With the rapid growth of cloud computing,the number of data centers(DCs)continuously increases,leading to a high-energy consumption dilemma.Cooling,apart from IT equipment,represents the largest energy consumption in ...With the rapid growth of cloud computing,the number of data centers(DCs)continuously increases,leading to a high-energy consumption dilemma.Cooling,apart from IT equipment,represents the largest energy consumption in DCs.Passive design(PD)and active design(AD)are two important approaches in architectural design to reduce energy consumption.However,for DC cooling,few studies have summarized AD,and there are almost no studies on PD.Based on existing international research(2005-2024),this paper summarizes the current state of cooling strategies for DCs.PD encompasses floors,ceilings,and layout and zoning of racks.Additionally,other passive strategies not yet studied in DCs are critically examined.AD includes air,liquid,free,and two-phase cooling.This paper systematically compares the performance of different AD technologies on various KPIs,including energy,economic,and environmental indicators.This paper also explores the application of different cooling design strategies through best-practice examples and presents advanced algorithms for energy management in operational DCs.This study reveals that free cooling is widely employed,with Artificial Neural Networks emerging as the most popular algorithm for managing cooling energy.Finally,this paper suggests four future directions for reducing cooling energy in DCs,with a focus on the development of passive strategies.This paper provides an overview and guide to DC energy-consumption issues,emphasizes the importance of implementing passive and active design strategies to reduce DC cooling energy consumption,and provides directions and references for future energy-efficient DC designs.展开更多
The Mountain Science Data Center(MSDC),founded in 2021 under the Institute of Mountain Hazards and Environment(IMHE),Chinese Academy of Sciences,manages the entire lifecycle of mountain science data.It integrates data...The Mountain Science Data Center(MSDC),founded in 2021 under the Institute of Mountain Hazards and Environment(IMHE),Chinese Academy of Sciences,manages the entire lifecycle of mountain science data.It integrates data from diverse sources,including debris flows,landslides,soils,ecology,geology,natural resources,basic geographic information,and socio-economic data.The center provides comprehensive services,including data collection,processing,analysis tools,modeling,and application support,offering reliable data backing for numerous research projects within the institute.Data management and services are accessible via the Mountain Science Data Center Portal(https://www.msdc.ac.cn/),ensuring long-term,stable,and trustworthy access to facilitate scientific research and institutional development.展开更多
Rapid developments in the electronic information industry drive the increased energy usage and carbon emission of data center buildings,prompting the focus on the energy efficiency and environmental sustainability.Exp...Rapid developments in the electronic information industry drive the increased energy usage and carbon emission of data center buildings,prompting the focus on the energy efficiency and environmental sustainability.Expanded operation envelopes of tropical data centers is assessed to analyze the potential for the building energy savings and carbon emission reduction through collaborative analysis of operation modes(OMs),supply air temperature(SAT),and outdoor air temperature(OAT).The OMs of compression vary with the setpoints of SAT,in which the average exergy efficiency of compressors at alternate operation mode is 6.8%and 8.0%lower than that of double and single compression operations.As SAT rises from 20℃to 32℃,the system exergoeconomic factor increases from 5.4%to 8.0%,and the average carbon cost decreases by 36.5%.Additionally,with just an 8.5%increase in exergy cost(i.e.,Case 8)at OAT rising from 30 to 34℃,the high SAT and low refrigerant charges provide considerable exergy cost advantages versus resisting the OAT fluctuations.Dynamic operation strategies are also proposed and compared to cope with the impacts of tropical environments.Compared to the 26℃SAT baseline,the average energy savings are 9.1-14.7%,indicating the ability to fully utilize outdoor and indoor conditions.展开更多
Air pollution in China covers a large area with complex sources and formation mechanisms,making it a unique place to conduct air pollution and atmospheric chemistry research.The National Natural Science Foundation of ...Air pollution in China covers a large area with complex sources and formation mechanisms,making it a unique place to conduct air pollution and atmospheric chemistry research.The National Natural Science Foundation of China’s Major Research Plan entitled“Fundamental Researches on the Formation and Response Mechanism of the Air Pollution Complex in China”(or the Plan)has funded 76 research projects to explore the causes of air pollution in China,and the key processes of air pollution in atmospheric physics and atmospheric chemistry.In order to summarize the abundant data from the Plan and exhibit the long-term impacts domestically and internationally,an integration project is responsible for collecting the various types of data generated by the 76 projects of the Plan.This project has classified and integrated these data,forming eight categories containing 258 datasets and 15 technical reports in total.The integration project has led to the successful establishment of the China Air Pollution Data Center(CAPDC)platform,providing storage,retrieval,and download services for the eight categories.This platform has distinct features including data visualization,related project information querying,and bilingual services in both English and Chinese,which allows for rapid searching and downloading of data and provides a solid foundation of data and support for future related research.Air pollution control in China,especially in the past decade,is undeniably a global exemplar,and this data center is the first in China to focus on research into the country’s air pollution complex.展开更多
The growth of computing power in data centers(DCs)leads to an increase in energy consumption and noise pollution of air cooling systems.Chip-level cooling with high-efficiency coolant is one of the promising methods t...The growth of computing power in data centers(DCs)leads to an increase in energy consumption and noise pollution of air cooling systems.Chip-level cooling with high-efficiency coolant is one of the promising methods to address the cooling challenge for high-power devices in DCs.Hybrid nanofluid(HNF)has the advantages of high thermal conductivity and good rheological properties.This study summarizes the numerical investigations of HNFs in mini/micro heat sinks,including the numerical methods,hydrothermal characteristics,and enhanced heat transfer technologies.The innovations of this paper include:(1)the characteristics,applicable conditions,and scenarios of each theoretical method and numerical method are clarified;(2)the molecular dynamics(MD)simulation can reveal the synergy effect,micro motion,and agglomeration morphology of different nanoparticles.Machine learning(ML)presents a feasiblemethod for parameter prediction,which provides the opportunity for the intelligent regulation of the thermal performance of HNFs;(3)the HNFs flowboiling and the synergy of passive and active technologies may further improve the overall efficiency of liquid cooling systems in DCs.This review provides valuable insights and references for exploring the multi-phase flow and heat transport mechanisms of HNFs,and promoting the practical application of HNFs in chip-level liquid cooling in DCs.展开更多
Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communi...Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communication has evolved into an increasingly prominent area of research in recent years.Here,we demonstrate DSP-free coherent optical transmission by analog signal processing in frequency synchronous optical network(FSON)architecture,which supports polarization multiplexing and higher-order modulation formats.The FSON architecture that allows the numerous laser sources of optical transceivers within a data center can be quasi-synchronized by means of a tree-distributed homology architecture.In conjunction with our proposed pilot-tone assisted Costas loop for an analog coherent receiver,we achieve a record dual-polarization 224-Gb/s 16-QAM 5-km mismatch transmission with reset-free carrier phase recovery in the optical domain.Our proposed DSP-free analog coherent detection system based on the FSON makes it a promising solution for next-generation,low-power,and high-capacity coherent data center interconnects.展开更多
National Population Health Data Center(NPHDC)is one of China's 20 national-level science data centers,jointly designated by the Ministry of Science and Technology and the Ministry of Finance.Operated by the Chines...National Population Health Data Center(NPHDC)is one of China's 20 national-level science data centers,jointly designated by the Ministry of Science and Technology and the Ministry of Finance.Operated by the Chinese Academy of Medical Sciences under the oversight of the National Health Commission,NPHDC adheres to national regulations including the Scientific Data Management Measures and the National Science and Technology Infrastructure Service Platform Management Measures,and is committed to collecting,integrating,managing,and sharing biomedical and health data through openaccess platform,fostering open sharing and engaging in international cooperation.展开更多
Data centers operate as physical digital infrastructure for generating,storing,computing,transmitting,and utilizing massive data and information,constituting the backbone of the flourishing digital economy across the ...Data centers operate as physical digital infrastructure for generating,storing,computing,transmitting,and utilizing massive data and information,constituting the backbone of the flourishing digital economy across the world.Given the lack of a consistent analysis for studying the locational factors of data centers and empirical deficiencies in longitudinal investigations on spatial dynamics of heterogeneous data centers,this paper develops a comprehensive analytical framework to examine the dynamic geographies and locational factors of techno-environmentally heterogeneous data centers across Chinese cities in the period of 2006–2021.First,we develop a“supply-demand-environment trinity”analytical framework as well as an accompanying evaluation indicator system with Chinese characteristics.Second,the dynamic geographies of data centers in Chinese cities over the last decades are characterized as spatial polarization in economically leading urban agglomerations alongside persistent interregional gaps across eastern,central,and western regions.Data centers present dual spatial expansion trajectories featuring outward radiation from eastern core urban agglomerations to adjacent peripheries and leapfrog diffusion to strategic central and western digital infrastructural hubs.Third,it is empirically verified that data center construction in Chinese cities over the last decades has been jointly influenced by supply-,demand-,and environment-side locational factors,echoing the efficacy of the trinity analytical framework.Overall,our findings demonstrate the temporal variance,contextual contingency,and attribute-based differentiation of locational factors underlying techno-environmentally heterogeneous data centers in Chinese cities.展开更多
Aiming at the problems such as low throughput and unbalanced load of data center network caused by traditional multipath routing strategy,a dynamic load balancing strategy for flow classification oriented to Fat-Tree ...Aiming at the problems such as low throughput and unbalanced load of data center network caused by traditional multipath routing strategy,a dynamic load balancing strategy for flow classification oriented to Fat-Tree topology based on the software defined network(SDN)architecture is proposed,named DLB-FC.Multi-index evaluation methods such as link state information and network traffic characteristics are considered.DLB-FC mechanism can dynamically adjust the flow classification threshold to differentiate between large and small flows.The scheme selects different forwarding paths to meet the transmission performance requirements of different flow characteristics.On this basis,an SDN simulation platform is built for performance testing.The simulation results show that DLB-FC algorithm can dynamically distinguish large flows from small flows and achieve load balancing effectively.Compared with equal-cost multi-path(ECMP),global first fit(GFF)and minmum total delay load routing(MTDLR)algorithms,DLB-FC scheme improves the network throughput and link utilization of the data center network effectively.The transmission delay is also reduced with better load balance.展开更多
Most data centers currently tap into existing power grids to draw the immense amount of electricity they need to operate.But many of the data centers that Google(Mountain View,CA,USA)plans to open in the next few year...Most data centers currently tap into existing power grids to draw the immense amount of electricity they need to operate.But many of the data centers that Google(Mountain View,CA,USA)plans to open in the next few years will boast their own power plants,an arrangement known as colocation[1].Under an agreement announced in December 2024,the company will site data centers in industrial parks where its partner Intersect Power of Houston,TX,USA,has installed clean power facilities[1,2].The first of these complexes is scheduled to come online in 2026[1].展开更多
With the advent of the digital economy,there has been a rapid proliferation of small-scale Internet data centers(SIDCs).By leveraging their spatiotemporal load regulation potential through data workload balancing,aggr...With the advent of the digital economy,there has been a rapid proliferation of small-scale Internet data centers(SIDCs).By leveraging their spatiotemporal load regulation potential through data workload balancing,aggregated SIDCs have emerged as promising demand response(DR)resources for future power distribution systems.This paper presents an innovative framework for assessing capacity value(CV)by aggregating SIDCs participating in DR programs(SIDC-DR).Initially,we delineate the concept of CV tailored for aggregated SIDC scenarios and establish a metric for the assessment.Considering the effects of the data load dynamics,equipment constraints,and user behavior,we developed a sophisticated DR model for aggregated SIDCs using a data network aggregation method.Unlike existing studies,the proposed model captures the uncertainties associated with end tenant decisions to opt into an SIDC-DR program by utilizing a novel uncertainty modeling approach called Z-number formulation.This approach accounts for both the uncertainty in user participation intentions and the reliability of basic information during the DR process,enabling high-resolution profiling of the SIDC-DR potential in the CV evaluation.Simulation results from numerical studies conducted on a modified IEEE-33 node distribution system confirmed the effectiveness of the proposed approach and highlighted the potential benefits of SIDC-DR utilization in the efficient operation of future power systems.展开更多
