1. Introduction to the Project

1.1 Project background

1.1.1 Product introduction

In order to promote high-quality development, we are committed to strengthening the construction of digital infrastructure, breaking through key core technologies, accelerating industrial digital transformation, enhancing the cultivation of digital economy talents, and optimizing the development environment of the digital economy. We are building a computing power center, constructing a social production platform and a creative content visualization platform, and building professional technical and algorithm reserves. In addition, we are also improving the support capability system for cloud video, CG creation, digital creativity, VR content production, education business and other businesses, building AI algorithm libraries, real-time rendering algorithm sets, and AI application module sets.

1.1.2 Market prospect

(1) Current situation of the computing power industry

As of the first half of 2024, the number of completed and under construction intelligent computing centers in China has exceeded 250, and the construction quantity and investment scale of intelligent computing centers have shown a rapid growth trend. For example, Beijing, Shanghai, Guangdong, Hebei and other places have already released development plans that include 2025. It is expected that by 2025, Beijing will have basically formed a pattern of clustered supply of intelligent computing resources, autonomous construction of intelligent computing facilities, precise empowerment of intelligent computing capabilities, green operation of intelligent computing centers, and systematic development of intelligent computing ecology.

The demand for the construction of computing power centers is rapidly increasing, especially in fields such as artificial intelligence and big data processing. As a public computing infrastructure, the intelligent computing center provides common computing power, data, and algorithm services, making intelligent computing a basic public service like hydropower. The core computing unit of the intelligent computing center adopts advanced artificial intelligence chips, which can significantly improve the efficiency of basic computing power utilization and algorithm iteration.

Edge computing has started large-scale practical applications in industries, government affairs, transportation, finance, the Internet, energy, retail, entertainment, agriculture and other fields. The traditional industry uses edge computing to deploy business applications locally, effectively reducing business latency and costs. At the same time, it combines with existing fragmented IT systems to achieve integrated governance and intelligent upgrading of cloud edge, enabling the industry to deepen digital transformation. The combination of Internet applications and edge computing technology has incubated various new Internet industry services with audio and video solutions as the main line, such as cloud games, cloud e-sports, AIGC reasoning, etc.

The industrial map of edge computing is constantly enriched. A number of edge open source projects have emerged in the industry. The standard system of edge computing overall architecture, key technologies, connectivity and other aspects at home and abroad is increasingly mature. The industrial chain cooperates and cooperates to explore and create a scenario oriented integrated whole stack edge solution to accelerate the large-scale promotion of edge applications.

(2) Market prospects of the computing power industry

With the promotion of the demand for industrial application scenarios, edge computing technology is constantly integrating with new technologies such as hardware, cloud computing and AI. The edge infrastructure that integrates software and hardware exhibits a highly integrated architecture to meet the diverse computing power needs of edge businesses; edge computing of the Internet of Things efficiently connects massive terminals, realizes cloud edge integration, and enables “intelligent connection of all things”; edge cloud native management of distributed heterogeneous resources, building dynamic, agile, and secure edge businesses; edge AI builds a service model of “central training, edge inference”, and continuously innovates lightweight large-scale models at the edge, which not only reduces inference latency and costs, but also enhances data security; the edge computing power Internet continues to explore, breaking through the perception, management and scheduling problems facing diverse and heterogeneous edge computing power.

Looking into the future, edge computing will develop in the direction of wider distribution, more intelligent management and more diversified services. With the large-scale application of “AI+” action in China, AIGC, the continuous innovation of big model technology will significantly increase the demand for AI inference computing. Edge computing, with its advantages of “low latency, low cost, wide distribution, and high security”, breaks through the application bottleneck of intelligent computing power in cross-architecture, cross-region, and cloud side collaboration scenarios through innovative frameworks such as “training while pushing”, and provides a solid support for the large-scale development of artificial intelligence technology.

Thanks to the high growth in computing power scale, the global computing industry has also maintained a steady growth momentum. In recent years, the development of China’s computing industry has benefited not only from the driving force of technological revolution, but also from the relevant policies that encourage the development of artificial intelligence and computing infrastructure. Layout from top-level planning, facility construction, industrial development to management standards, and guide the high-quality development of the entire computing industry, including artificial intelligence.

