Written by: Web4 Research Center

Recently, the Ministry of Industry and Information Technology issued a notice to organize a special action aimed at empowering the development of small and medium-sized enterprises (SMEs) through inclusive computing power. For the first time, it explicitly proposes exploring innovative services such as "Computing Power Banks" and "Computing Power Supermarkets", supporting SMEs to deposit idle computing resources, achieving flexible usage through cross-regional and cross-period scheduling, thereby reducing computing costs. The notice states that by the end of 2028, a comprehensive inclusive computing service system will be established that is wide-ranging, low-cost, and offers high-quality service, covering no less than 10 main industry categories of SMEs.

The policy's signal is clear and strong. Meanwhile, on the front line of policy implementation, pioneers have already begun their practices—Shanghai Telecom's "Computing Power Supermarket" has been launched and is operational, while the Henan Airport Intelligent Computing Center has completed the deployment of the Token API interface to support SMEs in calling domestic intelligent computing power on demand.

At three a.m., at the Shanghai Qingpu Intelligent Computing Center, rows of cabinets illuminate with thousands of server indicator lights flashing like stars in the sea. But at this moment, what is truly harnessing this powerful computing power is not a massive model training for some AI unicorn with billions of parameters, but an industrial simulation task from a small manufacturing enterprise in Suzhou. It ordered 4 hours of intelligent computing card power through Shanghai Telecom's "Computing Power Supermarket", and the whole process took less than two minutes.

At the same time, in Zhengzhou Airport District, Henan, the Henan Airport Intelligent Computing Center's machine room is also bustling with activity. The computing power of this largest intelligent computing center in central China has reached 10,000 P. If all 7 billion people in the world were to compute non-stop for 40 years, they would only complete the computing volume of one second of it. Just weeks ago, this colossal facility had just completed the Token API interface deployment, allowing SMEs to call domestic intelligent computing power without making a huge upfront investment.

This is China in the spring of 2026. Computing power, the core production factor of the AI era, is gradually transforming from the "heavy weapons" of a few giants to the "daily necessities" accessible to SMEs.

1. A "Silent Revolution": When Computing Power Becomes as Accessible as Electricity and Coal

Computing power is the core production factor for SMEs' digital transformation in the AI era, akin to electricity and coal in the industrial era, directly determining the innovation capacity and market competitiveness of enterprises.

This statement is not a promotional phrase from a tech company but an accurate summary of the positioning of the computing power industry by Zhang Linshan, a researcher at the National Development and Reform Commission's Macro-Economic Research Institute. Indeed, applications of large models, industrial simulations, intelligent R&D, and data analysis—these intelligent scenarios, once exclusive to large enterprises, have become key methods for SMEs to reduce costs and enhance efficiency while seizing niche markets.

However, the reality is not optimistic. Currently, the issue of mismatched supply and demand for computing power still exists. Although the total computing power in our country is large, there is a dislocation between the supply and demand in the eastern and western regions, and the overall utilization rate is low. More critically, the computing power demands of SMEs exhibit characteristics of "small batches, fragmentation, and temporariness", while traditional computing power services operate on a "long-term binding, large upfront payment" asset-heavy model, creating a complete misalignment between the two logics.

How serious is this contradiction? A set of data is sufficient to illustrate the problem. According to statistics from the National Bureau of Statistics, as of March this year, the average daily Token call volume in our country has exceeded 140 trillion, which is an increase of over 1,000 times compared to 100 billion at the beginning of 2024 and an increase of over 40% in the three months compared to 100 trillion at the end of 2025. The explosive growth in Token call volume indicates that AI applications are penetrating various industries at an unprecedented speed. However, during the same period, the annual growth rate of the total computing power in our country is about 30%, and the growth rate of supply is far behind the explosive demand.

This is not merely a problem of "not enough computing power". The deeper issue lies in the fact that a large amount of computing resources remain idle or are utilized inefficiently, while SMEs that urgently need computing power are deterred due to high thresholds. The computing power gap is becoming a crucial bottleneck restricting the innovative vitality of SMEs.