Anomaly detection is an important task for maintaining the performance of cloud data center.Traditional anomaly detection primarily examines individual Virtual Machine(VM)behavior,neglecting the impact of interactions...Anomaly detection is an important task for maintaining the performance of cloud data center.Traditional anomaly detection primarily examines individual Virtual Machine(VM)behavior,neglecting the impact of interactions among multiple VMs on Key Performance Indicator(KPI)data,e.g.,memory utilization.Furthermore,the nonstationarity,high complexity,and uncertain periodicity of KPI data in VM also bring difficulties to deep learningbased anomaly detection tasks.To settle these challenges,this paper proposes MCBiWGAN-GTN,a multi-channel semi-supervised time series anomaly detection algorithm based on the Bidirectional Wasserstein Generative Adversarial Network with Graph-Time Network(BiWGAN-GTN)and the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN).(a)The BiWGAN-GTN algorithm is proposed to extract spatiotemporal information from data.(b)The loss function of BiWGAN-GTN is redesigned to solve the abnormal data intrusion problem during the training process.(c)MCBiWGAN-GTN is designed to reduce data complexity through CEEMDAN for time series decomposition and utilizes BiWGAN-GTN to train different components.(d)To adapt the proposed algorithm for the entire cloud data center,a cloud data center anomaly detection framework based on Swarm Learning(SL)is designed.The evaluation results on a real-world cloud data center dataset show that MCBiWGAN-GTN outperforms the baseline,with an F1-score of 0.96,an accuracy of 0.935,a precision of 0.954,a recall of 0.967,and an FPR of 0.203.The experiments also verify the stability of MCBiWGAN-GTN,the impact of parameter configurations,and the effectiveness of the proposed SL framework.展开更多
The effect of gradient exhaust strategy and blind plate installation on the inhibition of backflow and thermal stratification in data center cabinets is systematically investigated in this study through numericalmetho...The effect of gradient exhaust strategy and blind plate installation on the inhibition of backflow and thermal stratification in data center cabinets is systematically investigated in this study through numericalmethods.The validated Re-Normalization Group(RNG)k-ε turbulence model was used to analyze airflow patterns within cabinet structures equipped with backplane air conditioning.Key findings reveal that server-generated thermal plumes induce hot air accumulation at the cabinet apex,creating a 0.8℃ temperature elevation at the top server’s inlet compared to the ideal situation(23℃).Strategic increases in backplane fan exhaust airflow rates reduce server 1’s inlet temperature from 26.1℃(0%redundancy case)to 23.1℃(40%redundancy case).Gradient exhaust strategies achieve equivalent server temperature performance to uniform exhaust distributions while requiring 25%less redundant airflow.This approach decreases the recirculation ratio from1.52%(uniformexhaust at 15%redundancy)to 0.57%(gradient exhaust at equivalent redundancy).Comparative analyses demonstrate divergent thermal behaviors:in bottom-server-absent configurations,gradient exhaust reduces top server inlet temperatures by 1.6℃vs.uniformexhaust,whereas top-serverabsent configurations exhibit a 1.8℃ temperature increase under gradient conditions.The blind plate implementation achieves a 0.4℃ top server temperature reduction compared to 15%-redundancy uniform exhaust systems without requiring additional airflow redundancy.Partially installed server arrangements with blind plates maintain thermal characteristics comparable to fully populated cabinets.This study validates gradient exhaust and blind plate technologies as effective countermeasures against cabinet-scale thermal recirculation,providing actionable insights for optimizing backplane air conditioning systems in mission-critical data center environments.展开更多
To improve the traffic scheduling capability in operator data center networks,an analysis prediction and online scheduling mechanism(APOS)is designed,considering both the network structure and the network traffic in t...To improve the traffic scheduling capability in operator data center networks,an analysis prediction and online scheduling mechanism(APOS)is designed,considering both the network structure and the network traffic in the operator data center.Fibonacci tree optimization algorithm(FTO)is embedded into the analysis prediction and the online scheduling stages,the FTO traffic scheduling strategy is proposed.By taking the global optimal and the multi-modal optimization advantage of FTO,the traffic scheduling optimal solution and many suboptimal solutions can be obtained.The experiment results show that the FTO traffic scheduling strategy can schedule traffic in data center networks reasonably,and improve the load balancing in the operator data center network effectively.展开更多
The rapid advancement of artificial intelligence(AI)has significantly increased the computational load on data centers.AI-related computational activities consume considerable electricity and result in substantial car...The rapid advancement of artificial intelligence(AI)has significantly increased the computational load on data centers.AI-related computational activities consume considerable electricity and result in substantial carbon emissions.To mitigate these emissions,future data centers should be strategically planned and operated to fully utilize renewable energy resources while meeting growing computational demands.This paper aims to investigate how much carbon emission reduction can be achieved by using a carbonoriented demand response to guide the optimal planning and operation of data centers.A carbon-oriented data center planning model is proposed that considers the carbon-oriented demand response of the AI load.In the planning model,future operation simulations comprehensively coordinate the temporal‒spatial flexibility of computational loads and the quality of service(QoS).An empirical study based on the proposed models is conducted on real-world data from China.The results from the empirical analysis show that newly constructed data centers are recommended to be built in Gansu Province,Ningxia Hui Autonomous Region,Sichuan Province,Inner Mongolia Autonomous Region,and Qinghai Province,accounting for 57%of the total national increase in server capacity.33%of the computational load from Eastern China should be transferred to the West,which could reduce the overall load carbon emissions by 26%.展开更多
With the increasing complexity and scale of hyperscale data centers,the requirement for intelligent,real-time power delivery has never been more critical to ensure uptime,energy efficiency,and sustainability.Those tec...With the increasing complexity and scale of hyperscale data centers,the requirement for intelligent,real-time power delivery has never been more critical to ensure uptime,energy efficiency,and sustainability.Those techniques are typically static,reactive(since CPU and workload scaling is applied to performance events that occur after a request has been submitted,and is thus can be classified as a reactive response.),and require manual operation,and cannot cope with the dynamic nature of the workloads,the distributed architectures as well as the non-uniform energy sources in today’s data centers.In this paper,we elaborate on how artificial intelligence(AI)is revolutionizing power distribution in hyperscale data centers,making predictive load forecasting,real-time fault detection,and autonomous power optimization possible.We explain how ML(machine learning)and RL(reinforcement learning)-based models have been introduced in PDN(power delivery networks)for load balancing in three-phase systems,overprovisioning reduction,and energy flow optimization from the grid to the rack.The paper considers the architectural pieces of the AI-led systems,such as data ingestion pipelines,anomaly detection frameworks,and control algorithms to manage the power switching,cooling synchronization,and grid/microgrid interaction.Practical use cases show the value of these systems on PUE,infrastructure reliability,and environmental footprint.Key implementation challenges,including data quality,legacy systemintegration,and AI decision-making governance,are also discussed.Last,the paper speculates on the future of autonomous DC power infrastructure where AI becomes not only an assistive resource to the operator but really takes control over infrastructure behavior end-to-end,from procuring energy,to phase balancing,to predicting maintenance.Integrating technology innovation with operational sustainability,AI-powered power distribution is emerging as a core competence for the Smart Digital Power Facility of the Future.展开更多
