China's Semiconductor Bottlenecks Are Disappearing

Washington's semiconductor debate increasingly centers on the question: Can China build an extreme ultraviolet lithography machine? EUV systems are essential for producing the world's most advanced chips efficiently, and China still cannot manufacture them at scale. That constraint matters. But a strategy built around one machine risks missing the larger change underway across China's semiconductor industry.  Meanwhile, making smaller chips is increasingly less important to the future of advanced semiconductors as physics constraints slow progress and advanced packaging creates a new avenue for improvement. 

Photoresist offers a useful example. This light-sensitive material allows manufacturers to print circuit patterns onto semiconductor wafers. Foreign suppliers still provide roughly 92 percent of the photoresist used by China's chipmakers, and Chinese producers remain especially weak in the most advanced grades. Yet one domestic supplier that began developing photoresist for immersion deep ultraviolet lithography in 2020 had reportedly qualified its product by late 2025 for use by leading chipmakers ChangXin Memory Technologies and Semiconductor Manufacturing International Corporation. As production increases, Chinese-made photoresist is likely to take a larger share of the domestic market.

China has also made gains in the equipment used to etch and deposit thin layers of material during fabrication. These tools become more important when manufacturers lack EUV and must repeat patterning steps to produce advanced chips. Shenzhen-based SiCarrier, which has close ties to Huawei, has emerged as an important domestic supplier. Its tools are supporting expanded production of advanced logic chips, while similar capabilities are becoming more important for memory manufacturing.

Serious constraints remain. Ion implantation and metrology are two of the hardest. China produces only a few dozen ion implanters each year and still depends heavily on imports. Of the 448 machines China imported in 2025, 306 came from the United States. The gap is largest in high-energy ion implanters used for production at 14 nanometers and below, where Chinese companies have not yet established mature alternatives. Metrology systems, which measure and inspect chips during production, present another durable challenge.

Domestic substitution also does not always eliminate foreign dependence. It often moves that dependence one step upstream. A Chinese equipment maker may replace a foreign machine while still relying on imported components, intermediate products or specialized materials. 

Policymakers should therefore look beyond the nationality of the final supplier and map the full chain of inputs that makes production possible.

The policy implication is straightforward. Export controls are no longer able to provide a durable advantage against an industrial system designed to replace foreign technology. At most, restrictions may slow progress in selected areas for a limited period. They also intensify China's incentive to develop substitutes, encourage Chinese customers to abandon foreign suppliers and reduce the revenue available for research and development by U.S. and allied companies. Continuing to treat controls as the centerpiece of strategy risks weakening the industrial base the United States needs to lead.

The United States should compete by widening its technological and commercial lead. That means expanding domestic production, supporting research, training workers, strengthening allied supply chains and keeping global markets open to trusted U.S. technology. Policymakers should track China's progress across equipment, materials and upstream components so public investment follows the capabilities that matter most. Protecting national security requires specific safeguards for genuine military risks, not broad commercial restrictions that encourage faster localization.

China has not eliminated its semiconductor vulnerabilities. Yet those vulnerabilities are shrinking, and several may disappear faster than policymakers expect. Over the next five to ten years, China is likely to develop a highly localized semiconductor ecosystem that challenges international suppliers across many parts of the industry. The United States will need a faster rate of innovation and a stronger industrial base because regulation cannot freeze today's supply chain in place.

Jost Wübbeke is Managing Partner at Sinolytics, a management consultancy focused on China. He is an expert on China’s industrial policy, supply chains, the automotive industry, and semiconductors. He advises board-level executives at multinational corporations and SMEs on navigating China’s political and business landscape amid growing geopolitical complexity. He holds a PhD on China’s industrial policy from the Free University of Berlin, was a research fellow at Tsinghua University, and studied International Relations and East Asian Economics in Berlin.

image: A schematic representation shows an electron under 12-qubit quantum dot gates. Fabricated on 300-millimeter wafers, Tunnel Falls leverages Intel’s most advanced transistor fabrication capabilities, such as extreme ultraviolet lithography (EUV) and advanced materials processing techniques. This makes the chip a single electron transistor and allows Intel to fabricate Tunnel Falls with few changes to a standard complementary metal oxide semiconductor (CMOS) logic processing line. (Credit: Intel Corporation)