2. "Computing Power Bank": Let Computing Power Flow Like Money

On April 2, 2026, the Ministry of Industry and Information Technology issued a notice to organize a special action aimed at empowering the development of SMEs through inclusive computing power, explicitly proposing the exploration of two innovative services: "Computing Power Banks" and "Computing Power Supermarkets". The policy immediately garnered significant attention in the industry.

Why are these two concepts so captivating? Because they fundamentally reconstruct the logic of computing power supply.

The core logic of the "Computing Power Bank" is not complicated. It draws on the traditional banking "deposit-loan" model: SMEs can "deposit" idle computing resources into a unified resource pool, and the platform enables flexible storage, demand-scheduling, and value realization through intelligent scheduling across regions and periods. In other words, where computing power used to be a bunch of silent servers that could only be used for self-use—if you bought them and didn't use them all, they would just sit idle—now it transforms into an asset that can be "deposited" or "withdrawn"; during idle times, resources can be deposited in the pool to earn returns, and during busy times, they can be withdrawn to meet needs.

The essence of the Computing Power Bank is to transform fragmented idle computing power into a circulating inclusive resource.

Some may ask: How can SMEs, which already have insufficient computing power, have idle computing resources to "deposit"? This is precisely a cognitive misconception. An industrial design SME may require a lot of computing power for rendering and simulations during peak project seasons, but during off-peak seasons, these servers are almost in a dormant state. An AI startup may face computing shortages during the day but have large amounts of idle power at night. Under the traditional model of building and using resources, such periodic idleness cannot be realized and is merely wasted. The Computing Power Bank, however, provides a pathway to activate this "dormant computing power".

The policy design of the Ministry of Industry and Information Technology goes further. The notice clearly requires the improvement of special zones for SMEs on national computing power platforms, the connection of computing resources on the China SME Service Network, and the national computing power internet service node zones, to facilitate precise matching of SME demands with computing resource supply. It also promotes flexible payment models based on "usage time", "calculated time", and Tokens, encouraging local relevant departments to subsidize "computing power coupons", "storage coupons", and "operational coupons" to SMEs that meet industrial orientations through SME platform special zones.

The intent of this set of combined measures is very clear: the Computing Power Bank addresses the problems of resource "storage" and "scheduling", while computing power coupons resolve the issue of "affordability", and platform special zones tackle the question of "discoverability". Together, they form a complete closed loop for the supply of inclusive computing power.

3. "Computing Power Supermarket": Buying Computing Power Like Online Shopping

If the Computing Power Bank addresses the issue of resource circulation, then the Computing Power Supermarket focuses on the consumer experience.

The "Computing Power Supermarket" is positioned as a public computing service portal that gathers various computing services and supports online consultation and transactions of computing power. In simpler terms, it is akin to an e-commerce platform in the computing power field—intelligent computing cards, multi-card clusters, bare-metal servers, and GPU cloud hosts are all clearly priced, allowing enterprises to log in to the platform and order on demand just like purchasing daily necessities, paying immediately for use.

This logic sounds simple, but it is revolutionary in the field of computing power. In the traditional model, for SMEs to access computing power, they had to either build their own data centers (with initial investments starting in the millions), or sign long-term leases (contract binding, large upfront payments), or simply give up using it altogether. The Computing Power Supermarket, by productizing, standardizing, and retailing computing power, allows businesses to only pay for what they actually use, with billing based on usage time, calculated time, or even Tokens, truly achieving "pay for what you use".

The practice of Shanghai Telecom's Computing Power Supermarket has already demonstrated the viability of this model. The platform connects the powerful computing resources of the Qingpu and Lingang "East-West Wings" intelligent computing centers, and is open to computing power suppliers, SMEs, and the public, supporting various computing service providers to operate on the platform. Currently, it has established a district-level computing operation zone for related areas, providing online ordering services for intelligent computing single cards/multi-card, bare-metal, and GPU cloud hosts, and possesses core functionalities such as multi-level account management and precise measurement and billing, providing "computing power + model + application" one-stop service support for clients in finance, government, education, and other industries.

The Henan Airport Intelligent Computing Center is also validating the feasibility of inclusive computing power using a different approach. As the first intelligent computing center in central China to fully deploy the DeepSeek model, this center allows SMEs to conveniently utilize leading domestic AI large models through an "instant access, immediate use" approach. The domestic AI chip enterprise Taichu Yuqi has built the intelligent computing foundation for it and completed the Token API interface deployment, meaning SMEs can call domestic intelligent computing power just like invoking APIs without making a large upfront investment. To reduce trial-and-error costs, this intelligent computing center also offers Token trial services to SMEs and research institutions from universities.