With the rapid development of technologies such as big data and cloud computing,data communication and data computing in the form of exponential growth have led to a large amount of energy consumption in data centers....With the rapid development of technologies such as big data and cloud computing,data communication and data computing in the form of exponential growth have led to a large amount of energy consumption in data centers.Globally,data centers will become the world’s largest users of energy consumption,with the ratio rising from 3%in 2017 to 4.5%in 2025.Due to its unique climate and energy-saving advantages,the high-latitude area in the Pan-Arctic region has gradually become a hotspot for data center site selection in recent years.In order to predict and analyze the future energy consumption and carbon emissions of global data centers,this paper presents a new method based on global data center traffic and power usage effectiveness(PUE)for energy consumption prediction.Firstly,global data center traffic growth is predicted based on the Cisco’s research.Secondly,the dynamic global average PUE and the high latitude PUE based on Romonet simulation model are obtained,and then global data center energy consumption with two different scenarios,the decentralized scenario and the centralized scenario,is analyzed quantitatively via the polynomial fitting method.The simulation results show that,in 2030,the global data center energy consumption and carbon emissions are reduced by about 301 billion kWh and 720 million tons CO2 in the centralized scenario compared with that of the decentralized scenario,which confirms that the establishment of data centers in the Pan-Arctic region in the future can effectively relief the climate change and energy problems.This study provides support for global energy consumption prediction,and guidance for the layout of future global data centers from the perspective of energy consumption.Moreover,it provides support of the feasibility of the integration of energy and information networks under the Global Energy Interconnection conception.展开更多
In recent years,dual-homed topologies have appeared in data centers in order to offer higher aggregate bandwidth by using multiple paths simultaneously.Multipath TCP(MPTCP) has been proposed as a replacement for TCP i...In recent years,dual-homed topologies have appeared in data centers in order to offer higher aggregate bandwidth by using multiple paths simultaneously.Multipath TCP(MPTCP) has been proposed as a replacement for TCP in those topologies as it can efficiently offer improved throughput and better fairness.However,we have found that MPTCP has a problem in terms of incast collapse where the receiver suffers a drastic goodput drop when it simultaneously requests data over multiple servers.In this paper,we investigate why the goodput collapses even if MPTCP is able to actively relieve hot spots.In order to address the problem,we propose an equally-weighted congestion control algorithm for MPTCP,namely EW-MPTCP,without need for centralized control,additional infrastructure and a hardware upgrade.In our scheme,in addition to the coupled congestion control performed on each subflow of an MPTCP connection,we allow each subflow to perform an additional congestion control operation by weighting the congestion window in reverse proportion to the number of servers.The goal is to mitigate incast collapse by allowing multiple MPTCP subflows to compete fairly with a single-TCP flow at the shared bottleneck.The simulation results show that our solution mitigates the incast problem and noticeably improves goodput in data centers.展开更多
How to effectively reduce the energy consumption of large-scale data centers is a key issue in cloud computing. This paper presents a novel low-power task scheduling algorithm （L3SA） for large-scale cloud data cente...How to effectively reduce the energy consumption of large-scale data centers is a key issue in cloud computing. This paper presents a novel low-power task scheduling algorithm （L3SA） for large-scale cloud data centers. The winner tree is introduced to make the data nodes as the leaf nodes of the tree and the final winner on the purpose of reducing energy consumption is selected. The complexity of large-scale cloud data centers is fully consider, and the task comparson coefficient is defined to make task scheduling strategy more reasonable. Experiments and performance analysis show that the proposed algorithm can effectively improve the node utilization, and reduce the overall power consumption of the cloud data center.展开更多
With the emerging diverse applications in data centers,the demands on quality of service in data centers also become diverse,such as high throughput of elephant flows and low latency of deadline-sensitive flows.Howeve...With the emerging diverse applications in data centers,the demands on quality of service in data centers also become diverse,such as high throughput of elephant flows and low latency of deadline-sensitive flows.However,traditional TCPs are ill-suited to such situations and always result in the inefficiency(e.g.missing the flow deadline,inevitable throughput collapse)of data transfers.This further degrades the user-perceived quality of service(QoS)in data centers.To reduce the flow completion time of mice and deadline-sensitive flows along with promoting the throughput of elephant flows,an efficient and deadline-aware priority-driven congestion control(PCC)protocol,which grants mice and deadline-sensitive flows the highest priority,is proposed in this paper.Specifically,PCC computes the priority of different flows according to the size of transmitted data,the remaining data volume,and the flows’deadline.Then PCC adjusts the congestion window according to the flow priority and the degree of network congestion.Furthermore,switches in data centers control the input/output of packets based on the flow priority and the queue length.Different from existing TCPs,to speed up the data transfers of mice and deadline-sensitive flows,PCC provides an effective method to compute and encode the flow priority explicitly.According to the flow priority,switches can manage packets efficiently and ensure the data transfers of high priority flows through a weighted priority scheduling with minor modification.The experimental results prove that PCC can improve the data transfer performance of mice and deadline-sensitive flows while guaranting the throughput of elephant flows.展开更多
文摘With the rapid growth of cloud computing,the number of data centers(DCs)continuously increases,leading to a high-energy consumption dilemma.Cooling,apart from IT equipment,represents the largest energy consumption in DCs.Passive design(PD)and active design(AD)are two important approaches in architectural design to reduce energy consumption.However,for DC cooling,few studies have summarized AD,and there are almost no studies on PD.Based on existing international research(2005-2024),this paper summarizes the current state of cooling strategies for DCs.PD encompasses floors,ceilings,and layout and zoning of racks.Additionally,other passive strategies not yet studied in DCs are critically examined.AD includes air,liquid,free,and two-phase cooling.This paper systematically compares the performance of different AD technologies on various KPIs,including energy,economic,and environmental indicators.This paper also explores the application of different cooling design strategies through best-practice examples and presents advanced algorithms for energy management in operational DCs.This study reveals that free cooling is widely employed,with Artificial Neural Networks emerging as the most popular algorithm for managing cooling energy.Finally,this paper suggests four future directions for reducing cooling energy in DCs,with a focus on the development of passive strategies.This paper provides an overview and guide to DC energy-consumption issues,emphasizes the importance of implementing passive and active design strategies to reduce DC cooling energy consumption,and provides directions and references for future energy-efficient DC designs.
