The competition between the United States and China over technology supremacy has moved well beyond trade tariffs and export controls. What began as a dispute over market access has hardened into a structural effort by Washington to limit Beijing’s access to advanced semiconductor technology, AI infrastructure, and critical hardware supply chains. The fault lines are now drawn in silicon, software, and the global networks that connect them.
The Export Control Architecture
The October 2022 export control rules issued by the U.S. Bureau of Industry and Security (BIS) marked a decisive shift. For the first time, Washington moved to comprehensively restrict China’s ability to acquire or produce advanced semiconductors — chips at or below the 14-nanometer threshold, along with the equipment needed to manufacture them. The rules were expanded in October 2023 and again in 2024, closing loopholes that had allowed Chinese firms to source chips through third-country intermediaries.
The controls target three layers simultaneously: the chips themselves, the equipment used to fabricate them (particularly extreme ultraviolet lithography machines from Dutch firm ASML), and the software tools required for chip design. This multi-layer approach reflects a deliberate strategy to deny China not just today’s technology, but the industrial capability to develop next-generation chips independently.
The successive BIS rule expansions documented by Reuters trace a progression from targeted sanctions against specific firms (Huawei, SMIC) toward a broader regime aimed at the entire Chinese advanced semiconductor sector.
China’s Countermoves
Beijing’s response has followed two tracks. The first is domestic substitution: massive state investment in home-grown chip design and fabrication. The China Integrated Circuit Industry Investment Fund — the so-called “Big Fund” — has channeled over $47 billion into the sector across two rounds of investment. State-backed firms including SMIC and CXMT are pursuing process nodes that lag the global frontier but are sufficient for a wide range of military, industrial, and consumer applications.
The second track is international sourcing. Chinese procurement networks have adapted, using front companies in Southeast Asia, the Middle East, and Eastern Europe to acquire restricted components. The RAND Corporation’s analysis of technology transfer risks highlights the persistent difficulty of enforcing export controls against determined state actors with deep commercial networks.
In 2023, Huawei’s release of the Mate 60 Pro — powered by a 7nm chip apparently fabricated by SMIC despite U.S. controls — demonstrated that China retains meaningful semiconductor capability, even if it remains well behind the leading edge. The episode embarrassed U.S. policymakers and intensified debate over the controls’ effectiveness.
The Allied Dimension
Washington has worked to multilateralize its controls, persuading the Netherlands and Japan to restrict exports of certain lithography equipment to China. However, allied cooperation has limits. European and Japanese firms have significant commercial exposure to the Chinese market, and governments in both regions have pushed back against what they see as unilateral American pressure to adopt broad restrictions that impose asymmetric costs on their industries.
The Financial Times has documented ongoing friction between the U.S. Commerce Department and European counterparts over the pace and scope of controls. ASML, whose extreme ultraviolet machines are essential to cutting-edge chip production, has faced Dutch government pressure to continue some sales to China even as Washington pushes for a complete cutoff.
South Korea and Taiwan — home to Samsung, SK Hynix, and TSMC — occupy particularly complex positions. Their governments have nominally aligned with U.S. policy, but their companies derive substantial revenue from China. TSMC alone generated approximately 10 percent of its revenue from Chinese customers before tightening its own compliance policies in 2023.
The AI Infrastructure Layer
The semiconductor competition intersects directly with the race for artificial intelligence capability. Advanced AI training requires massive clusters of high-performance GPUs — predominantly NVIDIA products. BIS has implemented successive rounds of controls targeting NVIDIA’s A100, H100, and subsequent chip generations for export to China.
NVIDIA responded by developing downgraded variants (the A800, H800) for the Chinese market, but BIS closed that loophole in October 2023. The result has been a scramble by Chinese AI firms — Baidu, ByteDance, Alibaba, Tencent — to stockpile available chips and to accelerate development of domestic alternatives. Huawei’s Ascend chips have emerged as the primary domestic substitute, though they lag NVIDIA’s performance benchmarks by a significant margin.
The Council on Foreign Relations backgrounder on the U.S.-China tech war frames this as a structural competition that will shape military, economic, and political power for decades. The assessment is well-founded: AI capability increasingly underpins everything from autonomous weapons systems to economic productivity gains.
Key Indicators to Watch
- Progress of SMIC’s sub-7nm process development and any evidence of yield improvements at advanced nodes
- Further BIS rule expansions targeting additional chip categories, memory, or new country-specific measures
- Huawei Ascend chip adoption rates among Chinese AI developers — a proxy for domestic substitution success
- Dutch and Japanese compliance with U.S. pressure on ASML and Tokyo Electron export restrictions
- Evidence of third-country diversion schemes and any resulting enforcement actions
- Chinese government announcements on Big Fund III capitalization and deployment
- TSMC and Samsung exposure management — any shifts in revenue mix away from China
Bottom Line
The U.S.-China technology decoupling is now structural and largely irreversible in its core architecture. Washington has demonstrated the political will to accept commercial costs in exchange for constraining China’s advanced chip access, and Beijing has accepted that self-sufficiency — however costly and slow — is its only viable path. The critical uncertainty is the speed of China’s domestic progress: if SMIC and its peers can close the process node gap within five to seven years, the controls will have bought time rather than a durable advantage. This dynamic connects directly to broader questions of industrial policy competition across Asia, where nations from India to Vietnam are now positioning to capture supply chain diversification opportunities.
Leave a Reply