The industry analysis believes that with the rapid development of artificial intelligence technology, higher requirements have been put forward for the development of intelligent computing power. Wang Juncheng, Director of the Institute of Informatization and Industrialization at the China Academy of Information and Communications Technology, stated that with the advent of the artificial intelligence era, the global intelligent computing industry has achieved rapid growth, and systematic innovation in areas such as computing, storage, interconnection, and software is the main path for upgrading the computing power industry. To achieve the upgrading of the intelligent computing industry, it is necessary to strengthen the systematic upgrading of the industrial chain, accelerate the ecological construction of the computing power industry, and promote the green development of computing power.

According to the China Academy of Information and Communications Technology, in the era of deep learning, the proportion of global intelligent computing power is less than 10%; in the era of big models, the proportion of global intelligent computing power has exceeded 60%; with the advent of the era of general artificial intelligence, it is expected that the global proportion of intelligent computing power will exceed 90%.

In the future, all computing devices will have AI capabilities, and CPU driven general-purpose servers will assume more diverse computing roles in AI applications, working together with GPUs and other accelerators to form heterogeneous computing systems that can more efficiently process AI tasks and provide more efficient intelligent computing services for cloud edge devices and industry intelligence scenarios.

1.1.3 Technical analysis

System architecture design: The construction of a computing power center needs to be designed around four main parts: foundation, support, functionality, and objectives. The basic part includes advanced AI theory and computing architecture; the supporting part involves the logic of computing power production, aggregation, scheduling, and release; the functional part provides platforms for computing power production and supply, data openness and sharing, intelligent ecological construction, and industrial innovation aggregation; the overall goal is to promote the industrialization of AI, industrial AI, and intelligent government governance.

Technological evolution: The development of computing power centers is based on the latest AI theories and computing architectures, with key technologies including AI chips, AI servers, and AI clusters. AI chips are the mainstream mode of current AI computing, and mainstream AI acceleration computing uses CPU systems equipped with heterogeneous acceleration chips such as GPU, FPGA, ASIC, etc.

Fusion computing technology: Data centers are expanding from a single scale to complex architecture fusion, and server architectures are gradually shifting from CPU centric to heterogeneous fusion computing, such as introducing heterogeneous computing accelerators such as GPUs, TPUs, ASICs, or FPGAs. This design concept transcends the boundaries of hardware and software, considering the entire data center as a collaborative system to improve efficiency.

1.1.4 Advantageous conditions of project construction

(1) Policy advantages

In recent years, China’s edge computing related policies have been successively introduced. The Action Plan for the High-quality Development of Computing Infrastructure proposed to “promote the coordinated deployment of edge computing, and promote the widespread distribution and coordinated development of ‘cloud edge’ computing power”. With the increase of national policy support, the construction of edge computing has entered a period of rapid growth. The application scenarios are gradually enriched, the market size is growing, and edge computing is moving towards a new stage of development.

(2) Industrial advantages

The industrial layout and policy support in Jilin Province provide a favorable development environment for the computing power center project. Jilin Province is implementing the “One Main and Six-Double” high-quality development strategy, vigorously cultivating the “Four Major Clusters”, developing the “Six New Industries”, and constructing the “Four New Facilities”, constantly seizing new fields and tracks, and building new advantages. This strategic layout provides broad space and strong support for the development of computing power center projects. In addition, Jilin Province promotes the deep integration of the real economy and the digital economy through reforms, improves the mechanism for promoting the construction of “Digital Jilin”, further reduces costs, expands applications, and improves efficiency, providing computing power support for industrial innovation, “smart transformation and digital transformation”, and digital government construction.

(3) Location advantages

Lishu County is located in the southwest of Jilin Province, in the hinterland of the Songliao Plain, with convenient transportation. The main transportation arteries include the Beijing-Harbin Railway, Siping-Meihekou Railway, Harbin-Dalian Railway, Changchun-Dalian Expressway, and National No. 102 Highway, all of which pass through the county. There are three railway stations in the county, with an annual throughput of over 3 million tons of goods. In addition, Lishu County is only 10 kilometers away from the transportation hub of Siping City in Northeast China, 100 kilometers away from the provincial capital Changchun, and 560 kilometers away from Dalian Port. It is the closest county to the sea in the north of Liaoning, Northeast China. ‌‌

(4) Talent advantages

Siping has some universities including Jilin Normal University, Boda College of Jilin Normal University, Jilin Agricultural Engineering Vocational and Technical College, Siping Vocational University, as well as the ongoing establishment of Hongzui Vocational and Technical College, Siping Medical College of Jilin University, and Siping Radio and Television University. Among them, Jilin Normal University is a key university in Jilin Province, with 16302 full-time undergraduate students, 1844 graduate students, 336 international students, and 15131 adult education students. It can cultivate thousands of higher education talents to the society every year. Surrounding cities such as Changchun, Harbin, and Daqing have numerous universities that can provide over 500000 high-level talents annually. These talents can provide technical support for project construction.