文摘The Mountain Science Data Center(MSDC),founded in 2021 under the Institute of Mountain Hazards and Environment(IMHE),Chinese Academy of Sciences,manages the entire lifecycle of mountain science data.It integrates data from diverse sources,including debris flows,landslides,soils,ecology,geology,natural resources,basic geographic information,and socio-economic data.The center provides comprehensive services,including data collection,processing,analysis tools,modeling,and application support,offering reliable data backing for numerous research projects within the institute.Data management and services are accessible via the Mountain Science Data Center Portal(https://www.msdc.ac.cn/),ensuring long-term,stable,and trustworthy access to facilitate scientific research and institutional development.
基金supported by the National Research Foundation,Singapore,funded under Energy Research Testbed and Industry Partnership Funding Initiative,part of the Energy Grid(EG)2.0 programme.
文摘Rapid developments in the electronic information industry drive the increased energy usage and carbon emission of data center buildings,prompting the focus on the energy efficiency and environmental sustainability.Expanded operation envelopes of tropical data centers is assessed to analyze the potential for the building energy savings and carbon emission reduction through collaborative analysis of operation modes(OMs),supply air temperature(SAT),and outdoor air temperature(OAT).The OMs of compression vary with the setpoints of SAT,in which the average exergy efficiency of compressors at alternate operation mode is 6.8%and 8.0%lower than that of double and single compression operations.As SAT rises from 20℃to 32℃,the system exergoeconomic factor increases from 5.4%to 8.0%,and the average carbon cost decreases by 36.5%.Additionally,with just an 8.5%increase in exergy cost(i.e.,Case 8)at OAT rising from 30 to 34℃,the high SAT and low refrigerant charges provide considerable exergy cost advantages versus resisting the OAT fluctuations.Dynamic operation strategies are also proposed and compared to cope with the impacts of tropical environments.Compared to the 26℃SAT baseline,the average energy savings are 9.1-14.7%,indicating the ability to fully utilize outdoor and indoor conditions.
基金supported by the National Natural Science Foundation of China(Grant No.92044303)。
文摘Air pollution in China covers a large area with complex sources and formation mechanisms,making it a unique place to conduct air pollution and atmospheric chemistry research.The National Natural Science Foundation of China’s Major Research Plan entitled“Fundamental Researches on the Formation and Response Mechanism of the Air Pollution Complex in China”(or the Plan)has funded 76 research projects to explore the causes of air pollution in China,and the key processes of air pollution in atmospheric physics and atmospheric chemistry.In order to summarize the abundant data from the Plan and exhibit the long-term impacts domestically and internationally,an integration project is responsible for collecting the various types of data generated by the 76 projects of the Plan.This project has classified and integrated these data,forming eight categories containing 258 datasets and 15 technical reports in total.The integration project has led to the successful establishment of the China Air Pollution Data Center(CAPDC)platform,providing storage,retrieval,and download services for the eight categories.This platform has distinct features including data visualization,related project information querying,and bilingual services in both English and Chinese,which allows for rapid searching and downloading of data and provides a solid foundation of data and support for future related research.Air pollution control in China,especially in the past decade,is undeniably a global exemplar,and this data center is the first in China to focus on research into the country’s air pollution complex.
基金funded by the Science and Technology Project of Tianjin(No.24YDTPJC00680)the National Natural Science Foundation of China(No.52406191).
文摘The growth of computing power in data centers(DCs)leads to an increase in energy consumption and noise pollution of air cooling systems.Chip-level cooling with high-efficiency coolant is one of the promising methods to address the cooling challenge for high-power devices in DCs.Hybrid nanofluid(HNF)has the advantages of high thermal conductivity and good rheological properties.This study summarizes the numerical investigations of HNFs in mini/micro heat sinks,including the numerical methods,hydrothermal characteristics,and enhanced heat transfer technologies.The innovations of this paper include:(1)the characteristics,applicable conditions,and scenarios of each theoretical method and numerical method are clarified;(2)the molecular dynamics(MD)simulation can reveal the synergy effect,micro motion,and agglomeration morphology of different nanoparticles.Machine learning(ML)presents a feasiblemethod for parameter prediction,which provides the opportunity for the intelligent regulation of the thermal performance of HNFs;(3)the HNFs flowboiling and the synergy of passive and active technologies may further improve the overall efficiency of liquid cooling systems in DCs.This review provides valuable insights and references for exploring the multi-phase flow and heat transport mechanisms of HNFs,and promoting the practical application of HNFs in chip-level liquid cooling in DCs.
