Group 1 - The core viewpoint of the article is that the recent decline in the A-share AI sector is primarily driven by lower-than-expected capital expenditures reported by major cloud companies like Tencent and Alibaba, raising concerns about the sustainability of the industry's growth [1][2] - The market's shift in focus from merely following the "NVIDIA chain" to valuing "domestic AI innovation" indicates a new pricing anchor for the domestic AI industry [1][2] - The decline in capital expenditures has led to heightened investor anxiety regarding the pace of AI application deployment and the actual investment in computing infrastructure [2][3] Group 2 - The slowdown in AI application deployment is attributed to the fact that while leading models can handle certain tasks, they often do not address fundamental human problems effectively, leading to a lack of necessary demand for AI applications [3][4] - The emergence of DeepSeek did not create a supply-demand imbalance in computing power; instead, there are structural excesses in the domestic computing power market, with a significant gap in advanced computing capabilities [4][5] - The capital expenditures of major internet companies are viewed as a key forward-looking indicator of AI progress, despite the potential for short-term fluctuations due to external factors [5][6] Group 3 - The anticipated launch of NVIDIA's new product, B40, is expected to drive a new wave of capital expenditure growth among major companies, potentially revitalizing market sentiment towards the domestic computing power sector [9][10] - The ongoing trade tensions between China and the U.S. have led to a shift in AI investment strategies, with companies diversifying their supply chains and focusing on domestic alternatives [10] - Predictions indicate that the proportion of AI servers in China sourced from domestic suppliers is expected to rise significantly, suggesting a long-term growth potential for domestic computing power [10]