Hot Spot Interpretation丨The demand for AI computing power is high, and central enterprises are making efforts to build computing power

Edited by: Xiao Ruidong

1. The demand for AI computing power is booming, and central enterprises are making efforts to build computing power

Event: Recently, the State-owned Assets Supervision and Administration Commission of the State Council held a special promotion meeting on artificial intelligence of central enterprises on "AI empowerment and industrial renewal", emphasizing that central enterprises should put the development of artificial intelligence in the overall work and make overall plans, further promote industrial renewal, and accelerate the layout and development of intelligent industries. Accelerate the construction of a number of intelligent computing power centers and carry out AI+ special actions. At the meeting, 10 central enterprises signed a proposal, stating that they would take the initiative to open up artificial intelligence application scenarios to the society.

2. Interpretation of hot spots

1. The core drive of artificial intelligence involves three aspects: data, algorithms and computing power, and the rapid development of AI applications brings long-term and massive computing needs, among which the technical infrastructure with high computing power can accelerate the speed of data processing and analysis, promote the application and optimization of complex algorithm models, and provide a broader space for artificial intelligence innovation. As the vane of the global artificial intelligence industry, Nvidia recently announced that the company's revenue in the fourth quarter of 2024 was US$22.1 billion, an increase of 265% year-on-year and 22% quarter-on-quarter, exceeding market expectations of US$20.41 billion. As the underlying cornerstone of the AI era, computing power is expected to attract market attention again, and relevant suppliers are expected to usher in the release of performance, focusing on investment opportunities related to the AI computing power industry chain, including AI server parts, server machines, computing power leasing, data centers and other links.

At the national level, computing power has become an important measure of national strength, and countries around the world are formulating their own AI strategies and policies to promote the development of the AI industry. Among these strategies, strengthening the construction of AI infrastructure is generally regarded as a key component, including computing resources, data resources, and human resources. In this context, the development of the artificial intelligence industry has received attention at the national level, and relevant policies have been favorable. From "Eastern Data and Western Computing" to "National Integrated Computing Network" to "Central Enterprise Artificial Intelligence Special Promotion Conference", it demonstrates the determination of policy support. This time, the State-owned Assets Supervision and Administration Commission (SASAC) has promoted the development of artificial intelligence in central enterprises, once again promoting the development of the industry, and is optimistic about the artificial intelligence sector that has successively issued standards and support policies by beneficiary countries and local governments.

From 2016 to 2022, the scale of computing power in mainland China increased by an average of 46% per year, the digital economy grew by 14.2%, and the GDP grew by 8.4%. According to IDC data, the scale of China's intelligent computing power will reach 414.1EFLOPS in 2023, with a compound growth rate of 33.9% from 2022 to 2027. At present, the country attaches great importance to the construction of computing power, and the AI demand + policy drives the continuous expansion of the intelligent computing power market. Domestic computing power is gradually accelerating, and the intelligent computing center is the key to the development of computing power. Intelligent computing power is artificial intelligence computing power, which is undertaken by various special chips such as GPU (graphics processing unit), ASIC (application-specific integrated circuit), FPGA (field programmable gate array), NPU (neural network processor) and other special chips to undertake the calculation of artificial intelligence scene applications, which has the advantages of better performance and lower energy consumption under specialized applications. Dozens of cities in mainland China are currently building or planning to build intelligent computing centers, with the overall layout mainly in the eastern region and gradually expanding to the central and western regions. In the future, with the continuous optimization and improvement of the layout of the intelligent computing center in mainland China, as well as the continuous innovation and unlocking of artificial intelligence application scenarios, the demand for intelligent computing power will be released to a greater extent, and the enabling role of the intelligent computing center will be further stimulated. With the continuous implementation of the intelligent computing center, related beneficiary links such as computing power optimization, liquid cooling and heat dissipation, domestic switches and other fields are expected to further expand the future performance growth space.

In the long run, the localization of computing power is the trend of the development of the domestic AI industry. According to iResearch's estimates, China's AI chip market will be about 38.5 billion yuan in 2022 and is expected to reach 216.4 billion yuan by 2027. According to IDC data, in 2022, China's AI accelerator card (open market) shipments will be about 1.09 million, of which Nvidia will account for 85% of the market, Huawei will account for 10%, Baidu will 2%, Cambrian and Suiyuan will all be 1%, and the localization of AI chips will have a broad space. At present, the main demanders of AI computing power in China include intelligent computing centers, Internet manufacturers, operators, financial and other industry customers. According to institutional research, these three types of customers have begun to actively seek domestic substitution of AI chips. As the localization of computing power continues to accelerate, it is recommended to pay attention to investment opportunities in Huawei's Kunpeng and Ascend industry chains.

