当前位置:网站首页>What is the charm of Guizhou? Why do Alibaba, Huawei and Tencent build data centers in Guizhou?
What is the charm of Guizhou? Why do Alibaba, Huawei and Tencent build data centers in Guizhou?
2022-06-23 14:14:00 【User 9363305】
At present, Guizhou has gathered more than big data enterprises 9000 home , The output value exceeds 1 100 billion yuan , It can attract tens of thousands of big data talents every year . Everyone knows that big data centers are electric tigers , The operating cost of the big data center mainly includes the power charge of the computer room 、 Broadband cost 、 Machine room construction and amortization 、 Labor cost and machine room rent, etc , The electricity cost of the computer room accounts for more than half of the total cost . At present, Guizhou is rich in power resources , Guizhou has a water power reserve of 1874.5 kilowatts , It ranks the sixth in hydropower reserves in China ; Among them, the amount that can be developed is 1683.3 kilowatts .
It is estimated that ,2020 In, the power consumption of the national data center is about 760 Million kilowatt hour . If converted to carbon dioxide emissions ,2020 In, the carbon dioxide emission of national data centers was nearly 4000 Ten thousand tons of . The overall look , The scale of power consumption and carbon emission of data center accounts for less than that of electricity in the whole society 、 chemical 、 Transportation and other high emission industries , But it is still a big emitter . In the future, the energy consumption and carbon emission of data center will increase steadily , Therefore, energy conservation and emission reduction is a key direction that must be anchored for the long-term development of the data center industry .
// Macro to micro , Drill down step by step
Realize the low-carbon operation of the data center by using graph flapping visualization technology , Help the data center achieve energy conservation and emission reduction .HT Support multiple ways of model rendering , Using lightweight 3D Modeling Technology , 1:1 High simulation , Based on 3D scene ,2D Supplemented by data panel , Digitally display the operation of data center equipment , Switch on and off the refrigeration equipment in time , Reduce the waste of water and electricity .
How to complete the 3D modeling of such a park ?
Under normal circumstances, by providing satellite cloud images 、 design sketch 、 Aerial view 、CAD chart 、 Scene photos and other information , Lightweight modeling by designers .
At present, we have BIM Model information for , Whether it can be directly used for the display of this scene ?
Building information model (Building Information Modeling,BIM) It is a kind of 3D model data in construction industry management information system , Engineering construction professionals can pass BIM Obtain relevant information and data , Plan more efficiently 、 Design 、 Build and manage buildings and infrastructure . take BIM The building data is superimposed on the map scene , Realization BIM + GIS Combined display . Through to BIM Separate the asset construction in different floors and floors of the model , Show details of the equipment .
Except for artificial modeling , Is there any other way ?
At present, tilt photography model can be generated by UAV aerial photography , Re pass GIS Show me the plan .
// Power monitoring Visualization
IDC Energy consumption and management and control system as a large overall data center display system , In the macro use scenario , There will be more focus on the overall data . By docking the number of data centers in the region 、 Energy consumption 、 Energy saving information 、 Energy saving and so on , Help the decision-maker to observe all kinds of objects as much as possible , Create a data intelligent decision-making platform for decision makers .
And as a monitoring system , early warning 、 There must be more alarms . Docking sensing network data , Through the event list display background real-time push data information , Realize real-time perception of display data , Set early warning threshold to realize data monitoring and early warning , Effective monitoring and monitoring of data business data operation situation . PUE (Power Usage Effectiveness), As an indicator of energy efficiency in data centers , Visually display through the data panel .
// Asset visualization
The asset visualization module will view the asset status of different drill down levels as the hierarchical relationship progresses . Asset management module has the function of life cycle management , Through the background data interface to achieve automatic loading and unloading , No manual operation is required , Using interface data to ensure the accuracy of display content , Finally, all the facilities on the platform are integrated . adopt 3D View to view key information , Such as virtual machine 、 System processes 、 Disk space occupation and other real-time data monitoring . Information that presents assets , Including but not limited to equipment type 、 model 、 Purchase and launch date 、 Location 、 Maintenance and other information , And automatically refresh the existing location information of the device .
The module supports fuzzy search and location of devices in the scene , The lens automatically moves to the current position of the positioning device , Click to pop up relevant device information , Assist the operation and maintenance management personnel to quickly find the required equipment . Meet the operation and maintenance management characteristics of different types of assets , Ensure that the asset information is complete and well documented .
// Capacity Management Visualization
Visualization can not only describe the objects seen by the naked eye with images , It can also express the information state of the device . Through the stands 、U position 、 bearing 、 Power consumption, etc , The operation and maintenance personnel of the computer room will have a clearer understanding of the current capacity situation , Such as the power load of the current computer room 、 Space left in the cabinet 、 Load bearing and storage capacity of each area of the computer room , Can be expressed visually , Effectively manage the capacity resources of the computer room , Let all kinds of resources in the computer room load more balanced .
The capacity visualization management function also supports spatial search function , Accurate statistics and presentation of used space and available space . Assist the staff to quickly find the right shelf space for new equipment , Precise positioning of the space required . The main functions are as follows :
Visualization slot :
Perspective of the current data center environment , The number of used seats and the number of remaining seats .
visualization U position :
Perspective of each cabinet U Bit usage .
Visual weight bearing :
See through the load of each cabinet .
Visualization power consumption :
Perspective of the total power consumption of each cabinet , And then understand the energy consumption distribution of the computer room .
When combined with monitoring Visualization , It can be connected with the dynamic environment monitoring system PDU(Power Distribution Unit, Power distribution unit ) Monitoring integration , Cabinet real-time power distribution statistics and computer room PUE(Power Usage Effectiveness) The exhibition of . You can view the amount of energy consumption by different areas , Such as the floor 、 room 、 Cabinet for search and Statistics . Operation and maintenance personnel no longer need to infer through the original data to establish a psychological image , It's about using your senses to quickly understand the situation .
// Pipeline Visualization
Pipeline visualization is usually divided into network link visualization 、 HVAC duct Visualization 、 Visualization of pipeline types such as electrical circuit visualization . If you use traditional manual modeling , Usually the cost is higher 、 The implementation cycle is long , And the visual content built up is of little significance in the scene , And the direction of visualization is to monitor business data , It's not the actual pipeline arrangement .
