Fe) is used. • (Partial) energy efficiency ratio (pumps and control, no fans) of > 20 is realistic. The electricity and water consumption of data centers is growing on a global scale. A shift towards liquid cooled racks in combination with thermally driven cooling can help to reduce the electricity and water demand associated with the necessary heat rejection. To quantify the potential of adsorptive heat transformation devices in reducing the electricity and water demand, the prediction of thermal efficiency, heat flow rates and energy efficiency ratio is required. To this end, a numerical model is newly developed using basic adsorption heat exchanger theory. This model can predict the necessary performance indicators with respect to temperatures and volume flow rates, heat exchanger design and adsorbent. The full performance map of a market-available adsorption chiller (71 points) and own measurements are used for calibration and rigorous validation of the model. An average deviation (experiment vs. calculation) of 8.3 % in terms of thermal efficiency and 7.2 % in terms of heat flow rates is achieved, indicating a very good agreement for a wide range of temperatures. At a moderate liquid cooled rack outlet temperature of 50 °C, a heat rejection temperature of 26 °C and a cold aisle inlet temperature of 18 °C the cooling power of the silica gel reference chiller of 5.3 kW can be increased by 59 % to 8.5 kW at a partial energy efficiency ratio (pumps and control, no fans) of > 20 by assuming MIL-100(Fe) as adsorbent on a flat-tube lamella heat exchanger. The model can be used in subsequent annual system simulations to quantify the savings in electrical power and water consumption, which strongly depend on the ambient conditions. [ABSTRACT FROM AUTHOR]

The error of regional models gradually increases during long-term integration, and it can be reduced with the constraint of large-scale circulation by assimilating global forecast system (GFS) fields to improve the forecast performance. However, with increasing forecast time, the GFS forecast error also significantly increases and varies widely. Directly assimilating GFS fields into the regional model without processing may generate negative impacts, whereas employing postprocessing correction methods can effectively mitigate GFS forecast errors. In this study, the GFS data of European Centre for Medium-Range Weather Forecasts (ECMWF) are first corrected using a convolutional long- and short-term memory network method. The correction model can correct multiple forecasts simultaneously by training only one model, and the correction results are less erroneous than those obtained by training multiple correction models for the different forecast times and pressure layers. The corrected ECMWF forecast fields are then assimilated into Weather Research and Forecasting Model (WRF) using the nudging method. This assimilation method generally improves the forecasts for precipitation above the moderate rainfall level at forecast times of 4–7 days, and reduces the errors of surface wind speed forecasts. This study improves the accuracy of correction models and the forecasts of persistent precipitation weather event with regional model. • A correction model for global forecast system (GFS) upper-air fields is proposed. • The correction model can correct multiple GFS forecasts by training only one model. • The correction model reduces the errors of U and V components of GFS wind fields. • Correction results are assimilated into a regional model as large-scale constraint. • Adjustments to the wind fields are important for improving precipitation forecasts. [ABSTRACT FROM AUTHOR]

Cloud services have become a demand in big cloud data centers for allocating energy-efficient resources and resource utilization. Most current solutions consider a small number of resource categories, which causes an imbalanced load or the needless activation of real servers. Thus, the virtual machine placement-based bin padding algorithm was proposed and analyzed with the equilibrium state of multi-resource utilization. Additionally, the impact of imbalanced resources causes weak virtual machine task allocation that induces poor VMs' scheduling activities. Therefore, this work incorporates efficient conjectured scheduling to prioritize the task based on computational time with trust parameters calculation. Moreover, to improve the performance, the new abatement strategy was used with efficient resource management, which generates a physical host to minimize energy burning and increase the quality of service to the data center by reducing overload. Implementing the proposed algorithm efficiently utilizes the memory of up to 290 MB for processing 90 MB files with a low computational time of 23 s Further algorithms tested on ten benchmark functions show that they outperform similar algorithms. Thus, the proposed model applies to individual servers in the cloud. [ABSTRACT FROM AUTHOR]

The increasing prevalence of high-performance computing data centers necessitates the adoption of cutting-edge cooling technologies to ensure the safe and reliable operation of their powerful microprocessors. Two-phase cooling schemes are well-suited for high heat flux scenarios because of their high heat transfer coefficients and their ability to enhance chip temperature uniformity. In this study, we perform experimental characterization and deep learning driven optimization of a commercial two-phase cold plate. The initial working design of the cold plate comprises a fin height of 3 mm , fin thickness of 0.1 mm , and a channel width of 0.1 mm.A dielectric coolant, Novec /HFE 7000, was impinged into microchannel fins through impinging jets. A copper block simulated an electronic chip with a surface area of 1 " × 1 ". The experiment was conducted with three different coolant inlet temperatures of 25 ° C , 36 ° C , and 48 ° C with varying heat flux levels ranging from 7.5 to 73.5 W cm 2. The effects of coolant inlet temperatures and flow rate on the thermo-hydraulic performance of the cold plate were explored. In two-phase flow, increasing coolant inlet temperature results in more nucleation sites and improved thermal performance consequently. Thermal resistance drops with flow rate in single-phase flow while it is not affected by flow rate in nucleate boiling region. An improvement in the design of the cold plate was carried out, with the goal of increasing the number of bubble sites and flow velocity at the root fins, by cutting the original fins and creating channels perpendicular to the original channels. Three design parameters, fin height, width of machined channels, and height of short fins preserved through machined channels, were defined. It was observed that widening the machined channels and cutting fins to some point can improve the thermal performance of the cold plate. However, removing fins excessively adversely affects the thermal performance of the cold plate because of loss of heat transfer surface area. Moreover, preserving the short fins through the machined channels decreases thermal resistance as they increase heat transfer surface area and nucleation sites. Furthermore, a deep learning-based compact model is demonstrated for the two-phase cold plate design in the specific range of geometry and flow conditions. The developed compact model is utilized to drive the single and multi-objective optimization to arrive at global optimal results. [ABSTRACT FROM AUTHOR]