基金supported by the National Natural Science Foundation of China(Grant Nos.62405250 and 62471404)the China Postdoctoral Science Foundation(Grant No.2024M762955)+1 种基金the Key Project of Westlake Institute for Optoelectronics(Grant No.2023GD003)the Optical Com-munication and Sensing Laboratory,School of Engineering,Westlake University.
文摘Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communication has evolved into an increasingly prominent area of research in recent years.Here,we demonstrate DSP-free coherent optical transmission by analog signal processing in frequency synchronous optical network(FSON)architecture,which supports polarization multiplexing and higher-order modulation formats.The FSON architecture that allows the numerous laser sources of optical transceivers within a data center can be quasi-synchronized by means of a tree-distributed homology architecture.In conjunction with our proposed pilot-tone assisted Costas loop for an analog coherent receiver,we achieve a record dual-polarization 224-Gb/s 16-QAM 5-km mismatch transmission with reset-free carrier phase recovery in the optical domain.Our proposed DSP-free analog coherent detection system based on the FSON makes it a promising solution for next-generation,low-power,and high-capacity coherent data center interconnects.
文摘National Population Health Data Center(NPHDC)is one of China's 20 national-level science data centers,jointly designated by the Ministry of Science and Technology and the Ministry of Finance.Operated by the Chinese Academy of Medical Sciences under the oversight of the National Health Commission,NPHDC adheres to national regulations including the Scientific Data Management Measures and the National Science and Technology Infrastructure Service Platform Management Measures,and is committed to collecting,integrating,managing,and sharing biomedical and health data through openaccess platform,fostering open sharing and engaging in international cooperation.
基金Major Program of National Social Science Foundation of China,No.21&ZD107。
文摘Data centers operate as physical digital infrastructure for generating,storing,computing,transmitting,and utilizing massive data and information,constituting the backbone of the flourishing digital economy across the world.Given the lack of a consistent analysis for studying the locational factors of data centers and empirical deficiencies in longitudinal investigations on spatial dynamics of heterogeneous data centers,this paper develops a comprehensive analytical framework to examine the dynamic geographies and locational factors of techno-environmentally heterogeneous data centers across Chinese cities in the period of 2006–2021.First,we develop a“supply-demand-environment trinity”analytical framework as well as an accompanying evaluation indicator system with Chinese characteristics.Second,the dynamic geographies of data centers in Chinese cities over the last decades are characterized as spatial polarization in economically leading urban agglomerations alongside persistent interregional gaps across eastern,central,and western regions.Data centers present dual spatial expansion trajectories featuring outward radiation from eastern core urban agglomerations to adjacent peripheries and leapfrog diffusion to strategic central and western digital infrastructural hubs.Third,it is empirically verified that data center construction in Chinese cities over the last decades has been jointly influenced by supply-,demand-,and environment-side locational factors,echoing the efficacy of the trinity analytical framework.Overall,our findings demonstrate the temporal variance,contextual contingency,and attribute-based differentiation of locational factors underlying techno-environmentally heterogeneous data centers in Chinese cities.
基金Supported by the National Natural Science Foundation of China(No.61672270)Jiangsu Provionce Teaching Reform Project for Cloud Computing Technology and Application Talent Training(No.201802130049).
文摘Aiming at the problems such as low throughput and unbalanced load of data center network caused by traditional multipath routing strategy,a dynamic load balancing strategy for flow classification oriented to Fat-Tree topology based on the software defined network(SDN)architecture is proposed,named DLB-FC.Multi-index evaluation methods such as link state information and network traffic characteristics are considered.DLB-FC mechanism can dynamically adjust the flow classification threshold to differentiate between large and small flows.The scheme selects different forwarding paths to meet the transmission performance requirements of different flow characteristics.On this basis,an SDN simulation platform is built for performance testing.The simulation results show that DLB-FC algorithm can dynamically distinguish large flows from small flows and achieve load balancing effectively.Compared with equal-cost multi-path(ECMP),global first fit(GFF)and minmum total delay load routing(MTDLR)algorithms,DLB-FC scheme improves the network throughput and link utilization of the data center network effectively.The transmission delay is also reduced with better load balance.
文摘Most data centers currently tap into existing power grids to draw the immense amount of electricity they need to operate.But many of the data centers that Google(Mountain View,CA,USA)plans to open in the next few years will boast their own power plants,an arrangement known as colocation[1].Under an agreement announced in December 2024,the company will site data centers in industrial parks where its partner Intersect Power of Houston,TX,USA,has installed clean power facilities[1,2].The first of these complexes is scheduled to come online in 2026[1].
基金supported in part by the National Natural Science Foundation of China under Grant 52177082in part by the Beijing Nova Program under Grant 20220484007.
文摘With the advent of the digital economy,there has been a rapid proliferation of small-scale Internet data centers(SIDCs).By leveraging their spatiotemporal load regulation potential through data workload balancing,aggregated SIDCs have emerged as promising demand response(DR)resources for future power distribution systems.This paper presents an innovative framework for assessing capacity value(CV)by aggregating SIDCs participating in DR programs(SIDC-DR).Initially,we delineate the concept of CV tailored for aggregated SIDC scenarios and establish a metric for the assessment.Considering the effects of the data load dynamics,equipment constraints,and user behavior,we developed a sophisticated DR model for aggregated SIDCs using a data network aggregation method.Unlike existing studies,the proposed model captures the uncertainties associated with end tenant decisions to opt into an SIDC-DR program by utilizing a novel uncertainty modeling approach called Z-number formulation.This approach accounts for both the uncertainty in user participation intentions and the reliability of basic information during the DR process,enabling high-resolution profiling of the SIDC-DR potential in the CV evaluation.Simulation results from numerical studies conducted on a modified IEEE-33 node distribution system confirmed the effectiveness of the proposed approach and highlighted the potential benefits of SIDC-DR utilization in the efficient operation of future power systems.
