On Wednesday, April 15, the market experienced consolidation. The Huabao Shanghai Science and Technology Innovation Board Artificial Intelligence Trading Open Ended Index Sec (589520), which focuses on the domestic AI industry chain, was active in the morning session, with its intraday gain reaching a high of 1.46%. It subsequently pulled back alongside the broader market in the afternoon, with its gains narrowing gradually, ultimately closing down 0.49%. However, looking at the daily K-chart, since the recent low on March 24, the ETF appears to have formed a pattern of fluctuating upward steps, potentially indicating a right-side signal of climbing from lower levels.
Among its constituent stocks, domestic GPU leader Cambricon Technologies Corporation Limited (688256) continued its strong rally for an eighth consecutive session, with a cumulative gain of 27.53%, and saw an intraday increase exceeding 6% today. Domestic ASIC leader VeriSilicon Microelectronics led gains by over 3%, followed by Winsense Technology, Anlogic Technology, Ezviz Network, ClouDr Group, and Roborock. The remaining 23 stocks declined, with Montage Technology falling over 4% and GeoStar dropping more than 3%, ranking among the top decliners and weighing on the index performance.
On the news front, on April 14, Sugon released the largest AI4S computing cluster in China to date, comprised of 60,000 domestically produced accelerator cards. This marks a critical step forward for China in the field of AI computing infrastructure. Sugon's full-stack self-development breaks overseas technological monopolies; the cluster includes six types of self-developed core chips, achieving overall international advanced levels and significantly boosting confidence in the domestic AI chip supply chain.
Industry insiders pointed out that from the launch of the 30,000-card core node in Zhengzhou on February 5 to the deployment of the 60,000-card computing cluster on April 14, Sugon completed the leap from 10,000-card level engineering breakthroughs to ultra-large-scale deployment in less than three months. Behind this milestone lies not only a technological breakthrough for domestic computing power but also signifies Sugon's ongoing business model transformation from selling hardware to offering computing power as a service, prompting a systemic reassessment across performance elasticity, competitive barriers, and valuation potential.
Amid changing external conditions, the winning bid share for domestic GPUs in intelligent computing center tenders has increased to over 35%, indicating market demand shifting from 'usable' to 'essential'. Sinolink Securities believes that 2026 will be a pivotal year for China's computing power demand transition from cloud training to dual-driven by training and inference, potentially accelerating the development of domestic AI chip manufacturers. Morgan Stanley noted that generative AI is driving sustained high growth in cloud provider capital expenditures, with explosive demand for AI GPUs, custom ASICs, and high-end memory, coupled with optimized supply-demand dynamics in the industry chain, suggesting the semiconductor industry is entering an AI-driven structural growth cycle.
Notably, domestic AI chip benchmark Cambricon Technologies Corporation Limited reported its first annual profit since listing on March 16, with revenue more than quadrupling year-over-year, signaling a transition for AI companies from the technology investment phase to the commercial return phase. Benefitting from AI computing power drivers, the global semiconductor industry's target of reaching a trillion-dollar output value is expected to be achieved by the end of 2026, ahead of the previous 2030 forecast, potentially heralding an inflection point for companies in the domestic AI industry chain.
The Huabao Shanghai Science and Technology Innovation Board Artificial Intelligence Trading Open Ended Index Sec (589520) and its feeder funds focus on the domestic AI industry chain. Its constituents include domestic GPU leaders, domestic ASIC leaders, and AI application leaders, with the semiconductor sector accounting for nearly half of the weighting, giving it strong offensive characteristics. The software sector weighting exceeds 30%, positioning it to potentially benefit from catch-up rallies in AI applications. Furthermore, the ETF is a margin trading security, making it an efficient tool for gaining exposure to domestic computing power.
Comments