AI data center power supply chain: Goldman Sachs favors "gas turbine blades and transformers" the most, which presents a major opportunity for Chinese companies of this type.
```
The global wave of AI data center construction is facing a severe “power bottleneck”, especially in the US market. Goldman Sachs believes this represents a historic opportunity for Chinese power supply chain companies with advantages in technology, cost, and delivery speed.
According to WindChaser Trading Desk, on January 14th, Goldman Sachs research team released a report showing that global AI data center electricity demand is surging, but there is a severe shortage of power generation equipment such as gas turbines and grid equipment such as transformers, with delivery times as long as several years.
With orders for European and American suppliers scheduled out to 2028, the shorter delivery cycles of quality Chinese suppliers have become a core competitive advantage. Goldman Sachs sees that the supply shortage-driven pricing premium allows Chinese suppliers to sell abroad at prices 10% to 80% higher than domestic, and even after counting tariffs and increased logistics costs, profit margins are still significantly improved.
In terms of product category preference, Goldman Sachs ranks: Gas Turbine Blades > Power Transformers > Electrical Components > UPS/Power Racks > Liquid Cooling > Server Power Supply Units (PSU).
Goldman Sachs estimates that the average compound annual growth rate (CAGR) of sales for Chinese companies during 2025-2030 may reach 23%, and by 2030 the overseas AI data center market will account for an average 23% of their revenue, with a global market share averaging 4%.
Global AI Data Center Capacity Surging, but US Power Bottleneck Is More Severe
Goldman Sachs points out that global AI data center (AIDC) capacity is rapidly increasing, but the US power bottleneck issue is far more severe than in other countries.
The report projects that from 2025 to 2030, global data center capacity will add about 73 GW, with the bulk of the increase concentrated in the US and China.
Despite the surge in demand, the US is facing severe challenges in power supply, with the bottleneck effect more prominent than in places like China:
Local Supply Severely Inadequate: The capacity of US domestic power equipment can only meet about 40% of its local demand, with a huge supply gap.Tight Grid Reserve Capacity: Compared to China, the US has much tighter effective power reserve capacity. The report shows that from 2025 to 2030, US grid average reserve is about 100 GW, and shortages will exacerbate; in the same period, China’s average reserve capacity is up to 300 GW, making its system much more abundant.Extremely High Growth in Demand: The report cites internal team forecasts that by 2030, US data center (AI and non-AI) power consumption will increase about 175% from 2023, far exceeding the national average compound growth rate of 2.6% for overall power demand.Forced to Seek Non-Traditional Suppliers: Because key equipment (such as gas turbines, transformers) have delivery cycles up to 3-5 years, US power companies and data center operators are increasingly willing to work with non-traditional suppliers to fill multi-year supply gaps. This creates market opportunities for qualified Chinese suppliers capable of rapid delivery.
Globally, the expansion of AI computing power is shifting from “how many GPUs are needed” to “where does the power come from and how fast can it be delivered.”
Under this trend, the US—due to limited domestic supply, tight grid reserves, and extremely rapid demand growth—is the region with the most severe power bottleneck, which makes suppliers capable of providing "Time-to-Power" solutions highly valuable.
Severe Shortage of Power Generators and Transformers to Continue Through 2028-2030
Goldman Sachs highlights that power generators and transformers are the most constrained supply chain bottlenecks in global AI data center construction, and this supply shortage cycle is expected to last through 2028-2030.
The report notes years-long shortages in power generation (especially gas turbines) and grid transmission (especially transformers). Capacity expansion is much slower than initially planned, while demand spikes, causing the supply-demand gap to persist.
Based on technological barriers and order pipeline, the report concludes the tight supply of gas turbines (especially blades) will last at least until 2028-2030.
Gas turbine blades must withstand extreme temperatures and physical stresses, involving single-crystal high-temperature alloys, specialized coatings, complex aerodynamic design, and intensive front-end manufacturing processes (such as wax tree making and assembly). This makes blades one of the highest technical barriers and slowest to expand capacity.
In addition, due to high customization and labor intensity, transformers face the most severe bottleneck.
