Driven by NVIDIA's latest earnings report significantly exceeding market expectations, the ChiNext Artificial Intelligence ETF (159363), which holds a significant position in leading optical module companies, saw its on-market price surge by 2.39% against the broader market trend today (February 26). Data indicates that funds had already begun accumulating positions in advance, with the ETF attracting a net inflow of 532 million yuan over the preceding four consecutive days.
Among its component stocks, leading optical module manufacturer Suzhou Tfc Optical Communication Co.,Ltd. rose over 9%, reaching a new all-time high. Furthermore, Xingchen Technology led gains by over 18%, followed by Runze Technology which advanced more than 11%, driving the index's strong performance against the market trend.
The catalyst was the latest financial results released early this morning (Beijing Time) by NVIDIA, the global leader in AI. The company reported fourth-quarter revenue for fiscal 2026 of $68.1 billion, a substantial 73% year-over-year increase, surpassing analyst expectations of $65.684 billion. Additionally, NVIDIA's provided guidance also exceeded market forecasts, significantly alleviating concerns about an "AI bubble" and demonstrating persistently strong demand generated by artificial intelligence.
Consequently, NVIDIA's stock price surged in after-hours trading, at one point climbing over 4%. NVIDIA CEO Jensen Huang stated in the earnings report that computing demand is growing exponentially, marking a tipping point for AI agents. Corporate adoption of AI agents is accelerating rapidly.
Industry insiders point out that the demand for high-speed, low-latency internal communication within AI computing clusters is exploding. This directly benefits leading domestic optical module manufacturers, such as Zhongji Innolight, Suzhou Tfc Optical Communication Co.,Ltd., and Eoptolink, whose demand for high-speed products like 800G optical modules is expected to remain robust.
As AI development shifts from infrastructure building to application deployment, the ChiNext Artificial Intelligence ETF (159363) and its off-exchange counterparts (Class A: 023407, Class C: 023408), which provide one-click exposure to both "AI computing power and applications," stand to benefit more directly from the growth dividends of AI technology commercialization. In terms of sector allocation, the ChiNext AI index allocates approximately 60% to computing power (leading optical module and IDC companies) and about 40% to AI applications, positioning it not only as a core "computing power" play but also as a genuine representative of "AI application" opportunities.
Comments