China accelerates the formulation of national integrated computing power network technical standards, while the largest communication ETF by index scale, ChinaAMC (515050), and the lowest fee ChiNext Artificial Intelligence ETF, ChinaAMC (159381), undergo fluctuations and adjustments
I'm LongbridgeAI, I can summarize articles.On May 26th, the Communication ETF ChinaAMC (515050) fell by 1.33%, and the ChiNext Artificial Intelligence ETF ChinaAMC (159381) fell by 1.06%. The National Bureau of Statistics is accelerating the formulation of technical standards for the national integrated computing power network to promote the optimized allocation of computing power resources. CITIC Securities expects that the demand for AI computing power will maintain rapid growth and suggests paying attention to the opportunities for value reassessment on the IDC and network sides brought about by domestic computing power investments
On May 26th, during the midday session, the three major indices fell again, with the communication equipment sector leading the decline. As of 14:00, the largest communication ETF by index size, ChinaAMC (515050), dropped by 1.33%. Among its holdings, Shengyi Technology hit the daily limit up, Huadian Technology rose over 8%, while Xinyi Technology, Shenzhen South Circuit, and Dalian Technology fluctuated upward. Unisplendour and Guangku Technology led the decline; the lowest fee rate ChiNext Artificial Intelligence ETF, ChinaAMC (159381), fell by 1.06%. As of the publication, the trading volume of the communication ETF, ChinaAMC (515050), exceeded 1 billion yuan.
In terms of news, according to CCTV News, the National Bureau of Statistics stated that China is accelerating the research and formulation of national integrated computing power network technical standards. A total of 12 relevant guiding technical documents have been issued, covering various aspects such as computing power monitoring and scheduling, computing power and electricity collaboration, and security protection, to promote the optimized allocation of national computing power resources. It is reported that the computing power network under construction is akin to the current power grid. Users do not need to buy generators because the power grid has achieved unified scheduling, allowing for on-demand use.
CITIC Securities stated that as AI commercialization accelerates by 2026, the demand for computing power has shifted from a single "training-driven" model to a dual "training + inference" drive, and it is expected to maintain rapid growth in the future. With continuous breakthroughs in domestic computing power chips in terms of process and architecture, increasing capacity and yield of advanced domestic processes, and the introduction of super nodes, domestic cloud vendors are expected to increase their computing power investments, leading to continuous expansion of domestic intelligent computing centers and promoting the formation of a closed loop of "chips - models - applications." It is recommended to focus on the IDC and network-side value reassessment opportunities brought about by the acceleration of domestic computing power investments.
The communication ETF, ChinaAMC (515050), tracks an index where the combined weight of CPO + PCB concept stocks accounts for nearly 80%, ranking first in the entire market (excluding overlapping stock weights), with the PCB concept stock weight exceeding 20%. This layout fully covers the entire chain from the scarce upstream optical chips to the fiber optic cables, AI servers, optical interconnects, and PCBs that benefit from new CPO/OCS technologies. It directly undertakes the dual drive of accelerated AI computing capital expenditure and technological iteration, serving as a core tool to capture the performance release of the communication and electronic computing power sectors. As of April 30, 2026, the top ten weighted stocks are Zhongji Xuchuang, Xinyi Technology, Luxshare Precision, Industrial Fulian, Dongshan Precision, Zhaoyi Innovation, Tianfu Communication, Huadian Technology, Hengtong Optic-Electric, and Huagong Technology, with the top ten weighted stocks accounting for 60.5% in total. Off-market connection (Class A: 008086; Class C: 008087)
The high-content AI theme ETF for optical modules: ChiNext Artificial Intelligence ETF, ChinaAMC (159381), has half of its index weight concentrated in the optical module CPO sector, while the other half covers domestic computing power and AI software application fields, forming a balanced layout of "hardware + applications." The weights of the three leading optical module companies, Xinyi Technology, Zhongji Xuchuang, and Tianfu Communication, account for over 41%. The top 10 weighted stocks are Zhongji Xuchuang (17.62%), Xinyi Technology (15.18%), Tianfu Communication (8.33%), Xiechuang Data, BlueFocus Communication Group, Runze Technology, Changxin Bochuang, Beijing Junzheng, Kunlun Wanwei, and Tonghuashun Currently, the fund size exceeds 2 billion yuan, with an on-market comprehensive fee rate of only 0.20%, which is the lowest for ETFs. Off-market connection (Class A: 025505; Class C: 025506)
