When Washington blocked China from buying advanced chips, the goal was to slow Beijing’s AI ambitions. It may have accelerated them instead.

CXMT Corp, a memory chipmaker once dependent on foreign technology, debuted on Shanghai’s STAR Market last month and surged 466% in a single session. The stock eclipsed Industrial and Commercial Bank of China to become the country’s most valuable listed company. CXMT raised $9.8 billion in one of China’s biggest IPOs in years, with proceeds going toward the chip capacity Beijing needs to win the AI race.

From Containment to Catalyst

US export controls cut off China’s access to the semiconductors that power AI systems. The logic was direct: no advanced chips meant no path to AI leadership. When Washington announced its crackdown on AI chip exports to China, the restrictions targeted Nvidia, AMD and their successors. The result, according to CXMT’s trajectory, was that China built rather than bought.

Beijing cleared the runway fast. CXMT went from filing to trading in under eight months through a preliminary review process reserved for strategically vital companies, a pace that normally takes years. When tech stocks slid in July and threatened to derail the debut, regulators and state funds intervened within days to stabilize sentiment.

Chris Miller, author of Chip War and professor at Tufts University, acknowledged the bind. “US firms have had greater access to capital, but financing costs are rising,” he told Bloomberg. Chinese domestic compute is still more expensive because domestic chips lag on quality, but the gap is closing.

The Cost Advantage America Didn’t See Coming

Beijing also rewired how it funds the race. China traditionally backed strategic industries with subsidies and state lending. Now it taps the $26 trillion sitting in household savings accounts, the largest pool of private capital in the world.

Chinese tech firms borrowed at an average coupon of 1.9% this year, more than 300 basis points below US peers paying 5.25%, the widest gap since at least 2015. That cost advantage compounds every quarter.

The IPO pipeline behind CXMT is long. DeepSeek is weighing a listing at a $71 billion valuation. Moonshot AI, whose Kimi K3 model rattled Silicon Valley, is filing in Hong Kong. Z.AI and MiniMax are next. UBS says China’s leading AI models cost less than 10% of what OpenAI and Anthropic spend to train. API prices for major Chinese models run below 20% of comparable global offerings.

Washington set out to limit China’s options. It may have forced Beijing to build better ones.

Source: BeInCrypto