US–China chip rivalry deepens as Beijing narrows gap with lower-cost AI push

Four years after the United States imposed sweeping export controls on advanced semiconductors, China is accelerating efforts to build a self-reliant chip industry—reshaping the global technology landscape and intensifying competition between the world’s two largest economies.

The restrictions, introduced under the Biden administration, targeted high-end chips used in artificial intelligence (AI), data centres and defense systems, with the aim of slowing Beijing’s technological rise. In response, China has doubled down on its long-standing ambition for semiconductor independence, outlined in its “Made in China 2025” strategy, according to DW.

Beijing has since poured hundreds of billions of dollars into domestic chip production, offering subsidies, tax incentives and other support to nurture local champions capable of competing with firms such as NVIDIA and TSMC.

China’s leading foundry, SMIC, reported record revenues of $9.3 billion last year, while HuaHong has been operating above full capacity, reflecting surging demand.

Despite this momentum, analysts say China remains far from achieving full self-sufficiency.

“Beijing wants to achieve chip self-sufficiency, but the current level is nowhere near it,” said Ryu Yongwook, an expert on US-China tech rivalry at the National University of Singapore.

The country still lags behind the US in research and chip design, and trails Taiwan and South Korea in advanced manufacturing capabilities.

Even so, China has made notable gains. According to the Rhodium Group, it now holds around 30% of the global market for so-called legacy chips—less advanced semiconductors widely used in cars, industrial machinery and consumer electronics.

“Chinese production expansion will drive down [chip] prices globally and put pressure on non-Chinese vendors,” said John Lee of East-West Futures.

“This is already happening in some sectors, such as silicon carbide wafers,” he added.

China has also advanced in higher-end chips, producing 7-nanometer-class processors now used in smartphones by Huawei. While comparable to chips produced by TSMC in 2018, they still fall short of today’s leading-edge 3-nanometer and 5-nanometer technologies in performance and efficiency.

Tim Rühlig of the European Union Institute for Security Studies said China faces structural limits.

“There is only so much that you can do without access to the US's most advanced chipset,” he said, adding that China may need “a decade or so” to catch up.

Facing these constraints, Beijing appears to be recalibrating its strategy. Rather than focusing solely on cutting-edge chips, it is prioritising practical AI applications that can run on less advanced hardware. This approach is gaining traction, particularly in developing markets where cost is a key factor.

Taipei-based TrendForce estimates Chinese AI platforms—including DeepSeek and Alibaba’s Qwen—captured about 15% of the global AI model market by late 2025.

This growing footprint poses a challenge to US tech giants such as Microsoft and Google, which are expected to invest heavily in AI infrastructure.

At the same time, structural factors could shape the outcome of the rivalry. The global market intelligence provider ICIS has warned that US data centres may face constraints due to pressure on the power grid. In contrast, China’s expanding energy capacity could support large-scale AI deployment.

“Cheap energy is a very important factor, not necessarily for chips but for AI and other advanced technologies,” Ryu said. “Cheap energy in China goes some way to make up for its relative chip inefficiency.”

Analysts say the race could evolve along several paths, ranging from continued US leadership to a split into rival technological ecosystems.

For now, the trajectory points to intensifying competition, as Chinese firms gain ground by offering lower-cost alternatives while steadily improving their technological capabilities.

By Sabina Mammadli