Moody's: Chinese AI Giants Double Capital Expenditure, Computing Power Gap Narrowing Rapidly
nashnova research
A new Moody's report says Chinese tech firms will roughly double AI capex from $65 billion to $140 billion this year — and thanks to lower build-out costs, the real computing-power gap with the U.S. is far smaller than the headline spending numbers imply.
How fast is Chinese AI capex growing?
Moody's projects Chinese tech majors will lift capex from roughly $65 billion in 2025 to about $140 billion this year — more than doubling.
By 2027, that figure is expected to rise further to approximately $165 billion.
This means → China's AI infrastructure push is a multi-year acceleration cycle, not a one-off surge.
Does spending less mean less computing power?
Moody's states plainly: U.S. hyperscalers outspend Chinese peers by a wide margin, yet the actual computing-power gap is far smaller than the dollar figures suggest.
In plain terms = judging AI competitiveness by capex alone significantly overstates America's lead.
Chinese firms benefit from three cost advantages: lower infrastructure build-out costs, targeted policy incentives, and cheaper green energy — delivering more compute per dollar spent.
This reflects a deeper point: the AI race hinges not on "who spends more" but on "how much compute each dollar buys."
Is America's chip edge enough?
Moody's acknowledges the U.S. retains a clear overall lead in cutting-edge semiconductor chips.
But the report's central question is whether that chip-technology advantage can sustain America's lead in the AI compute race as Chinese firms keep improving efficiency.
This means → Moody's sees the technology barrier as real — but its effective height is being steadily eroded by efficiency gains on the other side.
市场有风险，内容仅供研究参考，不构成投资建议。