1.2 Contents and scale of project construction

Covering an area of 16000 square meters, it will construct the data center, operation and maintenance center, monitoring center, substation, etc.

Build a social production platform and a creative content visualization platform, build professional technical and algorithm reserves, improve the support capability system for cloud video, CG creation, digital creativity, VR content production, education business and other businesses, and construct an AI algorithm library, real-time rendering algorithm set, and AI application module set.

1.3.Total investment of the project and capital raising

The total investment of the project is 200 million yuan, including the construction investment of 150 million yuan.

1.4 Financial analysis and social evaluation

1.4.1 Main financial indexes

After the project reaches the production capacity, its annual sales revenue will be 122 million yuan, its profit will be 37 million yuan, its investment payback period will be 7 years (after the tax, including the construction period of 1 years and its return on investment will be 18.5%.

Note: “10 thousand yuan” in the table is in RMB.

1.4.2 Social evaluation

Promoting technological innovation and economic development: The computing power center accelerates the iterative optimization of machine learning and artificial intelligence algorithms by providing powerful computing power, significantly improving research and development efficiency. For example, researchers can conduct large-scale experiments and simulations to accelerate the development of new materials and drugs. In addition, the construction and development of computing power centers have promoted applications in multiple fields such as intelligent manufacturing, smart cities, smart healthcare, and financial technology, driving innovation and development in the industry and bringing new business opportunities to enterprises.

Promoting industrial transformation and upgrading: As a new quality productivity in the era of “artificial intelligence+”, computing power centers are profoundly changing the structure and production methods of traditional industries. For example, the Beijing Super Cloud Computing Center provides high-quality computing power construction, platform services, and computing power operation services to assist enterprises in digital transformation and promote the high-quality development of the industry.

Improving data support and decision-making efficiency: The computing power center integrates and shares various data resources to enhance information processing efficiency and ensure the accuracy of decision-making basis. For example, in urban traffic management, the computing power center helps the government optimize traffic signal settings and reduce congestion by analyzing traffic flow data. In addition, the computing power center also promotes the provision of personalized intelligent services, conducts in-depth analysis of user data, helps enterprises understand customer needs, and provides customized products and services.

1.5 Cooperative way

Sole proprietorship, joint venture or cooperation

1.6 What to be invested by the foreign party

Funds, and other methods can be discussed in person.

1.7 Construction site of the project

Jilin Lishu Economic and Technological Development Zone

1.8 Progress of the project

The project proposal has been prepared.

2. Introduction to the Partner

2.1 Basic information

Name: Jilin Lishu Economic and Technological Development Zone

Address: East Section of Fenghua Street, Lishu Town

2.2 Unit overview

Jilin Lishu Economic and Technological Development Zone is located in Lishu County, Siping City, Jilin Province. It was established in October, 2005 and is a provincial-level development zone approved by the Development Office of Jilin Provincial Government. The total planned area of the development zone is 14.2865 square kilometers, divided into five blocks, with the specific area of 3.3049 square kilometers in Block 1, 4.3326 square kilometers in Block 2, 0.5607 square kilometers in Block 3, 0.4136 square kilometers in Block 4, and 5.6747 square kilometers in Block 5.

Six major industrial parks have been planned and constructed within the development zone, including the livestock product processing industrial park, pharmaceutical processing industrial park, high-tech industrial park, agricultural and sideline product processing industrial park, means of production industrial park, and mechanical processing industrial park. These parks have attracted a large number of enterprises to settle in, forming a diversified industrial layout.

In addition, the development zone also provides services such as factory rental and sales, warehouse rental and sales, and company registration, and enjoys preferential policies provided by Siping Municipal People’s Government, including the application for key enterprise support funds and the notification of scientific and technological achievement evaluation pilot units, aimed at promoting enterprise development and improving the business environment.

2.3 Contact method

Contact unit: Jilin Lishu Economic and Technological Development Zone

Contact person: Liu Ning

Tel: +86-16643433555

E-mail: jllsjjkfq@126.com

Contact method of the city (prefecture) where the project is located:

Contact unit: Siping Economic and Cooperation Bureau

Contact person: Wen Dacheng

Tel: +86-434-3260536

E-mail: spjhjjhk@163.com