Funds take profits in tech stocks after the rebound: What will keep physical AI active?
Cailianshe’s midday market review shows that the market diverged after last Friday’s strong rebound, with pronounced selling pressure on tech stocks. Physical AI remains active, but the sector’s upside potential and the staying power of autonomous driving are key to judging whether the rally can continue.
Original source：财联社First published on Unhidea：Original source date：
Key takeaway
Cailianshe considers the divergence after the rebound still acceptable, but profit-taking in tech stocks is pronounced. Whether physical AI can sustain its momentum depends on the sector’s upside potential and the staying power of autonomous driving.
Where is the main pressure after the rebound?
In morning trading on 26 October, the Shanghai Composite was relatively strong, while the ChiNext and STAR Market fluctuated weakly. Following last Friday’s strong rebound, the market diverged. Cailianshe judged the degree of divergence still acceptable, although the indices performed slightly worse than it expected.
Trading volume fell markedly from the previous session, though it was better than the author’s initial scenario of a contraction of around 20%. The slightly better volume did not eliminate profit-taking pressure, and ChiNext and STAR Market stocks and high-flying shares remained relatively weak.
Why are tech stocks a focus when assessing selling pressure?
Chips and semiconductors were among the more active areas in last Friday’s rebound, but Yuexin Semiconductor, cited in the article, did not maintain its strength that morning, with its share price below the previous close. The author therefore judged that funds were taking profits in tech stocks, while computing-power hardware performed even more weakly.
Outside technology, oil and gas, shipping, property, consumer stocks and media all showed strength, with funds flowing into a range of areas. The author paid closer attention to retail and department stores, as well as short dramas and games, believing these segments had relatively greater staying power.
How can physical AI move from an active theme to a trend?
Robotics and physical AI remained active amid the divergence, but the sector’s upside was constrained. The author identified whether Daye Co., Ltd. could lock in its daily price limit again in the afternoon as a point to watch: if it failed to do so, the sector’s room to advance could be affected.
Market trading showed signs that physical AI was tilting toward autonomous driving. The author believes autonomous driving could be the first physical AI application to be deployed on a large scale; only if the rally sustains itself could it signal a shift from thematic speculation to a trend-driven market.
What connects ThunderSoft to the real-world deployment of physical AI?
The article focuses on ThunderSoft, with the investment thesis centered on the operating system as a system-level interface: vehicles and robots need unified coordination of different types of computing power, real-time control and device-cloud collaboration, and the company uses AI-native operating systems to connect these elements.
On the automotive side, domain controllers from its subsidiary Changxing Intelligent Driving already support L2.9 to L3 urban autonomous driving and can be extended to L4 unmanned delivery and cleaning scenarios; Dishui AIOS and AIBOX work together to bring large models into vehicles. The business link here is the systems integration of autonomous driving and in-vehicle AI.
On the robotics side, the company is advancing an engineering path from proof of concept to mass production using “AIOS + on-device AI + chip platforms.” The article reports that it has served more than 100 major clients, achieved a repeat purchase rate of 60%, and partnered with the Beijing Humanoid Robot Innovation Center to promote large-scale applications across production-line clusters.