基金supported in part by National Natural Science Foundation of China under Grant 62071078in part by Sichuan Province Science and Technology Program under Grant 2021YFQ0053。
文摘Anomaly detection is an important task for maintaining the performance of cloud data center.Traditional anomaly detection primarily examines individual Virtual Machine(VM)behavior,neglecting the impact of interactions among multiple VMs on Key Performance Indicator(KPI)data,e.g.,memory utilization.Furthermore,the nonstationarity,high complexity,and uncertain periodicity of KPI data in VM also bring difficulties to deep learningbased anomaly detection tasks.To settle these challenges,this paper proposes MCBiWGAN-GTN,a multi-channel semi-supervised time series anomaly detection algorithm based on the Bidirectional Wasserstein Generative Adversarial Network with Graph-Time Network(BiWGAN-GTN)and the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN).(a)The BiWGAN-GTN algorithm is proposed to extract spatiotemporal information from data.(b)The loss function of BiWGAN-GTN is redesigned to solve the abnormal data intrusion problem during the training process.(c)MCBiWGAN-GTN is designed to reduce data complexity through CEEMDAN for time series decomposition and utilizes BiWGAN-GTN to train different components.(d)To adapt the proposed algorithm for the entire cloud data center,a cloud data center anomaly detection framework based on Swarm Learning(SL)is designed.The evaluation results on a real-world cloud data center dataset show that MCBiWGAN-GTN outperforms the baseline,with an F1-score of 0.96,an accuracy of 0.935,a precision of 0.954,a recall of 0.967,and an FPR of 0.203.The experiments also verify the stability of MCBiWGAN-GTN,the impact of parameter configurations,and the effectiveness of the proposed SL framework.
基金financially supported by the Basic Research Funds for the Central Government“Innovative Team of Zhejiang University”under contract number(2022FZZX01-09).
文摘The effect of gradient exhaust strategy and blind plate installation on the inhibition of backflow and thermal stratification in data center cabinets is systematically investigated in this study through numericalmethods.The validated Re-Normalization Group(RNG)k-ε turbulence model was used to analyze airflow patterns within cabinet structures equipped with backplane air conditioning.Key findings reveal that server-generated thermal plumes induce hot air accumulation at the cabinet apex,creating a 0.8℃ temperature elevation at the top server’s inlet compared to the ideal situation(23℃).Strategic increases in backplane fan exhaust airflow rates reduce server 1’s inlet temperature from 26.1℃(0%redundancy case)to 23.1℃(40%redundancy case).Gradient exhaust strategies achieve equivalent server temperature performance to uniform exhaust distributions while requiring 25%less redundant airflow.This approach decreases the recirculation ratio from1.52%(uniformexhaust at 15%redundancy)to 0.57%(gradient exhaust at equivalent redundancy).Comparative analyses demonstrate divergent thermal behaviors:in bottom-server-absent configurations,gradient exhaust reduces top server inlet temperatures by 1.6℃vs.uniformexhaust,whereas top-serverabsent configurations exhibit a 1.8℃ temperature increase under gradient conditions.The blind plate implementation achieves a 0.4℃ top server temperature reduction compared to 15%-redundancy uniform exhaust systems without requiring additional airflow redundancy.Partially installed server arrangements with blind plates maintain thermal characteristics comparable to fully populated cabinets.This study validates gradient exhaust and blind plate technologies as effective countermeasures against cabinet-scale thermal recirculation,providing actionable insights for optimizing backplane air conditioning systems in mission-critical data center environments.
基金supported by National Natural Science Foundation of China(No.62163036).
文摘To improve the traffic scheduling capability in operator data center networks,an analysis prediction and online scheduling mechanism(APOS)is designed,considering both the network structure and the network traffic in the operator data center.Fibonacci tree optimization algorithm(FTO)is embedded into the analysis prediction and the online scheduling stages,the FTO traffic scheduling strategy is proposed.By taking the global optimal and the multi-modal optimization advantage of FTO,the traffic scheduling optimal solution and many suboptimal solutions can be obtained.The experiment results show that the FTO traffic scheduling strategy can schedule traffic in data center networks reasonably,and improve the load balancing in the operator data center network effectively.
基金supported by the Scientific&Technical Project of the State Grid(5700--202490228A--1--1-ZN).
文摘The rapid advancement of artificial intelligence(AI)has significantly increased the computational load on data centers.AI-related computational activities consume considerable electricity and result in substantial carbon emissions.To mitigate these emissions,future data centers should be strategically planned and operated to fully utilize renewable energy resources while meeting growing computational demands.This paper aims to investigate how much carbon emission reduction can be achieved by using a carbonoriented demand response to guide the optimal planning and operation of data centers.A carbon-oriented data center planning model is proposed that considers the carbon-oriented demand response of the AI load.In the planning model,future operation simulations comprehensively coordinate the temporal‒spatial flexibility of computational loads and the quality of service(QoS).An empirical study based on the proposed models is conducted on real-world data from China.The results from the empirical analysis show that newly constructed data centers are recommended to be built in Gansu Province,Ningxia Hui Autonomous Region,Sichuan Province,Inner Mongolia Autonomous Region,and Qinghai Province,accounting for 57%of the total national increase in server capacity.33%of the computational load from Eastern China should be transferred to the West,which could reduce the overall load carbon emissions by 26%.
文摘With the increasing complexity and scale of hyperscale data centers,the requirement for intelligent,real-time power delivery has never been more critical to ensure uptime,energy efficiency,and sustainability.Those techniques are typically static,reactive(since CPU and workload scaling is applied to performance events that occur after a request has been submitted,and is thus can be classified as a reactive response.),and require manual operation,and cannot cope with the dynamic nature of the workloads,the distributed architectures as well as the non-uniform energy sources in today’s data centers.In this paper,we elaborate on how artificial intelligence(AI)is revolutionizing power distribution in hyperscale data centers,making predictive load forecasting,real-time fault detection,and autonomous power optimization possible.We explain how ML(machine learning)and RL(reinforcement learning)-based models have been introduced in PDN(power delivery networks)for load balancing in three-phase systems,overprovisioning reduction,and energy flow optimization from the grid to the rack.The paper considers the architectural pieces of the AI-led systems,such as data ingestion pipelines,anomaly detection frameworks,and control algorithms to manage the power switching,cooling synchronization,and grid/microgrid interaction.Practical use cases show the value of these systems on PUE,infrastructure reliability,and environmental footprint.Key implementation challenges,including data quality,legacy systemintegration,and AI decision-making governance,are also discussed.Last,the paper speculates on the future of autonomous DC power infrastructure where AI becomes not only an assistive resource to the operator but really takes control over infrastructure behavior end-to-end,from procuring energy,to phase balancing,to predicting maintenance.Integrating technology innovation with operational sustainability,AI-powered power distribution is emerging as a core competence for the Smart Digital Power Facility of the Future.
