AI Photo: VCG

A Bloomberg report recently claimed that almost half of the US data centers planned for this year are expected to be delayed or canceled due to severe shortages of critical electrical components. This predicament serves as a prism, reflecting a truth that is often overlooked: The ultimate limit of computing power is electricity, while the ultimate limit of electricity is energy strategy. America's AI prowess on computer chips and cutting-edge software is being hamstrung by the country's inability to manufacture the electrical parts.

In the field of AI, the US may take a temporary advantage in the development of advanced-process chips and cutting-edge large-scale models, but without the support of a robust energy infrastructure, these advantages will prove difficult to translate into tangible computing capacity and industrial competitiveness. The key to winning the AI race lies in access to energy. The challenges currently facing the US - particularly in the manufacturing of power equipment and the large-scale deployment of green energy - are being significantly amplified by the explosive growth in demand for AI computing power.

Viewed from a deeper perspective, this awkward "power shortage" is not merely a technical or industrial issue. Rather, it represents the inexorable response of the laws of economic globalization to the US' unilateralist policies. When selecting sites, many US AI data centers prioritize - above all else - the stable power generation. Although electrical components such as transformers, switchgear, and batteries account for less than 10 percent of total project costs, they are an absolute necessity for both construction and operation. Domestic manufacturing capacity for these equipment in the US is woefully insufficient, with a significant portion relying on imports from China. 

For a long time, the US has routinely invoked "national security" as a pretext to impose unilateral tariffs on relevant Chinese products, as well as implementing draconian import controls, and conducting so-called security reviews. Through a strategy of "small yard, high fences" and "decoupling," it has attempted to forcibly exclude China from its critical supply chains. 

Yet, the ultimate result has been to leave its own cutting-edge AI strategic projects with a "chokehold" around their necks.

The tide of economic globalization has long since permeated every capillary of global industrial and supply chains; any attempt to swim against this current and to construct a self-contained, closed system is ultimately destined to falter under the immense burden of prohibitive costs, inefficiencies, and time delays, and eventually backfires upon itself. 

The AI industry chain - spanning from upstream chip design and manufacturing and key equipment production and material supplies to midstream computing infrastructure layout and power supply, and extending to downstream application deployment and data flow - is already deeply embedded within a global network of specialized division and collaboration. The recurring struggles plaguing the development of AI data centers in the US not only expose the futility of the US industry's push to "decouple from China," but also clearly reflect the new characteristics of globalization emerging within the AI supply chain.

As a key global supplier of power infrastructure equipment, China plays an indispensable role within the AI data center industry chain. Attaining this status is no mere coincidence; rather, it reflects China's comprehensive advantages across the entire industrial chain. In terms of industrial layout, China controls virtually the entire supply chain for electrical equipment - spanning everything from raw material extraction and processing to finished product manufacturing - thereby establishing a complete industrial ecosystem. This enables the stable, large-scale supply of the diverse electrical equipment required by AI data centers, such as transformers, switchgear, and batteries, a distinct advantage that few other nations can rival.

In today's world characterized by an increasingly intricate global division of labor and rapid technological iteration, no single nation or bloc can dominate the landscape or unilaterally control the entirety of a supply chain. As the competitive advantage of Chinese enterprises gradually shifts from low labor costs toward core technologies and standard-setting, the "piece of the puzzle" represented by "Made in China" has become increasingly indispensable within global production and supply chains. 

If the US wishes to resolve its current predicament regarding power shortages, the only viable path is to abandon its fixation on unilateralism and return to a course that respects economic laws, adheres to the principles of comparative advantage, and seeks mutual benefit through open cooperation. Given the strong complementarity between their respective industrial sectors, Chinese enterprises can engage with the US market through models such as localized manufacturing and technological collaboration to achieve mutually beneficial outcomes and jointly drive innovative development within the global energy storage industry.

The current predicament facing AI data center construction in the US serves as a stark reminder: Mutually beneficial cooperation is the fundamental path to driving the development of the global data industry, whereas unilateralism and trade protectionism will only hinder industrial progress and ultimately undermine the own interests of those who apply it. 

As the global AI race continues to intensify, particularly at the new critical juncture where computing power and electricity supply are inextricably intertwined, the US faces a pivotal choice. Will it continue to endure the embarrassment of "power shortages" amid the push for so-called de-risking from China, thereby eroding its own competitive advantages and development opportunities? Or will it abandon zero-sum thinking and proactively integrate into the new global trend of digital economic development, which is grounded in open cooperation? This constitutes a true test of America's developmental mindset and strategic vision.