What are the development trends of college data centers in the future?

In July 2021, the Ministry of Education and other six departments issued the "Guiding Opinions on Promoting the Construction of a High -quality Education Support System for Promoting New Education Infrastructure Construction and Construction of High -quality Education Support System", marking the official ones of "Education New Infrastructure".

The guidance pointed out in the "Construction of a New Data Center" that the provincial education administrative department supports the construction of education cloud through a hybrid cloud model to provide convenient and reliable computing storage and disaster recovery services for educational institutions in the region. Planning and integrating the data center of the "low and small scattered" of education administrative departments and schools, and not encouraging county -level education administrative departments and primary and secondary schools to build data centers. Encourage regional and universities to share super computing resources and artificial intelligence computing resources, and provide basic computing power tools.

In the "Education New Infrastructure", the new data center shoulder the role of platform -type infrastructure. This role will further promote the construction speed of the current transformation of "cloudization" of college data centers. In general, the construction of data centers in universities in the future should focus on the following development trends:

1. "marginalization" of college data centers

The "marginalization" here does not refer to unimportant, but on the basis of ensuring localization of localization, college data centers will be given more marginal computing attributes. From the perspective of the national integrated big data center construction system, future data centers will develop in two directions. On one side is the large -scale and large -scale cloud data center relying on the cloud computing platform, and on the other hand, the edge data center close to the application side.

Among them, large -scale and ultra -large -scale data centers mainly solve the operations and storage problems of massive data. The edge computing data center mainly meets business needs such as ultra -low latency, high -real time, high security, and localization. The development of the two directions is only different functional positioning, not isolated. The two supplements each other to jointly promote the development of Internet data processing business.

In the future, the development of college data centers will be more flexible. It can provide private cloud services internally, but also integrates public cloud services in a "mixed cloud" method, and has both edge computing attributes. This is determined by the following factors:

(1) With the increasing maturity of the public cloud business, universities will migrate more business to the cloud in the future;

(2) Some important businesses need to be deployed with the help of "mixed clouds" to achieve the business "dual -living";

(3) From the perspective of security, colleges and universities need to have stable data centers to ensure that important data assets such as student status archives, scientific research data, and personnel wages are operated in private clouds;

(4) In terms of business stability, for interactive teaching, immersive experience, intelligent security, etc., such as high -delay and bandwidth business requires the help of local marginal calculations.

Although the data center of universities in the future is also called big data centers, under the architecture of the national integrated big data center system, its trend of calculating the development of data centers to the edge will be more obvious.

Second, college data center collection about sharing formula

We can basically be interrupted from the spiritual connotation transmitted by the new infrastructure of education: in the future, the construction route of college data centers will be the inevitable trend.

In colleges and universities, in addition to the data centers and computing centers of the unified planning and construction of the school, the phenomenon of independent construction computing resources such as departments, research institutes, and laboratories is more common. In the future, the systemic development of big data centers is the general trend. The information departments of colleges and universities should exert subjective initiative and explore the road of data center integration and development. Eliminate blind construction and repeated construction, achieve optimized allocation of resources, and improve resource utilization.

Specifically, by constructing a new type of data center in colleges and universities, the sharing of super computing resources and artificial intelligence computing resources in the school, and providing basic computing power tools for scientific research in the school. On this basis, further improve the platform construction system such as data centers and business processes to meet the needs of business departments open scenarios; build a data from the data to form a unified and authoritative data source; promote collaborative office and resource sharing, enhance school teaching and scientific research research Level and comprehensive governance capabilities.

At present, when colleges and universities formulate the "Fourteenth Five -Year Plan" information development plan, they basically follow the above development ideas. But in fact, the connotation of intensive shared development concepts is more broad.

The guidance of "Education New Infrastructure" proposes to "encourage regions and colleges and universities to share super computing resources and artificial intelligence computing resources, and provide basic computing power tools". At the same time, "support the provincial education administrative department Regional educational institutions provide convenient and reliable computing storage and disaster recovery services. " If selective reading, it is often easy to ignore keywords such as "region" and "region". Through the establishment of large -scale supercomputing resources and artificial intelligence computing resources, it provides shared basic computing power tools for educational institutions in the region, which is precisely a higher -level value appeal in the education field to promote computing resources.

Objectively speaking, compared with the sharing of computing resources in the school, the regional computing power sharing is not easy. Before the formation of a large -scale effect and substantial benefits, colleges and universities lack internal cause driving force. At the same time, this is also beyond the scope of the university itself, not one or more universities that can be promoted. Therefore, in the future, the provincial or municipal education administrative departments will also need to study the specific implementation methods of resource sharing and intensive development based on the guidance of new infrastructure in educational infrastructure, provide supporting policies support, and promote college intensive shared data centers from top to bottom. Construction.

Third, college data center green energy -saving small miniature

The data center needs to run uninterrupted 24 hours, which is a typical high -energy consumption industry, of which electricity costs account for more than 50%of the operating cost of data center. With the development of 5G, artificial intelligence, and the Internet of Things, data demand has grown rapidly, the number of data centers has increased day by day, and the scale of electricity consumption has also risen sharply. Therefore, energy -saving carbon reduction in data centers has become an important factor in promoting our country to achieve the goal of "dual carbon". The three departments such as the Ministry of Industry and Information Technology put forward in the "Guiding Opinions on Strengthening the Construction of Green Data Center" to vigorously promote the creation of green data centers, operation and maintenance, and transformation, guide the data center to take high -efficiency, clean, intensive, and circular green development paths to achieve data centers to achieve data centers Sustainable and healthy development. And for the new data center and the green transformation of old data centers, they have proposed more complete design specifications and indicators.

In the context of "energy consumption dual control" and "dual carbon", the state has continuously introduced policy optimization of the "old small scatter" stock data center, especially the optimization of key indicators such as PUE and cabinet power density. There is no doubt that colleges and universities will actively integrate this process in using data center transformation. With the help of new technologies and software and hardware equipment, it is not only a policy requirement, but also an objective needs for the development of data centers.

Like most electronic products, the life cycle of IT devices is usually 3 to 5 years, and the power density can be doubled every 5 years. To ensure good compatibility and scalability, the life cycle span of the data center infrastructure is 10 to 15 years. With the new and update of IT products, IT devices will continue to develop to high density in the future, and the computing and storage capacity will continue to grow. This allows data centers to continue to develop towards high density, and high density means saving greater space.

Therefore, in the future, the transformation of the use of data centers will focus on technological upgrades, not the expansion of the computer room. Coupled with the needs of green energy conservation and flexible management, the future data centers of universities will move towards the road of green, energy saving and miniaturization.

In summary, the development trend of college data centers in the context of new infrastructure is becoming clearer and clear. As an information management personnel, we should strengthen policy interpretation, think about the current status and future of the industry, and grasp the development trend of college data centers. Avoid or take less detours when formulating development planning to promote faster and better development of colleges and universities.

*This article is one of the research results of Jiangsu Modern Education Technology Research Smart Campus Special Project (Project Number: 2021-R-96738)

Author: Bian Yongtao (China University of Mining and Technology)

- END -