The demand for AI computing power drives the increase in electricity demand; however, each power supply method for Datacenters has its own shortcomings. It is expected that clean, efficient, and cost-effective gas turbine power generation will become the best choice.
According to Zhito Finance APP, Soochow Securities released a Research Report stating that the demand for AI computing power drives the increase in generator demand, but the power supply methods for datacenters each have their flaws. It is expected that clean, efficient, and low-cost Gas Turbine generation will become the best choice for powering AI computing centers outside of China. Due to the imperfect infrastructure of the global power grid, Gas Turbines are generally the main power supply equipment for datacenters. The firm is Bullish on the potential benefits for Chinese equipment manufacturers in the supply chain. It is recommended to achieve sales of electric-powered fracturing equipment paired with Gas Turbine Generator sets in the North American market, and to sign exclusive agreements with leading Gas Turbine companies like Yantai Jereh Oilfield Services Group (002353.SZ). It is suggested to pay attention to overseas leading components suppliers like Anhui Yingliu Electromechanical (603308.SH), Liande Co., Ltd. (605060.SH), and domestic replacement suppliers for Diesel Engines like Weichai Heavy Machinery (000880.SZ).
The main viewpoints of soochow securities are as follows:
The explosion of AI computing power brings about an increase in Generator demand, and the gas turbine Generator Industry Chain is expected to benefit significantly.
The demand for AI computing power is driving the increase in generator demand. According to Gartner's predictions, by 2027, the electricity demand for running AI-optimized servers is expected to reach 500 terawatt-hours per year, which is 2.6 times the level of 2023. The power supply methods for datacenters include grid power, Diesel generation, Gas Turbine generation, Nuclear Power, geothermal, and Wind Power, but the construction cycles for geothermal, Nuclear Power, and Wind Power are relatively long, combined with the imperfect global power grid infrastructure. Therefore, Gas and RBOB Gasoline are a better short-term solution to electricity shortages. However, Diesel generation will produce pollution issues; thus, clean, efficient, and low-cost Gas Turbine generation has become the best choice for powering AI computing centers outside of China. For instance, Musk opted to use Gas Turbines to supply power to datacenters due to grid issues, and datacenters in Ireland have also chosen to directly use Gas Turbines for power supply, leading to the expansion of the entire industry chain market scale.
USA Market: Gas turbines are the main power supply, with a bullish outlook for suppliers in China within the Industry Chain regarding potential benefits.
The USA is the largest AI computing power market globally, with enormous future development potential. Due to the imperfect power grid infrastructure in the USA, Gas Turbines are generally the main power supply equipment for datacenters. The firm is Bullish on the potential benefits for Chinese equipment manufacturers in the supply chain. ① Yantai Jereh Oilfield Services Group (integrator, buying Gas Turbines for generator set integration): The company can supply Gas Turbine Generator sets paired with electric fracturing equipment for the oil service industry in the North American market, possessing industry chain experience. The company is actively expanding into the datacenter sector, with Caterpillar's subsidiary Solar as its core competitor. ② Anhui Yingliu Electromechanical: The company can produce Gas Turbine Blades and is bound to global Gas Turbine leaders GE and Siemens, providing great flexibility as a component supplier. ③ Liande Co., Ltd.: The company is a supplier of castings for Caterpillar Gas Turbines and Diesel Engines and will benefit directly.
China market: Diesel engines/gas turbines serve as backup power sources, and there are bullish prospects for domestic replacement opportunities due to the growth in computing power demand.
Unlike the USA market, China has a well-established electrical utilities infrastructure. The power supply for domestic datacenters can be solved by the grid, requiring only diesel engines/gas turbines as backup power. Currently, the domestic diesel engines/gas turbines are still dominated by overseas leaders, but with limited production capacity from foreign enterprises, there are bullish breakthrough opportunities for domestic replacements driven by the increasing demand for computing power. It is recommended to pay attention to Weichai Heavy Machinery and China Yuchai International, among others.
Risk Warning: Datacenter capital expenditures are below expectations; demand for Gas Turbine Generators is lower than expected; international trade disputes.
