Hello! I'm your financial analysis assistant. Today, we're going to talk about a concept that might sound a bit technical, but it's actually closely related to everyone's future life: the Computing Power Network.
If we compare the internet to a “highway” for transporting information, then the Computing Power Network is like a “superbrain factory” that processes that information. Recently, the government has outlined a detailed plan and schedule for building this “superbrain factory,” and the investment effort is quite substantial.
Let me break down this news into five key points in simple language to help you fully understand the logic behind it.
1. The government has elevated “computing power” to the same strategic level as “power grids” and “water supply networks”
First, let's understand the context: in the past, when we talked about infrastructure, we thought of building roads, bridges, providing electricity, and ensuring water supply. Now, the government has proposed a set of “six networks,” one of which is the Computing Power Network.
What does this mean? It means that computing power is no longer just an internal asset of technology companies; it has become a basic public resource, just like electricity and water.
- Status upgrade: The Computing Power Network has been officially included in the country's 109 major projects for the 14th Five-Year Plan (2026-2030). It's like the government saying, “Over the next five years, we will focus on building this ‘digital highway.’ This is a national priority, not an optional task.”
- Core goal: To create a nationwide integrated computing power network. In simple terms, this means connecting computer resources across the country so that wherever computing power is needed, it can be quickly accessed, just like turning on a faucet at home to get water, without having to carry water from afar.
In one sentence: Computing power has officially become a national infrastructure, on par with water and electricity, and is the core of the new infrastructure for the country's future development.
2. Computing power needs to increase by six times within five years – this is a “arms race”
There are some astonishing figures in the news that we need to look at carefully:
- Current situation: By 2025, China's intelligent computing power was 1590 EFLOPS (a measure of how many billion calculations can be performed per second).
- Goal: By 2030, this number needs to reach 9800 EFLOPS.
- Growth rate: This means that the computing power will need to increase by 6.2 times over the five years from 2026 to 2030, with an average annual growth rate of 45%.
Why such a rapid increase?
It's because while people used to use their phones for basic tasks like messaging and watching videos, now they use AI for coding, designing, and even controlling robots.
- Surging demand: The training and operation of large AI models (like ChatGPT) require massive amounts of computing power. It’s like going from cooking with coal to building rockets; the demand for computing power is growing exponentially.
- Quality as well as quantity: Expert Zhu Keli pointed out that this is not just about building more data centers; it’s due to the explosive growth of the AI industry. Without enough computing power, AI systems won’t function effectively, just like a phone would be useless without electricity.
In one sentence: Because AI is so popular, the government wants to increase computing power by six times within five years as part of a major expansion project to support the AI era.
3. 4 trillion yuan in investment: Opportunities for the public and private sector
Where will the money for such a large project come from, and how much will it cost?
- Total investment: It is estimated that the construction of the Computing Power Network will require an additional 4 trillion yuan during the 14th Five-Year Plan period.
- Who will fund it: The National Development and Reform Commission has emphasized that the construction of the computing power network will be primarily funded by private enterprises.
What does this mean for individuals or investors?
- A huge market for private investment: Since the investment is mainly coming from enterprises, it will flow into areas such as technology, communications, and data center construction. For private businesses, this represents a vast market opportunity. The government provides the framework (setting standards and building infrastructure), while enterprises build the data centers, purchase chips, and provide services, allowing everyone to benefit.
- Industrial chain stimulation: This 4 trillion yuan will not only go into building data centers but also into chip manufacturing, server production, power supply, cooling systems, and network communications, driving the entire tech industry and creating many jobs.
In one sentence: With 4 trillion yuan in investment, led by private enterprises, the tech industry is set to experience a wave of significant investment, providing ample opportunities for private capital.
4. The “East Data, West Computing” initiative: Leveraging the West’s resources for the East
You may have heard of the “East Data, West Computing” initiative; this news discusses its advanced version:
- Problems: The East (e.g., Shanghai, Beijing) has high demand but high electricity costs, expensive land, and challenges with heat dissipation; the West (e.g., Guizhou, Gansu, Inner Mongolia) has cheap electricity, a cool climate, and ample land, but low local demand.
- Solution: The Computing Power Network allows the transfer of computing tasks from the East to the West for processing.
- Metaphor: The East is like a busy city center with many orders but limited space; the West is like a large warehouse in the suburbs. With the new superhighway (the Computing Power Network), orders can be sent directly to the western warehouses for processing and then the results quickly sent back to the city center.
- Current situation and layout: The country has already established an “8+10+3” layout (8 major hubs, 10 clusters, and 3 computing-electricity coordination areas). Eighteen hubs already handle 87.3% of the national intelligent computing capacity, indicating that the western regions have become a key force.
- Future trend: The plan aims to build a multi-level system of “hubs, regions, and edges.”
- Hubs: The western clusters handle heavy computing tasks like training large AI models.
- Edges: Local cities and factories can respond in real-time; for example, autonomous vehicles need millisecond-level responses, so data must be processed locally rather than being transmitted over long distances.
In one sentence: By using the West’s cheap electricity and space, the “East Data, West Computing” initiative aims to optimize resource allocation, saving money and being environmentally friendly.
5. Key variables: Domestic chips and the commercialization of AI
Finally, let’s look at two key factors that will determine the success of the Computing Power Network, as mentioned in the news:
1. Can domestic chips catch up?
- Chips (such as GPUs) are at the core of the Computing Power Network. Currently, high-end chips are mainly imported.
- The plan explicitly states the need to increase support for domestic chips.
- Interpretation: If domestic chip technology advances and production capacity expands, it will not only break monopolies but also significantly reduce costs, making computing power accessible to small and medium-sized enterprises. This is essential for industrial autonomy. If domestic chips fail to keep up, the Computing Power Network will be dependent on others.
2. Can AI become profitable?
- Even with a well-built computing network, it needs to be used and generate value.
- The report indicates that AI is shifting from experimental to commercial applications.
- Interpretation: If AI models remain just laboratory tools without commercial value, all the computing power will be wasted. Only when AI is applied in industries like manufacturing, healthcare, and finance to help businesses save money and improve efficiency will demand for computing power continue to grow.
Additionally, there’s the benefit of “inclusive computing power”:
- The plan includes initiatives to support small and medium-sized enterprises with affordable computing power. This means that in the future, these companies won’t need to spend millions on servers; they can rent computing power on demand, just like cloud storage, significantly lowering the barriers to entrepreneurship and innovation.
In one sentence: The success of the Computing Power Network in the future depends on whether domestic chips can overcome technical barriers and whether AI applications can be successfully commercialized. The government will also promote the widespread use of computing power, making it accessible to small businesses.
Suggestions for individuals:
- Pay attention to industry opportunities: If you work in or are interested in the tech industry, areas like data center construction, domestic chips, computing power management software, and green energy are promising fields for the next five years.
- Understand the new trend: In the future, computing power will be as important as data flow in measuring a company’s competitiveness. For individuals, learning to use AI tools and understanding the logic behind computing power will be part of new digital literacy.
- Investment perspective: Although the 4 trillion yuan in investment is significant, be cautious of “conceptual hype” and focus on companies with real technical barriers, orders, and the ability to solve domestic substitution issues, rather than just following the latest trends.
In summary, this news outlines a blueprint for China’s digital economy over the next five years. The construction of the Computing Power Network is not just about making AI more powerful but also about making society more efficient and cost-effective.