In recent years, with the rapid development of the digital economy and the widespread application of information technology, computing power has become an important support for social development.

The total scale of my country's computing power has grown at an average annual rate of nearly 30% in the past five years, reaching 1.97 trillion floating-point operations per second, ranking second in the world.

However, traditional computing power methods are often accompanied by high energy consumption and high emissions. Promoting the green and low-carbon development of computing power has become an industry consensus and a need of the times.

  Agency data shows that up to 30% of servers in commercial and enterprise data centers are in a "sleeping" state, consuming power but not providing useful information services.

In order to cope with this situation, some technology companies began to carry out technological innovation based on existing computing power.

  "A lot of data is 'cold data' that is never used, and only 5% of the 'hot data' is really frequently used, which means that a lot of computing power resources are actually wasted." He Zhengyu, chief technology officer of Ant Group, said that realizing computing Key technical elements of greening include supporting multiple supply chains, comprehensive cloud nativeization, moving to serverless, implementing platform engineering, and establishing a green continuous observation mechanism.

  In exploring the path of low-carbon and energy-saving data centers, the greening of the computing power supply chain has attracted much attention.

Relevant staff of Sugon said: "We have now built a complete intelligent manufacturing system solution that reduces energy consumption per unit output value by 30%, which is equivalent to reducing the burning of about 2 kilograms of standard coal for each server produced. For finance, electricity , manufacturing and other specific industries, we have also launched liquid cooling overall solutions to accelerate the construction of a better and sustainable green computing era."

  Massive and complex computing requirements are causing the overall power consumption of data centers to increase at a rate of more than 10% every year.

At present, the total electricity consumption of data centers nationwide accounts for 2% of the entire society, and electricity costs account for 50% of the total operating costs of data centers.

According to data from the China Academy of Information and Communications Technology, it is expected that by 2030, the total energy consumption of data centers will reach about 380 billion kilowatt hours.

  While improving computing efficiency, it is crucial to reduce the energy consumption of the equipment behind the data.

Recently, China Academy of Information and Communications Technology, National Energy Group Ningxia Electric Power Co., Ltd., Sugon, Beijing University of Posts and Telecommunications, China Telecom Co., Ltd. Ningxia Branch, etc. jointly released a new generation of distributed green energy intelligent computing integrated deployment action cooperation plan. Integrated liquid-cooled intelligent computing servers are placed in wind power towers in Ningxia, allowing data centers to directly use green power.

This method no longer requires a supporting power transmission system, greatly reducing losses in power transmission, thereby achieving optimal configuration standards for computing power settings, liquid cooling, and energy storage modules for this type of platform, and ultimately achieving high-end computing and efficient computing networks. The triple goals of operation and green energy saving.

  Reducing carbon emissions from the source of power production and increasing the proportion of renewable energy used in computing power infrastructure are becoming important supports for the overall green and low-carbon computing power industry chain.

  Walking into the edge of the Tengger Desert in Zhongwei City, Ningxia, you will be amazed by the dense photovoltaic power stations and the large white data center of the Western Cloud Base.

Ningxia is the national hub node of the national integrated computing network and the new national Internet exchange center. The western cloud base has built 14 large and ultra-large data centers, and the size of the data center racks in use reaches 67,000 standard racks.

At the same time, Ningxia is an important transmission channel for "West-to-East Power Transmission" and the country's first comprehensive new energy demonstration area.

At present, Ningxia's new energy installed capacity exceeds 36 million kilowatts, its new energy utilization rate ranks first in the northwest, and it is listed as the first batch of green electricity trading pilot provinces, which constitutes the convenient advantage of using green electricity.

  In order to promote the green and low-carbon development of the "East-to-West" project, State Grid Zhongwei Electric Power Company closely cooperates with government departments to carry out full-chain and full-cycle services from power grid planning, installation and connection to energy efficiency services.

Comprehensive data center layout, construction progress, local grid structure, power supply reliability requirements, etc., carry out scientific planning and appropriately arrange investments in advance.

Ningxia has also set up a special service class to formulate a data center power connection plan on a "one household, one case" basis. It also actively promotes cloud computing companies to be included in the scope of cross-regional and cross-provincial power transactions, and strives for preferential electricity prices and subsidy policies. It is expected to reduce electricity consumption for companies every year. The cost is more than 10 million yuan; real-time monitoring of data center load conditions, online analysis and visual display of equipment power consumption, cooling and other power structures and energy consumption levels, helping users formulate energy efficiency improvement strategies and helping users save energy and reduce consumption.

  Tuo Zhaobing

  (Economic Daily)