“Due to the complexity of scenarios, different load characteristics, and different user business characteristics, edge computing presents diversified characteristics in the application of power supply and distribution systems.
Due to the complexity of scenarios, different load characteristics, and different user business characteristics, edge computing presents diversified characteristics in the application of power supply and distribution systems.
In the face of complex edge computing scenarios, what kind of power supply and distribution solution should be chosen, since it can ensure high reliability, it can also reduce TCO? Is there a standardized power supply and distribution solution that can match the global edge computing scenario?
In response to these problems, Li Xu, general manager of the ACP product department of Vertiv Technology (Vertiv, NYSE: VRT) in China, presented in the third “V-LoGeek 100 Lessons”-Vertiv Technology (Vertiv) 2020 online seminar Give a comprehensive interpretation.
Intricate edge computing scenarios
From the perspective of the development process of the data center, it has gone through three stages, namely cloud computing, edge computing and edge-cloud collaboration.
As we all know, as a large computing power platform, cloud computing is aimed at global business, providing non-real-time large-scale storage and large-scale computing. Edge computing provides real-time processing of local services to meet low-latency service requirements, and emphasizes rapid decision-making and rapid execution of customers’ local services.
So, what is edge collaboration?
From the perspective of customer business, it is to allocate and coordinate the computing power of cloud computing and edge computing accordingly, which will bring the effect of one plus one greater than two. For example, if cloud computing is compared to strategy, then edge computing is like tactics. Only when strategy and tactics match can we win the entire battle.
According to a third-party research report, by 2022, the global edge computing business scale will reach 6.7 billion U.S. dollars, and this figure will maintain a rapid growth of more than 30%. Data shows that with the continuous development of digital business, 75% of enterprise-generated data will be created and processed in locations other than traditional centralized data centers or the cloud, and 84% of enterprises are incorporating edge computing into their corporate plans.
In view of the current intricate application scenarios of edge computing, Vertiv has scientifically classified four major technical scenarios based on key customer needs based on years of research and practice, namely data-intensive, experience delay-sensitive, and machine delay-sensitive. , Life support type, and each technical scenario includes multiple subdivided business scenarios.
At the same time, in accordance with the corresponding needs of different edge computing technology scenarios, Vertiv has introduced integrated, standardized, and modular data center solutions, which are used in education, agriculture, HPC, manufacturing, power and other industries. Edge computing scenarios have been widely used.
Edge computing requires higher reliability guarantee
From the perspective of users, the users of edge computing are mainly divided into three categories: telecom operators, industrial markets, enterprises and the Internet of Things.
Users in almost all industries have huge business needs for edge computing, and while maintaining their own business characteristics, they also have great commonalities in edge computing needs.
First of all, the huge demand for edge computing has also brought diversification of application scenarios. Secondly, for many companies, edge computing services often do not have sufficient planning time, and their business requirements are generally non-linear. For sudden edge computing services, it is necessary to achieve rapid response and rapid delivery, and optimize TCO. In addition, unified management and simplified operation and maintenance are also very important for edge computing scenarios.
The development of edge computing requires the support of critical infrastructure. With the widespread deployment of edge computing, there are many pain points in critical infrastructure applications, including wide distribution, diversified demand types, lack of professional delivery capabilities, and flexibility and low cost. Important core requirements for critical infrastructure applications.
In comparison, the operating environment of edge computing is more severe than that of cloud computing data centers, because edge computing is usually in a non-standardized computer room environment, which cannot guarantee that key infrastructures such as power supply and distribution systems and air conditioning can maintain ideal operating conditions.
Take the power supply and distribution architecture as an example. Generally, 2N architecture is rarely used in edge computing, but N+1 redundancy design or even stand-alone power supply architecture is adopted. This is due to the harsh operating environment of edge computing, on the other hand, There are fewer and fewer power supply redundancy backups for edge computing.
In this regard, how to design and select the corresponding power supply and distribution solutions?
Let’s first understand what factors are affecting the security of the data center. According to a third-party survey report, in the statistics of data center downtime reasons, UPS system failures accounted for 29%, and 24% were caused by human operating errors, and most of the operating errors occurred in the power supply and distribution system.