therefore HT This paper presents a unique algorithm based on pipeline visualization to generate pipelines connected between devices , Including network interface 、 HVAC lines ( Water cooled air conditioning 、 Water tower 、 water chilling unit 、 Cooling pump 、 Constant humidity machine ) Artificial modeling or data generation )、 Electrical lines ( transformer 、 Power Distribution Box 、 The battery pack 、 Battery cabinet 、 Power monitoring communication cabinet 、 Switch cabinet 、 Main control cabinet ) And other link visualization functions . The system can be connected with network line 、 electrical 、 HVAC automatic control system for data docking , Automatic pipeline generation by algorithm , Display the running state and connection state of the equipment in the form of visualization and animation , Therefore, the connection relationship and link direction can make the operation and maintenance personnel clear .
// Dynamic environment monitoring Visualization
Communication power supply and machine room environment monitoring system ( Referred to as moving ring monitoring system ) It is for the power cabinets distributed in each machine room 、UPS、 Air conditioner 、 Battery and other power equipment and door magnets 、 infrared 、 Broken window 、 Water immersion 、 Temperature and humidity 、 Remote measurement of various parameters of the machine room environment such as smoke detector 、 Telesignaling 、 Remote adjustment and remote control , Real time monitoring of its operating parameters , Diagnose and handle faults , Record and analyze relevant data , A computer control system for centralized monitoring and maintenance of equipment . Monitoring center of the moving ring monitoring system 、 Monitoring sub center equipment includes database server 、 Business management server 、 Terminal equipment and network access equipment, etc , The monitoring station consists of a field monitoring and acquisition unit 、 Monitoring module 、 Network communication equipment 、 Various types of sensors 、 Transmitter, etc . When the monitoring station is not equipped with integrated video monitoring system equipment , Visual monitoring unit can be set , Including image acquisition 、 Processing unit, etc , Communication channel adopts IP Network or 2M Special line transmission mode .
(1)UPS monitor
Monitor the parameters and status of the equipment , Parameters include input and output voltage 、 electric current 、 power 、 The voltage of the battery pack 、 Temperature etc. ; States include rectifiers 、 inverter 、 The battery 、 The state of the load and other components , Display and record the change curve of each parameter , And record and deal with all kinds of alarm state . It can realize the real-time power distribution statistics of cabinet according to different areas ( floor 、 room 、 Cabinet ) Check and count the energy consumption of equipment .
(2) Three dimensional thermal cloud image
Through docking the Internet of things devices in the data center ( Temperature and humidity sensor ), Get the temperature information in the scene , Render a three-dimensional thermal cloud effect . At present, many data centers have installed and used low, medium and high vertical sensors , The traditional thermal cloud image is to attach the rendered image to the flat floor , Using a three-dimensional thermal cloud image , Can bring a more intuitive visual experience , It can locate the heating condition of equipment on each cabinet faster .
(3) Simulation of airflow perception
According to the data from the temperature and humidity sensors , Form a real-time air distribution map CFD (Computational Fluid Dynamics), Make it clearly see the flow direction and distribution of cold and hot air in the machine room . So that the indirect evaporative cooling scheme can maximize the use of natural cold sources , Not only can “ Refrigeration on demand ”, It can better meet the evaluation index .
(4) Energy saving monitoring
Green energy saving 、 Sustainable development is the key to the future development of the data center industry “ The main theme ”, Issued by the state 《 Guidance on accelerating the establishment and improvement of a green, low-carbon and circular economic development system 》 It is clearly pointed out in :“ We should accelerate the continuous upgrading of green infrastructure , Do a good job in large and medium-sized data centers 、 Green construction and transformation of network computer room , Establish a green operation and maintenance system .” The three aspects of the largest power consumption in the data center are the computer room air conditioner 、IT Equipment and UPS Power Supply , Up to... Of the total power consumption of the data center 93%, therefore , The energy-saving monitoring of the data center should also focus on these three aspects , meanwhile , In the future, the data center will change from a traditional data center to a computing power center , The industry is more concerned about the cost per unit of computing power 、 Energy consumption per unit of computing power and customer experience , The data center construction mode has also changed from resource driven to innovative value driven .
Real time statistics of annual energy-saving electricity and energy-saving benefits .
(5) Access control
The access control system is set to separate the public area from the controlled area , Swipe your card 、 password 、 The fingerprint 、 The traditional manual release is replaced by vein . Install access card reader or personnel access gate at the entrance and exit to be controlled , Set permission control points in the technical defense center , Set the access control authority for each cardholder . The basic function of the access control system is to control all important areas 、 The access doors of important places and the main passageways shall be monitored and controlled .HT Based on HTML5 Standard component library , It's seamless HTML5 Various multimedia functions , Support the integration of various video resources to form a unified video stream , Can be found in 2D、3D Mark the camera object on the situation map and associate it with its video signal source , Call the corresponding monitoring video through scene interaction , Meet the real-time situation awareness of the scene by the operation and maintenance personnel 、 Historical data backtracking comparison 、 Emergency response plan and other monitoring requirements .
(6) Early warning
The system has perfect fault warning 、 Event self diagnosis 、 Analysis and other functions , For a performance index system that exceeds the performance threshold , Be able to give fault alarm or early warning and inform the corresponding operation and maintenance management personnel , And do to save historical information and alarm events . Before equipment failure and other hazards that need to be guarded against occur , According to the laws summarized in the past or the possible precursors observed , Send emergency signals to relevant departments , Report hazardous situations , So as to avoid the hazard occurring without knowledge or insufficient preparation , So as to minimize the loss caused by the harm .
Although Guizhou's economy is not very developed , However, with its unique natural advantages, it has become one of the provinces in which major Internet technology enterprises build data centers . More Than This , Due to its underdeveloped economy, Guizhou has issued a series of preferential policies , This is undoubtedly not a good way to save costs for many enterprises who want to build data centers .
Now , More and more Internet technology enterprises have settled here , What's more, Guizhou's digital economy is growing 、 Rise one after another , It has ranked first for many years .
边栏推荐
- 2021-04-15
- Tinder security cooperates with Intel vPro platform to build a new pattern of software and hardware collaborative security
- Wechat applet pop up the optional menu from the bottom
- 【课程预告】基于飞桨和OpenVINO 的AI表计产业解决方案 | 工业读表与字符检测
- Edge and IOT academic resources
- [Course preview] AI meter industry solution based on propeller and openvino | industrial meter reading and character detection
- One way linked list implementation -- counting
- 微信小程序之flex属性
- Detailed explanation of kubernetes log monitoring system architecture
- How to install the DTS component of SQL server2008r2 on win10 64 bit systems?
猜你喜欢