1. Artificial Intelligence AIETF (515070) and its feeder fund (008585/008586)

The AI AIETF tracks the CSI Artificial Intelligence Theme Index (index code: 930713.CSI, index abbreviation: CS Artificial Intelligence) and selects representative companies from the companies that provide basic resources, technologies and application support for artificial intelligence as sample stocks, reflecting the overall performance of artificial intelligence theme companies. The index is included in the IOSCO Financial Benchmarks Principles Assurance Report as at 30 September 2018.

2. Chip ETF (159995) and its feeder fund (008887/008888)

The chip ETF tracks the CNI Semiconductor Chip Index (980017.CNI, index abbreviation: CNI Chip) to reflect the market performance of listed companies related to the chip industry in the A-share market. As a representative index of the semiconductor chip industry, the CNI Semiconductor Chip Index can reflect the market opportunities in the industry.

（2）

3. Cloud Computing ETF (516630) and its feeder fund (019868/019869)

The Cloud Computing ETF tracks the CSI Cloud Computing and Big Data Theme Index (index code: 930851.CSI, index abbreviation: Cloud Computing Index) and selects 50 A-shares of listed companies whose business involves the provision of cloud computing services, big data services and hardware equipment related to the above services as sample stocks to reflect the overall performance of cloud computing and big data theme stocks.

Data source: Minsheng Securities, Pacific Securities, Wind, as of 2024.2.22, the risk level of this product is R4 (medium and high risk), the above funds are index funds, and there are major risks such as the deviation of the underlying index return from the average return of the stock market, the fluctuation of the underlying index, and the deviation of the return of the fund portfolio from the return of the underlying index. and other specific risks such as the risk of deviation from the performance of the target ETF, and the historical performance of the market or the underlying product is not indicative of the future. Subscription: A fund will be charged a one-time subscription fee with no sales service fee when subscribing, while Class C will have no subscription fee but a sales service fee. There may be a big difference in the long-term performance of the two due to different fees and establishment times, etc., please refer to the product periodic report for details. Before investing in a fund, investors should carefully read the fund's "Fund Contract", "Prospectus" and "Product Key Facts Statement" and other fund legal documents, fully understand the risk-return characteristics and product characteristics of the fund, and fully consider their own risk tolerance according to their own investment objectives, investment period, investment experience, asset status and other factors, and make rational judgment and prudent investment decisions on the basis of understanding the product situation and sales suitability opinions, and independently bear the investment risk.

For ETF funds, investors investing in the Fund are exposed to potential risks such as tracking error control not reaching the agreed target, suspension of services by index compilers, suspension of trading of constituent bonds, the risk of deviation between the return of the underlying index and the average return of the stock market, the risk of fluctuation of the underlying index, the risk of deviation between the return of the fund's portfolio and the return of the underlying index, the risk of change in the underlying index, the risk of discount and premium of the trading price of the fund shares in the secondary market, the risk of error in the subscription and redemption list, the risk of incorrect decision-making with reference to IOPV and the calculation of IOPV, Delisting risk, the risk of failure of investors' subscription and redemption, the risk of realizing the redemption consideration of fund shares, the risk of derivatives investment, etc.

For ETF feeder funds, the fund's assets are mainly invested in the target ETF, and in most cases a high proportion of target ETF investment will be maintained, and the net value of the fund may fluctuate with the net value of the target ETF, and the risks associated with the target ETF may directly or indirectly become the risks of the ETF feeder fund. The specific risks of ETF feeder funds also include: tracking deviation risk, risk of performance difference with the target ETF, risk of discontinuation of services by the index compiler, risk of change of the underlying index, risk of suspension or default of constituent bonds, etc.

This material is not intended as any legal document, the views are for reference only, all information or opinions expressed in the material do not constitute investment, legal, accounting or tax advice, and our company does not make any guarantee for the final action advice on the content of the material. Under no circumstances shall the Company be liable to any person for any loss arising from the use of any content in this material. The market is risky, and you need to be cautious when entering the market.

National Business Daily