Punching above their weight: how China’s AI giants stretch each dollar in compute race
Chinese AI firms are narrowing the computing power gap with US counterparts despite lower spending, leveraging cheaper domestic costs and state support, according to a Moody’s Ratings report. This efficiency allows them to secure significantly more computing capacity per …
Intelligence analysis by Gemini 2.5 Flash
A new report by Moody’s Ratings suggests that the substantial spending gap between US and Chinese tech giants in artificial intelligence may not translate into the expected advantage for American firms. Chinese companies are reportedly 'punching above their financial weight' by securing far more computing power per dollar, thanks to lower buildout costs, targeted policy incentives, an…
Imagine two kids building with LEGOs. One kid has a huge allowance and buys tons of expensive sets. The other kid has less money but finds cheaper bricks, gets help from grown-ups, and uses clever tricks to build bigger things with less cash. That's like how Chinese AI companies are building powerful computers for AI – they spend less money than US companies but get more computing power because they're smart about how they buy and build things.
Analysis
The competitive landscape in artificial intelligence is often viewed through the lens of capital expenditure, with US tech giants historically outspending their Chinese counterparts by significant margins. However, a recent report from Moody's Ratings introduces a nuanced perspective, suggesting that this financial disparity does not fully reflect the actual gap in computing capacity. Chinese AI firms are demonstrating remarkable efficiency, effectively stretching each dollar to acquire more computing power, thereby narrowing the divide with American peers at a fraction of the cost.
Moody's Ratings
According to the Moody's Ratings report, while US hyperscalers' budgets dwarfed those of Chinese companies, the physical gap in computing capacity was not nearly as wide as the headline figures implied. This assessment is critical because it shifts the focus from raw spending to strategic resource allocation and operational efficiency. The report underscores that factors beyond mere investment size are playing a pivotal role in shaping the global AI compute race, challenging the assumption that financial might alone guarantees technological superiority.
Chinese tech firms benefit from several distinct advantages that enable this cost-effective expansion. Lower domestic buildout costs for data centers and infrastructure are a primary driver, allowing them to establish and scale computing facilities more economically. Additionally, targeted policy incentives from the state provide crucial support, directing resources and fostering an environment conducive to AI development. Access to cheaper green energy further reduces operational expenses, making the long-term running of these compute-intensive operations more sustainable and affordable.
US$140 billion
The report projects a significant increase in capital expenditure by China's major tech companies, with spending set to more than double to approximately US$140 billion this year, a substantial leap from US$65 billion in 2025. This upward trajectory is expected to continue, reaching US$165 billion by 2027. While these figures are still lower than those of their US counterparts, the context provided by Moody's suggests that this investment will yield disproportionately higher returns in terms of actual computing power. The ability to achieve more with less capital expenditure means that China's growing investments will have a magnified impact on its overall AI capabilities.
This strategic efficiency has profound implications for the global AI competition. Despite the US maintaining a clear overall lead in cutting-edge semiconductor chips, the findings from Moody's indicate that the compute race is not solely about chip supremacy or raw financial outlay. Instead, it's a multifaceted challenge where operational efficiency, government support, and cost advantages can significantly alter the competitive balance. China's approach allows it to build robust AI infrastructure, even with existing limitations in advanced chip technology, by optimizing every dollar spent on compute resources.
Key points
- Chinese AI firms are narrowing the computing power gap with US counterparts despite lower capital expenditure.
- A Moody's Ratings report indicates China secures significantly more computing power per dollar.
- Lower domestic buildout costs, targeted policy incentives, and cheaper green energy contribute to China's efficiency.
- China's major tech companies' capital expenditure is projected to more than double to US$140 billion this year, up from US$65 billion in 2025.
- The US still holds an overall lead in cutting-edge semiconductor chips, but spending figures don't tell the whole story.
China's ability to achieve more computing power with less capital could foster greater innovation and competition in the global AI landscape, potentially leading to more diverse AI developments and applications. This efficiency might also encourage other nations to explore similar cost-effective strategies for AI infrastructure development.
The reliance on state support and potentially less transparent cost structures in China could lead to an uneven playing field, making it harder for market-driven economies to compete purely on efficiency. Furthermore, the continued focus on compute power, regardless of cost, could intensify the global race for resources and talent, potentially exacerbating geopolitical tensions.