OpenVINOTM 2022.1中AUTO插件和自动批处理的最佳实践

Input adjustment of wechat applet

php接收和发送数据

微信小程序之flex属性

Quartus call & Design d Trigger - simulation & time sequence Wave Verification

实战 | 如何制作一个SLAM轨迹真值获取装置?
![[deeply understand tcapulusdb technology] tmonitor background one click installation](/img/0a/742503e96a9b51735f5fd3f598b9af.png)
[deeply understand tcapulusdb technology] tmonitor background one click installation

From the establishment to the actual combat of the robotframework framework, I won't just read this learning note

Technology sharing | do you understand the requirements of the tested project?

Shell process control - 39. Special process control statements
随机推荐
【深入理解TcaplusDB技术】单据受理之表管理
SQLserver2008r2安装dts组件不成功
Intel ® extensions for pytorch* accelerate pytorch
When pandas met SQL, a powerful tool library was born
Face registration, unlock, respond, catch all
深入剖析MobileNet和它的变种
如何正确计算导入Excel的行数(POI/NPOI)
php接收和发送数据
Groovy map operation
Win10 64位系统如何安装SQL server2008r2的DTS组件?
Linear regression analysis of parent-child height data set
vulnhub靶机Os-hackNos-1
微信小程序之下拉菜单场景
Error when Oracle enters sqlplus
IEEE Transaction期刊修改过程记录
通过 OpenVINO Model Server和 TensorFlow Serving简化部署
2022 college entrance examination quarterly essay winners announced
MIT 6.031 Reading5 : Version Control学习心得
How to install the DTS component of SQL server2008r2 on win10 64 bit systems?
Detailed description of Modelsim installation steps