Opinion: capacity, geography, and the rise of advanced nodes in the semiconductor Industry

Historically growing at 6.4% CAGR, the semiconductor device industry is now poised for a significantly higher growth rate in the coming five years, driven primarily by demand from servers, automotive electronics, and industrial applications. Within this ecosystem, semiconductor manufacturing remains a central pillar of the value chain. Foundry revenues, combining IDM production and open foundry services, account for roughly 50% of total semiconductor device revenues, reflecting the strategic importance of fabrication capacity.

This 3^rd opinion article is authored by Claire Troadec, Director of Global Semiconductor Activities at Yole Group, who brings more than two decades of experience in semiconductor technology, market analysis, and strategic business development. Leveraging close interactions with leading industry players and deep technical expertise, Claire provides a comprehensive perspective on the evolving dynamics of semiconductor manufacturing.

This Opinion analysis builds on Yole Group’s extensive portfolio of market intelligence, including reports such as Status of the Semiconductor Foundry Industry 2026, Status of the Semiconductor Devices Industry 2026, Greater China Semiconductor Device Industry 2026 etc. Discover here the full Global Semiconductor collection.

Advanced nodes below 10nm are expected to generate more revenue than the next three legacy-node categories combined!

Semiconductor manufacturing capacity remains heavily concentrated in Asia, particularly in Taiwan, South Korea, and mainland China. Taiwan leads advanced-node production through TSMC, while South Korea, especially Samsung, plays a major role in both logic and memory technologies.

Meanwhile, China is significantly expanding its manufacturing capacity, particularly in mature nodes. Continued investment could allow the country to reach 30% of global foundry capacity by 2031, reinforcing its role as a major manufacturing hub even though its share of semiconductor device sales remains relatively modest.

The global semiconductor landscape is therefore increasingly defined by capacity allocation across both regions and technology nodes.

Although advanced technologies attract the most attention in the industry, the majority of global manufacturing capacity remains concentrated in legacy nodes, particularly 90–45nm and 40–22nm processes. These nodes support a wide range of applications, including automotive electronics, industrial systems, and consumer devices, and continue to generate a large share of semiconductor revenues.

From a regional perspective, China, Taiwan, and parts of Southeast Asia host a large share of this mature-node capacity, reflecting years of incremental capacity expansion and lower manufacturing costs.

However, the situation differs significantly for the most advanced nodes (<10nm), where capacity is far more geographically concentrated.

Manufacturing capacity for advanced nodes below 10nm is largely concentrated in Taiwan and South Korea, dominated by TSMC and Samsung. These two players account for the vast majority of global advanced-node wafer capacity.

Other regions are attempting to close the gap. In Japan, companies such as Kioxia and Rapidus are investing in advanced semiconductor production, while the United States, led by Intel and Micron, is seeking to rebuild domestic advanced-node capabilities.

However, the number of companies capable of competing in this space has narrowed dramatically. Today, only three manufacturers, TSMC, Samsung, and Intel, remain active competitors in advanced-node logic manufacturing. All three are racing to introduce 2nm-class technologies, with high-volume manufacturing expected around 2026.

Complicating comparisons between manufacturers is the increasing ambiguity of node naming conventions. Process nodes are no longer tied to a specific physical dimension but instead reflect company-specific performance benchmarks and marketing strategies. For instance, Intel’s 18Å node offers performance similar to the 3nm nodes announced by TSMC, while Samsung’s naming scheme falls between the two approaches.

China’s semiconductor strategy reflects a capacity-driven approach. Although the country accounts for only 6% of global semiconductor device sales, it already represents roughly 21% of global foundry capacity, mainly concentrated in mature nodes.

This manufacturing base has enabled Chinese companies to be cost leaders in several device categories. Early progress was seen in optoelectronics and sensors, while power and analog semiconductors are now major focus areas.

Memory technologies are expected to follow, fueled by sustained investment and unprecedented AI demand. Supply constraints could open new opportunities for Chinese manufacturers such as YMTC and CXMT in automotive and consumer segments, while established memory leaders prioritize higher-end markets like servers and data centers.

Demand for AI has significantly increased interest in advanced semiconductor technologies. However, this demand has created a growing economic divide between node generations.

Advanced nodes (<10nm) are experiencing rapid revenue growth and strong pricing power, driven by high demand and limited capacity. In contrast, legacy nodes (>10nm) have faced significant price erosion, despite representing the majority of global manufacturing capacity.

Looking toward 2031, the semiconductor manufacturing landscape will be increasingly shaped by the distribution of node-specific capacity across regions.

Advanced nodes below 10nm are expected to generate more revenue than the next three legacy-node categories combined, even though mature nodes will continue to dominate global capacity.

In this scenario, the United States, Taiwan, Korea, and China will dominate the semiconductor foundry landscape and push other regions into irrelevance.

As described in Yole Group’s Global Semiconductor collection, the semiconductor manufacturing landscape is entering a decisive phase. While legacy nodes continue to anchor global capacity, advanced nodes are rapidly capturing value, driven by AI demand and limited supply. This growing divide is reshaping competitive dynamics, reinforcing regional concentration, and accelerating strategic investments worldwide.

As the industry moves toward 2031, understanding the interplay between capacity, geography, and technology nodes will be critical for anticipating market shifts and identifying new opportunities.

💬 Connect with Yole Group’s analysts to explore deeper insights, access detailed data, and discuss how these trends will impact your business and strategic roadmap.

Claire Troadec is Director of Global Semiconductor activity at Yole Group.

Based on her valuable experience in the semiconductor industry, Claire is managing the expansion of the technical and market expertise of the global semiconductor activities. Daily interactions with leading companies allow these analysts to collect a large amount of data and cross their vision of market segments’ evolution and technology breakthroughs.

In addition, Claire’s mission is focused on the management of business relationships with leading companies of this sector and the development of market research and strategy consulting activities inside the Yole Group.

Claire Troadec holds a Master’s degree in Applied Physics specializing in Microelectronics from INSA (Rennes, France). She then joined NXP Semiconductors and worked for 7 years as a complementary metal-on-silicon oxide semiconductor (CMOS) process integration engineer at the IMEC R&D facility. During this time, she oversaw the isolation and performance boosting of CMOS technology node devices from 90 nm down to 45 nm. She has authored or co-authored seven US patents and nine international publications in the semiconductor field and managed her own distribution company before joining Yole Group in 2013.

As well as monitors:

Source: www.yolegroup.com