基金supported by National Natural Science Foundation of China(61472042)Corporation Science and Technology Program of Global Energy Interconnection Group Ltd.(GEIGC-D-[2018]024)
文摘With the rapid development of technologies such as big data and cloud computing,data communication and data computing in the form of exponential growth have led to a large amount of energy consumption in data centers.Globally,data centers will become the world’s largest users of energy consumption,with the ratio rising from 3%in 2017 to 4.5%in 2025.Due to its unique climate and energy-saving advantages,the high-latitude area in the Pan-Arctic region has gradually become a hotspot for data center site selection in recent years.In order to predict and analyze the future energy consumption and carbon emissions of global data centers,this paper presents a new method based on global data center traffic and power usage effectiveness(PUE)for energy consumption prediction.Firstly,global data center traffic growth is predicted based on the Cisco’s research.Secondly,the dynamic global average PUE and the high latitude PUE based on Romonet simulation model are obtained,and then global data center energy consumption with two different scenarios,the decentralized scenario and the centralized scenario,is analyzed quantitatively via the polynomial fitting method.The simulation results show that,in 2030,the global data center energy consumption and carbon emissions are reduced by about 301 billion kWh and 720 million tons CO2 in the centralized scenario compared with that of the decentralized scenario,which confirms that the establishment of data centers in the Pan-Arctic region in the future can effectively relief the climate change and energy problems.This study provides support for global energy consumption prediction,and guidance for the layout of future global data centers from the perspective of energy consumption.Moreover,it provides support of the feasibility of the integration of energy and information networks under the Global Energy Interconnection conception.
基金supported in part by the HUT Distributed and Mobile Cloud Systems research project and Tekes within the ITEA2 project 10014 EASI-CLOUDS
文摘In recent years,dual-homed topologies have appeared in data centers in order to offer higher aggregate bandwidth by using multiple paths simultaneously.Multipath TCP(MPTCP) has been proposed as a replacement for TCP in those topologies as it can efficiently offer improved throughput and better fairness.However,we have found that MPTCP has a problem in terms of incast collapse where the receiver suffers a drastic goodput drop when it simultaneously requests data over multiple servers.In this paper,we investigate why the goodput collapses even if MPTCP is able to actively relieve hot spots.In order to address the problem,we propose an equally-weighted congestion control algorithm for MPTCP,namely EW-MPTCP,without need for centralized control,additional infrastructure and a hardware upgrade.In our scheme,in addition to the coupled congestion control performed on each subflow of an MPTCP connection,we allow each subflow to perform an additional congestion control operation by weighting the congestion window in reverse proportion to the number of servers.The goal is to mitigate incast collapse by allowing multiple MPTCP subflows to compete fairly with a single-TCP flow at the shared bottleneck.The simulation results show that our solution mitigates the incast problem and noticeably improves goodput in data centers.
基金supported by the National Natural Science Foundation of China(6120200461272084)+9 种基金the National Key Basic Research Program of China(973 Program)(2011CB302903)the Specialized Research Fund for the Doctoral Program of Higher Education(2009322312000120113223110003)the China Postdoctoral Science Foundation Funded Project(2011M5000952012T50514)the Natural Science Foundation of Jiangsu Province(BK2011754BK2009426)the Jiangsu Postdoctoral Science Foundation Funded Project(1102103C)the Natural Science Fund of Higher Education of Jiangsu Province(12KJB520007)the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions(yx002001)
文摘How to effectively reduce the energy consumption of large-scale data centers is a key issue in cloud computing. This paper presents a novel low-power task scheduling algorithm （L3SA） for large-scale cloud data centers. The winner tree is introduced to make the data nodes as the leaf nodes of the tree and the final winner on the purpose of reducing energy consumption is selected. The complexity of large-scale cloud data centers is fully consider, and the task comparson coefficient is defined to make task scheduling strategy more reasonable. Experiments and performance analysis show that the proposed algorithm can effectively improve the node utilization, and reduce the overall power consumption of the cloud data center.
基金supported part by the National Natural Science Foundation of China(61601252,61801254)Public Technology Projects of Zhejiang Province(LG-G18F020007)+1 种基金Zhejiang Provincial Natural Science Foundation of China(LY20F020008,LY18F020011,LY20F010004)K.C.Wong Magna Fund in Ningbo University。
文摘With the emerging diverse applications in data centers,the demands on quality of service in data centers also become diverse,such as high throughput of elephant flows and low latency of deadline-sensitive flows.However,traditional TCPs are ill-suited to such situations and always result in the inefficiency(e.g.missing the flow deadline,inevitable throughput collapse)of data transfers.This further degrades the user-perceived quality of service(QoS)in data centers.To reduce the flow completion time of mice and deadline-sensitive flows along with promoting the throughput of elephant flows,an efficient and deadline-aware priority-driven congestion control(PCC)protocol,which grants mice and deadline-sensitive flows the highest priority,is proposed in this paper.Specifically,PCC computes the priority of different flows according to the size of transmitted data,the remaining data volume,and the flows’deadline.Then PCC adjusts the congestion window according to the flow priority and the degree of network congestion.Furthermore,switches in data centers control the input/output of packets based on the flow priority and the queue length.Different from existing TCPs,to speed up the data transfers of mice and deadline-sensitive flows,PCC provides an effective method to compute and encode the flow priority explicitly.According to the flow priority,switches can manage packets efficiently and ensure the data transfers of high priority flows through a weighted priority scheduling with minor modification.The experimental results prove that PCC can improve the data transfer performance of mice and deadline-sensitive flows while guaranting the throughput of elephant flows.