From Shanghai to Henan, both paths point in the same direction: making computing power transition from "a luxury for the few" to "a daily necessity for the many".

4. Target 2028: The Roadmap for Inclusive Computing Power

If the practices in Shanghai and Henan are the "present continuous tense", then the timetable set by the Ministry of Industry and Information Technology for inclusive computing power is the "future roadmap" for the entire industry.

The notice clearly states that by the end of 2028, a comprehensive inclusive computing service system will be basically established, covering a wide range with low costs, excellent services, vibrant ecology, and strong talent, covering no less than 10 categories of the 15 industry categories applicable to SMEs, further strengthening public service efforts for computing power applications for SMEs, and significantly reducing the barriers for SMEs to acquire and use computing power.

This goal can be interpreted from three dimensions.

First, breadth of coverage. Covering no less than 10 categories among the 15 industry categories means that inclusive computing power will penetrate the key sectors of the national economy, including manufacturing, education, agriculture, and finance. This is not an "optional item" but a rigid requirement at the policy level. The notice also specifies that it will focus on digital transformation pilot cities for SMEs, characteristic industrial clusters, and innovative industrial clusters, centering on key processes in manufacturing SMEs such as R&D design, production, equipment operation, and supply chain management, and expedite the promotion of low-cost, lightweight, and easily deployable inclusive computing power solutions.

Second, depth of service. Broad coverage is just the first step; what is truly critical is providing quality services, vibrant ecology, and strong talent. To this end, the notice has deployed five key tasks—computing power resource allocation enhancement action, inclusive computing power service supply action, key industry enterprise empowerment action, industrial ecology collaborative construction action, and talent cultivation capacity building action. From resource deployment to talent cultivation, this is a system engineering that covers the entire chain of computing power services.

Third, mechanism innovation. Promoting a "1+N" pairing assistance mechanism between large, medium, and small enterprises is a significant highlight of this special action. The notice encourages large enterprises to establish normalized and institutionalized pairing assistance relationships with SMEs and to share their experiences in computing power technology, operation and maintenance management, cost control, and scenario implementation. It also organizes activities to facilitate the communication and connection of large, medium, and small enterprises, encouraging leading enterprises to open up their internally idle computing resources when releasing demand scenarios to SMEs. This "big-to-small" model provides institutional guarantees for the virtuous cycle of the computing power ecosystem.

5. From "Building" to "Using":

A Profound Shift in Industry Logic

Looking across the design of the entire special action, one core signal is clear: the focus of China's computing power industry is shifting from "building" to "using".

In the past few years, our country has invested heavily in the construction of computing power infrastructure. Intelligent computing centers have blossomed everywhere, and the total computing power scale has continued to rise. But with the infrastructure built, who will use it, how to use it, and can it be afforded? These questions have not received equal attention. As a result, a large amount of computing resources remain idle, with low utilization rates, starkly contrasting with the strong computing demand from SMEs.

The significance of this special action lies in the fact that it no longer pursues "larger scale", but aims for "higher efficiency"; it no longer focuses on "who builds it", but rather on "who uses it"; it no longer emphasizes "technological advancement", but stresses "service inclusiveness". This represents a shift in industry logic from supply-driven to demand-driven.

Of course, the development of any new model is not accomplished overnight. For the Computing Power Bank and Computing Power Supermarket to truly take off nationwide, they still face a series of practical challenges. Issues such as the latency of cross-regional scheduling, interface standardization of computing power products from different vendors, data security and privacy protection in computing transactions, and market mechanism design for pricing idle computing power—all need to be gradually explored and resolved in practice. Just as Heidegger revealed when questioning the essence of technology, the true value of technology lies not in its tool-like nature, but in how it changes the relationship between man and the world. The ultimate significance of inclusive computing power may not be about enabling more enterprises to use AI, but rather making AI truly a tool in everyone's hands, rather than a privilege for a few.

When computing power flows like electricity and water.

Innovation will no longer have barriers.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。