China's Chip Output Set to Triple by 2035 as Lithography Remains the Weak Link

The Production Surge Nobody Expected

China's semiconductor industry is preparing for a production explosion that will reshape the global chip landscape over the next decade. Domestic foundries working at 7-nanometre nodes and below are set to expand output at a 46 per cent compound annual rate through 2035, according to Goldman Sachs analysis. That pace dwarfs the 17 per cent growth trajectory anticipated for the rest of the world, signaling a dramatic narrowing of the advanced chip supply gap that has long constrained Beijing's technology ambitions.

At DailyTechWire, we've tracked the steady drumbeat of fab announcements across Wuxi, Hefei, and Shenzhen over the past eighteen months. What Goldman's projection clarifies is the scale at which domestic capacity will come online once yield rates stabilize and process recipes mature. The numbers suggest China could triple its share of global advanced node output by the middle of the next decade, a shift with profound implications for supply chain geography, export control enforcement, and the competitive position of incumbents in Taiwan and South Korea.

Yet the same analysis underscores a stubborn constraint: lithography equipment, the precision tooling required to pattern circuits at the atomic scale, remains the sector's Achilles heel. While production volume is climbing, the inability to secure or replicate cutting-edge exposure systems means China's foundries will continue to depend on workarounds, legacy tools, and incremental domestic substitutes that lag the frontier by multiple generations.

Why Lithography Still Dictates the Ceiling

Advanced chip manufacturing hinges on extreme ultraviolet lithography, the technology that enables sub-10-nanometre patterning. A single Dutch firm holds effective monopoly control over EUV systems, and export restrictions introduced in 2019 and tightened repeatedly since have locked Chinese fabs out of the latest platforms. The result is a bifurcated strategy: foundries push older deep ultraviolet tools to their physical limits through multi-patterning techniques, while state-backed consortia pour resources into developing indigenous EUV alternatives that remain years from commercial readiness.

Goldman's forecast acknowledges this bifurcation. The 46 per cent growth rate reflects aggressive capacity expansion using available equipment, much of it procured before controls took full effect or sourced through secondary channels. But the bank's analysts note that yield penalties and higher production costs associated with DUV-based advanced nodes will persist, capping margin potential and limiting the types of chips Chinese foundries can produce economically. High-performance computing processors and AI accelerators, which demand bleeding-edge density and power efficiency, remain particularly challenging without EUV access.

The lithography bottleneck also shapes investment priorities. Beijing has funneled an estimated tens of billions of dollars into equipment R&D since 2020, with lithography receiving the largest share. Progress has been incremental: domestic immersion DUV systems have entered pilot production at select fabs, but performance and reliability still trail established vendors. EUV remains further out, with prototype light sources and optics still in laboratory phases. Industry observers expect a five- to seven-year gap before Chinese EUV tools, if they arrive, reach parity with current-generation foreign systems, by which time the frontier will have moved further ahead.

Demand Drivers and the Domestic Market Cushion

What makes the 46 per cent growth rate plausible is the sheer scale of China's captive semiconductor demand. The country consumes more than half of global chip production, much of it imported. Automotive electronics, consumer devices, industrial automation, and data centre infrastructure all represent massive, price-sensitive markets where domestic foundries can compete on cost and proximity even if they trail on absolute performance.

The automotive sector has emerged as a particularly strategic beachhead. Electric vehicle platforms require hundreds of chips per vehicle, spanning power management, sensor fusion, battery control, and infotainment. Many of these functions can be served by 28-nanometre and 14-nanometre nodes, sweet spots for Chinese foundries that have achieved stable yields and secured domestic equipment for much of the process flow. As EV production ramps across Guangdong, Jiangsu, and Anhui, the co-location of chip supply and automotive assembly creates a vertically integrated advantage that offshore foundries struggle to match on delivery time and customization.

Consumer electronics present a similar, if more contested, opportunity. Handset processors and application-specific chips for smart home devices increasingly incorporate domestically sourced silicon, driven by both cost considerations and supply chain resilience mandates. While flagship smartphone models still rely on imported cutting-edge chips, the mid-tier and budget segments, which account for the majority of unit volume in China and across Southeast Asia, are transitioning to domestic nodes. This segment alone can absorb substantial foundry capacity and sustain utilization rates that justify continued expansion.

The Geopolitical Overhang

Goldman's projection arrives against a backdrop of escalating export controls and reciprocal measures. Washington has expanded the list of restricted chip-making tools multiple times since 2022, adding inspection, deposition, and etch systems to the original lithography bans. Allied governments in Tokyo, Seoul, and The Hague have aligned restrictions on their own equipment makers, tightening the net around China's equipment supply chain.

Beijing's response has been two-pronged: accelerate domestic substitution and cultivate alternative suppliers in jurisdictions less susceptible to US pressure. The former strategy underpins the 46 per cent growth forecast, as state-directed capital flows into fab construction and equipment R&D regardless of near-term return on investment. The latter has seen outreach to equipment makers in Europe and Asia, though results have been mixed as governments close loopholes and firms weigh reputational and regulatory risk.

The risk of further restrictions looms over the entire trajectory. If export controls extend to maintenance, spare parts, or software updates for already-installed tools, yield rates and uptime at Chinese fabs could deteriorate, undermining the growth path Goldman envisions. Conversely, any relaxation of controls, whether through policy shifts or enforcement gaps, would accelerate the timeline and potentially push growth rates even higher.

What Full Independence Really Means

The Goldman analysis highlights an uncomfortable truth for both Beijing and Washington: surge in production volume does not equal technological sovereignty. China's foundries can churn out unprecedented quantities of 7-nanometre wafers using adapted processes and legacy tools, but the performance, power efficiency, and cost structure of those chips will lag products made with unrestricted access to the latest lithography, metrology, and materials science.

This gap matters less for certain applications and critically for others. Surveillance systems, industrial controllers, and consumer IoT devices tolerate the trade-offs. AI training clusters, high-frequency trading infrastructure, and next-generation mobile base stations do not. The result is a tiered semiconductor landscape: China achieves self-sufficiency in volume and in categories where performance thresholds are forgiving, while remaining dependent on imports or smuggling for the frontier.

For the broader industry, the 2035 projection signals a future of persistent bifurcation. Global supply chains will increasingly route around Chinese capacity for leading-edge products, while Chinese capacity absorbs domestic demand and competes aggressively in mature and mid-range nodes. Foundries in Taiwan, South Korea, and eventually the United States will focus on the high end, protected by equipment access and process know-how, but facing margin pressure as Chinese competitors flood the volume segments.

Lithography will remain the dividing line. Until domestic EUV systems reach commercial viability or geopolitical winds shift enough to reopen equipment access, China's semiconductor industry will grow in scale and sophistication but remain one generation behind, perpetually catching up but never quite closing the gap. The 46 per cent growth rate is impressive, even transformative, but it unfolds within boundaries set by machines that remain out of reach.