If the direct failure of the UPS system and human operation errors are included, the overall power supply and distribution system affects the operating safety of the data center accounted for more than 50%. In other words, the safety and stability of the power supply and distribution system plays a vital role in the continuity of business development.
Then analyze what factors are affecting the reliability of the UPS system. Summarizing the reasons for the failure of small and medium UPS systems through big data, it can be found that the factors of harsh environment and PCB fouling and corrosion accounted for 51.9% of the reasons for UPS system failures, which is consistent with the harsh physical environment in the edge computing scenario.
At the same time, fan failure, rectifier failure, inverter failure, and relay failure will all affect the reliability of the UPS system. Therefore, to provide stable power supply and distribution guarantee for edge computing, first of all, we need to pay attention to the reliability of the power supply and distribution system itself.
According to the analysis of the factors affecting the reliability of the UPS system, to improve the reliability of the UPS system, it is necessary to effectively deal with the impact of wet and dusty edge computing scenarios, inverter bypass switching, high temperature & heavy load scenarios, mechanical vibration scenarios, and operation and maintenance capabilities. Challenges posed by system reliability.
In response to these challenges, Vertiv has given targeted designs in the development of its UPS products, such as independent air duct design, strict wet dust testing, parallel SCR output relays, and increased output power to 1, 9 Intensity seismic design, active maintenance, etc.
How to achieve reliable compatibility with TCO
For edge computing scenarios, it is not only necessary to pursue high reliability, but also to pursue lower TCO. Modular data centers have become the first choice for edge computing construction solutions and can achieve the best TCO.
Modular data centers can provide targeted solutions for the demands of different scales and different businesses. Through prefabricated integration and standardized delivery, delivery quality and delivery efficiency can be improved, and operation and maintenance interfaces can be unified.
At present, modular data centers face three major problems in the construction of power supply and distribution systems: low integration, simple patchwork according to space and functional partitions, fragmented communication between system equipment, and a little chainless; complex engineering, physical connection workload between equipment Large, the on-site power intelligent upgrade is difficult; inefficient operation and maintenance, equipment operation status is difficult to grasp, and the system equipment monitoring interface is not unified.
In the face of these problems, the demand for smart converged power supply and distribution solutions in modular data centers is becoming more and more urgent. Its primary requirement is to require cost optimization, step-by-step investment, and increase the rate of return on investment.
At the same time, it can realize the intelligent integration of software and hardware, highly integrated, easy to install, and can be expanded and changed flexibly in actual operation. In addition, unified management to simplify the operation and maintenance of a core demand for the power supply and distribution system of the modular data center.
In response to the new development trend and application requirements of the power supply and distribution system, Vertiv has continuously launched targeted innovative products and solutions based on years of in-depth practical experience in the industry.
Among them, the newly launched three-in-one intelligent power module can be called the optimal TCO integration solution for medium and large modular data centers. It combines the input power distribution cabinet, UPS and output precision power distribution cabinet into one. With integrated intelligent integration, 67% of the land occupation and installation time can be saved, and the cable cost between the cabinets is “0”, and all power distribution can be intelligent and refined load management.
At the same time, this product is a typical full power chain intelligent integration, full power chain management does not miss any key nodes, breaking the product boundary between power distribution and power supply.
For more small edge computing scenarios, Vertiv has also introduced a targeted low-power UPS with independent air duct design, output power factor of 1, and strong adaptability.
In response to the critical production and high reliability requirements of distributed power supply in industrial edge computing scenarios, Vertiv has launched a highly efficient and flexible integrated UPS designed with a built-in output isolation transformer, which has higher system reliability.
It is worth mentioning that Vertiv has also perfectly solved the problem of how to balance the high efficiency of the bypass with the high reliability of the double conversion by launching the VI dynamic online mode. Compared with the traditional ECO mode, the VI dynamic online mode has a qualitative improvement in reliability, comprehensive performance of energy efficiency, and short-circuit protection.
In addition, Vertiv has been actively exploring the application of lithium batteries and has launched HPL lithium battery cabinets, which can provide high-reliability lithium battery solutions, as well as more advantageous charge and discharge cycles and a smaller footprint .
Based on the analysis of complex edge computing scenarios, it is helpful to choose the most suitable power supply and distribution solution to meet business needs, and achieve the ultimate goal of not only ensuring the reliability of the system, but also achieving the optimal TCO.