Summary of Key Points
In 2026, industries representing new forms of productive forces, such as humanoid robots, semiconductors, and computing power, are entering a period of accelerated development: robots are moving from laboratories to commercial use, with explosive growth in production and financing; mergers and acquisitions in the semiconductor and digital economy sectors are intensifying, showing clear trends towards strengthening and expanding industry chains; capital is heavily invested in the primary market, while there are disagreements in the secondary market, but long-term opportunities are emerging. Technology experts predict that physical AI (embodied intelligence) will still take 2-5 years to reach a level comparable to ChatGPT, and the "necessary bubbles" in cutting-edge fields will drive long-term progress.
I. Humanoid Robots: From "Concept Shows" to Profit-Making Machines
2026 is a critical turning point for the robotics industry—robots are finally evolving from laboratory prototypes into profitable products. Data from the Ministry of Industry and Information Technology shows that the national production of humanoid robots was only 20,000 units in 2025, but it exceeded 40,000 units in the first half of 2026, with an expected annual total of over 100,000 units. Leading companies have already started to generate substantial profits: for example, Yushu Technology sold 5,500 pure humanoid robots in 2025, generating revenue of 1.7 billion yuan and a net profit of nearly 600 million yuan, with a gross margin exceeding 60% (which is much higher than many other industries).
Investors are also betting heavily on this trend: global funding for embodied intelligence was only 2.7 billion dollars in 2023, but it is expected to exceed 15 billion dollars in 2026. Companies are no longer just creating demonstration videos; they are bringing real-world applications to exhibitions—such as Leju Robotics, which replicated its production line at the WAIC exhibition, where robots were used to open boxes and load materials, immediately attracting orders from industry leaders like FAW and ZTE. Public expectations for robots have shifted from "science fiction toys" to "tools that can help with work."
II. Robot Investments: Some Investors Hit the Right Targets, but Institutions Say the Market Hasn't Yet Reached a Full Upsurge
Primary market investors have already recognized the potential of this sector. Zhang Lei, an early investor in Yushu Technology, invested 2.5 billion yuan in 2023, and the company's valuation has likely increased significantly since then. His logic is simple: ChatGPT acts as the "brain," and robots provide the "body"; combining the two will liberate human labor, leading to the emergence of leading companies.
However, institutions in the secondary market are more cautious. Huatai Securities believes that the current market is still in its early stages (not yet at a point of full growth) due to disagreements regarding initial production volumes and quality of demand. Jiaoyin International is more practical, suggesting that wheeled dual-arm robots (which are simpler than full-size humanoid robots) may be the first to generate profits due to their stability in industrial applications such as logistics and quality inspection.
III. Semiconductor Mergers and Acquisitions: Leaders Focus on Strengthening Their Supply Chains
The semiconductor industry saw a surge in mergers and acquisitions in 2026, with total amounts approaching 80 billion yuan, all aimed at strengthening supply chains and overcoming domestic bottlenecks. For example, SMIC spent 40.6 billion yuan to acquire the remaining shares of SMIC North, gaining full control over its Beijing 12-inch wafer production facility, which allows for more efficient capacity management and improved procurement of mature manufacturing processes. Other companies, such as Hua Hong and Jinghe Integration, are also acquiring assets to expand their specialized manufacturing capabilities and reduce dependence on foreign suppliers.
Design and testing sectors are also undergoing consolidation: Huatian Technology acquired Hua Yi Microelectronics to enter the power device market, while Gailun Electronics bought Rui Cheng Xin Wei to enhance its IP portfolio, both moves aimed at reducing reliance on imports.
IV. Computing Power & Digital Economy: Traditional Enterprises Cross Industries to Seize AI Infrastructure
AI models require massive computing power, making infrastructure a target for mergers and acquisitions. Traditional companies are entering this field: Dongguangyang, which operates in aluminum foil and pharmaceuticals, spent 8 billion yuan to acquire Dongshu No.1, thereby gaining access to IDC leader Qinhuai Data and transitioning to a dual-business model of "materials + intelligent computing"; Shensanda A invested 6.4 billion yuan in China Electronics Cloud to build a nationally controlled cloud platform.
The digital economy sector is even more active: Xiangcai Securities plans to merge with Dazhihui to create a rare combination of "broking + internet traffic"; traditional company Wanlong Optoelectronics acquired Zhongkong Information to enter the smart city and industrial software markets; even ST Yushun is transforming through acquisitions in data technology. Mergers are seen as the fastest way for these companies to capitalize on the benefits of the digital economy.
V. What Do Technology Experts Think?
There is both consensus and disagreement within the tech community about the future of AI:
- Challenges for Physical Intelligence: Yao Maoqing, CEO of Mi Feng Technology, points out that robots need to overcome three major hurdles to move from demonstration to practical use: limited and expensive data (difficult to obtain for real-world interactions), a lack of a unified "physical language" (rules vary across different robots and applications), and high trial-and-error costs (serious damage can result if robots malfunction).
- The Necessity of Bubbles: Kevin Kelly, founder of Wired magazine, believes that there may be bubbles in cutting-edge AI models (such as those that consume resources without generating immediate profits), but investments in infrastructure (like computing power and chips) are necessary for long-term progress. He also notes that AI will gradually transform society, with lower-cost, localized models becoming more competitive.
In summary, new forms of productive forces are moving from speculative concepts to practical applications. Capital and companies are actively making moves, but true maturity will take time. Investors need to focus on tangible indicators such as mass production, profitability, and repeat purchases to assess the potential of these developments.
(Note: The companies mentioned in this article are used for illustrative purposes only and do not constitute investment advice.)