The average age of US and European power grids is already 35-40 years, which struggles to meet the soaring power demand of the AI era, further increasing pressure on transformer replacement and upgrades.
Each transformer requires custom design for specific impedance, cooling, tap changing, overload, and earthquake resistance standards. This “one design per unit” mode makes production highly dependent on skilled labor and difficult to quickly expand output via capital expenditure.
Major Migration Toward 800V DC Architecture Driven by Technological Change
As AI cabinet power density breaks through 100kW and heads toward 1MW, conventional copper wire transmission approaches the physical limit.
To carry massive current, data center electrical architecture is undergoing a dramatic transformation: moving from traditional AC to 800V DC architecture, which can save roughly 5-15% in energy consumption.
Driven by capacity build-out and technology upgrade, Goldman Sachs expects power supply products for AI data centers covered under China’s industrial tech will maintain a 39% compound annual growth rate before 2030.
The research report emphasizes this will be especially favorable for component suppliers with DC tech reserves, particularly leaders in relay and capacitor sectors.
Chinese Suppliers Seize Overflow Demand with Fast Delivery and Competitive Products
With waiting times of 3-5 years for key components, Chinese suppliers are uniquely positioned to capture overflow demand.
Goldman Sachs believes the core competitive advantage of Chinese suppliers lies not just in cost, but in faster delivery and competitive products. Specifically:
Delivery Speed: This is the decisive factor for Chinese suppliers to seize opportunities. Faced with multi-year waits, end customers now prioritize “delivery speed” over historical supplier preference. Chinese suppliers who can provide reliable products and sharply shorten delivery cycles will be favored in the market.
Product Competitiveness: The report stresses the opportunity belongs only to a handful of leading Chinese companies that meet international standards in technical strength, quality control, and scale production. These firms typically specialize in HVDC tech, high-density power conversion, and their products meet the rigorous technical standards and rapid expansion needs of AI data centers.
Analysis finds that this speed- and product-driven market entry can translate into significant financial benefits.
Compared to domestic sales, the price premium on key product categories for Chinese firms can reach 10–80%. Despite extra tariffs and logistics costs, the large premium creates significant gross margin expansion, delivering highly visible profit improvement for suppliers.
Goldman Sachs analyzes globally competitive Chinese suppliers able to seize these opportunities, mainly involving six product categories:
Gas Turbine Blades: Capitalizing on advanced manufacturing technical barriers in single-crystal high-temperature alloys, specialized coatings, and complex aerodynamic design, these products are among the highest in technical threshold and supply chain tightness globally.Power Transformers: Due to high customization, labor intensity, and lengthy certification cycles, these are the most severe grid-side bottlenecks. Chinese suppliers, with shorter delivery cycles, enjoy key advantages in meeting overseas demand.HVDC (High-Voltage Direct Current) Relays: As critical components for data centers upgrading to 800V DC architecture, suppliers with advanced HVDC expertise and tech can secure highly visible global market share here.Aluminum Electrolytic Capacitors and Supercapacitors: These electrical components are fundamental to data center power systems (such as UPS, rectifiers, power factor correction), with demand responding quickly to AI data center build-out. Supercapacitors especially serve as accelerators for peak load modulation in the medium to long term.Uninterruptible Power Supply / DC Power Rack Systems: Through ODM (Original Design Manufacturer) models, these can provide high-power supply solutions to overseas markets (especially the US), and are actively developing next-generation 800V DC architecture products.Liquid Cooling Solutions: With AI server power density rising sharply, demand for advanced liquid cooling systems (including cold plates, manifolds, CDU/RPU, etc.) for chips like GPU/ASIC is surging, giving these products structural growth space and value creation.
~~~~~~~~~~~~~~~~~~~~~~~~
The above content comes from WindChaser Trading Desk.
For a more detailed interpretation, including live commentary and frontline research, please join the WindChaser Trading Desk Annual Membership
Risk Warning and DisclaimerThe market carries risks, and investment should be cautious. This article does not constitute personalized investment advice, nor does it consider the particular investment goals, financial circumstances, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their specific situation. Investing based on this carries your own responsibility.
```