"Morpheme Economy" rises: listed companies ramp up full-chain layout

Securities Daily reporter Liu Xiao Yi

The concept of “Token” continues to gain traction, and the A-share computing power sector is strengthening. On March 25, stocks such as Shenzhen TeFa Information Co., Ltd., 263 Network Communication Co., Ltd., and Shanghai Tianji Technology Co., Ltd. surged to their daily limit, while several other stocks, including Beijing Guanghuan Xinwang Technology Co., Ltd. and Guangdong AoFei Data Technology Co., Ltd., saw significant gains.

A Token is the smallest unit of information processed by large models, possessing three core attributes: it is measurable, can be priced, and is tradable. It serves as the basic unit for charging and settling AI services. If a large model is an intelligent machine, a Token is like “every unit of electricity” consumed while this machine operates, and it is also the fundamental measurement unit for charging services provided by this machine.

Data from the National Data Bureau shows that the daily average Token usage in China is experiencing explosive growth, reaching 100 billion in early 2024 and soaring to 100 trillion by the end of 2025. As of March 2026, the daily average has surpassed 140 trillion, with a growth of over a thousand times in two years.

“The ‘Token economy’ is expected to reshape the business logic of AI, shifting large model companies from ‘selling capabilities’ to ‘selling usage’.” Yuan Shuai, deputy secretary-general of the Zhongguancun Internet of Things Industry Alliance, told Securities Daily reporters. Under this model, corporate revenue will no longer rely on long-term project delivery but instead depend on the scale of real-time concurrent interactions. This change can drive the AI industry from being purely technology-driven to being driven by refined operations, enabling high-frequency and high-stickiness application scenarios to swiftly translate into explosive performance growth.

In light of the potential commercial transformation brought by the “Token economy,” global tech giants are accelerating their strategic layouts. Jensen Huang, founder and CEO of Nvidia, introduced “Token Economics” at the 2026 GTC conference, arguing that future data centers will no longer be warehouses for storing files, but “factories” for producing Tokens. At the same time, Nvidia also launched a full-stack hardware system to reduce Token production costs.

On March 16, Alibaba Group Holding Limited officially established the Alibaba Token Hub (ATH) business group, which focuses on “creating Tokens, delivering Tokens, and applying Tokens,” directly overseen by Group CEO Wu Yongming.

“The Token economy will bring tremendous growth opportunities.” Li Peng, senior vice president of Huawei Technologies Co., Ltd., predicts that by 2030, the scale of China’s AI-related industries will exceed 10 trillion yuan.

Surrounding the production, usage, distribution, and settlement of Tokens, domestic listed companies are accelerating the entire industry chain layout with upstream computing power infrastructure as the core, midstream platform services as the link, and downstream applications as the focus, leading to a continuous increase in industry concentration.

Upstream computing power infrastructure is the core foundation for Token production, becoming the most direct beneficiary. For example, XieChuang Data Technology Co., Ltd. (hereinafter referred to as “XieChuang Data”) achieved significant year-on-year growth in both revenue and net profit in 2025, with plans to invest 21.2 billion yuan in purchasing high-performance servers for large-scale computing cluster construction.

Geng Kangming, chairman of XieChuang Data, told Securities Daily reporters that current AI industry giants are generally optimistic about the continuous growth of computing power. OpenAI, Anthropic, Microsoft, Google, and others are all laying out intelligent agents, with inference models like DeepSeek pushing agents into the implementation phase.

Midstream platforms serve as the scheduling and settlement hub, undertaking key functions such as Token distribution, billing, and computing power scheduling. For instance, Zhongbei Communication Group Co., Ltd. has already established operational advantages at scale, currently operating computing power exceeding 17,000P and providing intelligent computing services to clients such as Kingsoft Cloud Holdings Limited.

In the downstream application field, listed companies leverage industry advantages to promote the integration of Token technology with vertical scenarios, accelerating commercial monetization. Major model and robotics AI application manufacturers all belong to this field. For example, iFlytek Co., Ltd. achieves large-scale monetization in education, healthcare, and office scenarios relying on its Spark large model; Shenzhen UBTECH Robotics Corp. empowers its robots with large models to support intelligent interaction and embodied intelligence implementation.