With the advent of the information age, the scale of data centers has developed unprecedentedly, in which the cooling system consumes a lot of energy. Therefore, it is crucial to adjust the operating parameters in real-time to maximize energy savings. In this article, the mathematical models of each component are established, and the functions of system power consumption and chip temperatures are obtained for the liquid-cooled data center based on cold plates. The method proposed in this article has good accuracy through experimental verification and can be used for the optimization of operating parameters. The server chips in data centers need to be maintained within a safe range, so we take the minimum system power consumption as the optimization goal and the chip temperature as the constraint condition to calculate the optimal cooling tower wind volume, primary side flow rate, and secondary side flow rate under different environmental temperatures and heat loads, and fit all three for real-time optimization and regulation. The results show that the intelligent control proposed in this article can save 42.7% energy and reduce PUE to 1.16 under variable heat load, and save 30.6% energy and improve PUE by 4.3% under variable environmental temperature. The intelligent control method described in this article provides guidance for real-time optimization in data centers. • The models of each equipment in the liquid-cooled data center based on cold plates were established. • The effect of three operating parameters on system power consumption and chip temperature are analyzed. • Operating parameters of the system for different ambient temperatures and thermal loads are optimized in real time. • The energy consumption of the cooling system is reduced and PUE of data center is improved. [ABSTRACT FROM AUTHOR]

• An improved empirical wavelet transform (IEWT) is proposed. • A new wave forecast method called IEWT-LSTM is introduced. • The forecast results by the LSTM, EWT-LSTM, and IWET-LSTM are analyzed. • IEWT-LSTM has significantly advantage and it is more superior than EWT-LSTM and LSTM. Due to strong non-linearity, ocean surface gravity waves are difficult to directly and accurately predict, despite their importance for a wide range of coastal, nearshore, and offshore activities. To minimize forecast errors, a hybrid combined improved empirical wavelet transform decomposition (IEWT) and long-short term memory network (LSTM) model has been proposed. Data from National Data Buoy Center buoys deployed in the North Pacific Ocean are taken as an example to verify the models. Wave forecasts using the LSTM, EWT-LSTM, and IWET-LSTM models are compared with the observations at 6, 12, 18, 24 and 48 h forecast windows. Consequently, IEWT-LSTM is superior to EWT-LSTM or LSTM models, especially for larger waves at longer long forecast windows. [ABSTRACT FROM AUTHOR]

In this paper, we present a novel clustering approach that eliminates the need for predefined cluster centres and cluster counts, addressing common limitations in traditional clustering algorithms. We achieve this by allowing each data point within the problem space to serve as a potential cluster centre, determined by the presence of nearby data points. Three distinct algorithms are introduced: i) Identifying valid data points for clustering by assessing their presence within the specified vicinity of each potential cluster centre, allowing data points to belong to multiple clusters, ii) Calculating the fuzzy membership of shared data points among the clusters and iii) Retaining shared data points in the cluster with the highest membership and removing them from others. A comprehensive evaluation of this method using synthetic data and benchmark datasets from well-known problems demonstrates promising performance. Our approach offers increased flexibility, ease of use, and enhances the fidelity of cluster centres to the underlying data distribution. [ABSTRACT FROM AUTHOR]

Cloud Computing is a robust model that allows users and organizations to purchase required services per their needs. The model offers many services such as storage, platforms for deployment, convenient access to web services, and so on. Load Balancing is a common issue in the cloud that makes it hard to maintain the performance of the applications adjacent to the Quality of Service (QoS) measurement and following the Service Level Agreement (SLA) document as required from the cloud providers to enterprises. Cloud providers struggle to distribute equal workload among the servers. An efficient LB technique should optimize and ensure high user satisfaction by utilizing the resources of VMs efficiently. This paper presents a comprehensive review of various Load Balancing techniques in a static, dynamic, and nature-inspired cloud environment to address the Data Center Response Time and overall performance. An analytical review of the algorithms is provided, and a research gap is concluded for the future research perspective in this domain. This research also provides a graphical representation of reviewed algorithms to highlight the operational flow. Additionally, this review presents a fault-tolerant framework and explores the other existing frameworks in the recent literature. [ABSTRACT FROM AUTHOR]