At the close, the Shanghai Composite Index rose 0.11%, the Shenzhen Component Index rose 0.52%, and the ChiNext Index rose 0.61%. In terms of ETFs, Bosera Science and Technology Innovation Board Artificial Intelligence ETF (code: 588790) closed up 1.01%.

On March 18, the market rose and fell throughout the day, and the three major indexes rose slightly. In terms of sectors, ports, robots, high-voltage fast charging, gold concepts and other sectors were among the top gainers, while liquor, film and television, retail, games and other sectors were among the top decliners. At the close, the Shanghai Composite Index rose 0.11%, the Shenzhen Component Index rose 0.52%, and the ChiNext Index rose 0.61%. In terms of ETFs, Bosera Science and Technology Innovation Board Artificial Intelligence ETF (code: 588790) closed up 1.01%, and among the constituent stocks, Hengxuan Technology (688608. SH) rose 5.74%, and Neoway Technology (688159. SH) rose 4.63%, VeriSilicon (688521. SH) rose 3.53%.

On the news side, on March 17, local time, the NVIDIA AI event GTC 2025 officially kicked off in San Jose, USA, focusing on AI agents, robotics and the future of accelerated computing, and will launch new CPO switches, new computing power cards and a new generation of AI chips. It is worth mentioning that this NVIDIA GTC conference will set up a "Quantum Day" for the first time, and executives from more than 10 leading companies in the quantum computing industry will share the stage to discuss the technical status, potential and future development direction of quantum computing.

Tianfeng Securities believes that the breakthrough in intelligent and graphic perception technology displayed at this conference may indicate a larger-scale popularization and wide application of robots in the future, which is expected to promote major manufacturers to enter the humanoid robot industry; The performance requirements brought about by the progress of AI intelligence are expected to promote the development of the key hardware industry of robots in the direction of high precision and high reliability.

In addition, the Suzhou Municipal Bureau of Industry and Information Technology recently solicited public opinions on the "Several Measures to Accelerate the Development of the AI Chip Industry in Suzhou (Draft for Comments)". It is mentioned that the backbone core enterprises should be strengthened. Focus on key directions such as GPU general-purpose chips, ASIC special chips, FPGA semi-customized chips, integrated storage and computing chips, silicon photonic chips, etc., increase investment attraction, accelerate the introduction of a number of high-quality projects and head enterprises with strong driving force, and provide comprehensive support for key projects in terms of space security, site construction, and talent introduction. Promote AI chip enterprises to improve their ability to integrate resources through mergers and acquisitions, expand their business to the upstream and downstream of the industrial chain, carry out mergers and acquisitions of high-quality AI chip enterprises, and encourage county-level cities (districts) to reward enterprises for mergers and acquisitions. AI chip companies will be awarded 1 million yuan and 400,000 yuan respectively for being awarded the individual champion of the manufacturing industry and the national specialized and special new "little giant". A maximum of 1 million yuan will be given to the recognized AI chip high-tech enterprises.

With the rapid growth of global AIDC power consumption and the significant increase in the total scale of computing power in China, the development and application of AI technology is bringing unprecedented opportunities for the construction of computing infrastructure. According to Semianalysis, global AIDC power consumption will grow rapidly from 49GW in 2023 to 96GW in 2026, with 90% of the growth driven by AI. At the same time, data from China's Ministry of Industry and Information Technology also shows that by the end of the third quarter of 2024, the total scale of China's computing power has reached 268E, and it is expected to exceed 300E in 2025. This not only reflects the strong driving effect of AI technology on the demand for computing power, but also indicates that in the next few years, with the continuous progress of AI technology and the expansion of application scenarios, the scale of investment in computing infrastructure will continue to expand, and related industry chain enterprises are expected to benefit from it.