TSMC is the undisputed leader in the global foundry industry. Its industry position can be summarized as follows: it is not only the pioneer of the industry but also the absolute leader. With continuous massive investments in advanced processes and technological leadership, it has built extremely high technical and ecological barriers, maintaining over half of the global market share for a long time. This has led top global tech companies like Apple, NVIDIA, and Qualcomm to heavily rely on its manufacturing capabilities. This comprehensive advantage in technology, scale, and customer ecosystem has granted it near-monopoly pricing power and influence over the global high-end chip supply chain, making it the cornerstone of the entire digital world. The Chinese foundry industry is striving to catch up; what investment opportunities exist in the foundry industry under the trend of domestic substitution?

1. Industry Landscape

First, let’s clarify the foundry model—specifically the TSMC model, which refers to foundries that do not design or sell their own branded chips, but instead provide chip manufacturing services exclusively for fabless companies. Today, we will only discuss companies primarily engaged in foundry services. The foundry industry in mainland China presents a pattern of “one strong leader and many strong competitors.”

One Strong Leader: SMIC is the absolute leader and the only pure foundry in mainland China capable of mass-producing 14nm and more advanced processes. It also has significant capacity and scale advantages in mature processes, making it the closest competitor to the “TSMC model.” Technological Gap: The most advanced node in mass production (14nm) lags 2-3 generations behind TSMC (which has 3nm in mass production and 2nm in research).Capacity Scale: Although it is the largest in China, its global market share (about 5%) is not on the same level as TSMC (about 60%).Equipment and Material Limitations: Due to international sanctions, acquiring the most advanced semiconductor manufacturing equipment (such as EUV lithography machines) is extremely difficult, severely restricting its speed of advancement to more advanced processes.

Many Strong Competitors: Here are a few typical examples.

Huahong Semiconductor: Focused on specialty process platforms such as power semiconductors, embedded memory, and analog chips, it is highly competitive in niche areas. It operates in a “TSMC-like model” but with deep differentiation.Differentiation Strategy: Huahong has not engaged in a head-to-head competition in advanced logic processes like SMIC and TSMC but has chosen specialty processes as its main track, achieving world-leading levels.Differentiation Strategy: Huahong has not engaged in a head-to-head competition in advanced logic processes like SMIC and TSMC but has chosen specialty processes as its main track, achieving world-leading levels.

Hefei Integrated Circuit: Mainly focuses on foundry services for display driver chips, CIS image sensors, etc., and has grown rapidly, ranking among the top ten globally.Model: Focused Pure Foundry in Niche Markets. It started from the single field of display driver chips and quickly became a global market leader, then gradually expanded into CIS, MCU, and other adjacent fields. This is a “point-to-surface” growth path, different from TSMC’s “full bloom” giant model.Model: Pure foundry focused on analog/power/automotive applications. They focus on mature processes and specialty processes with huge market demand, especially analog chips and power devices, filling the industrial gaps in South and East China. Their model is more akin to “specialty process foundry,” serving a large market that does not require the most advanced logic processes.

2. Investment Logic Analysis

SMIC, the leading foundry in mainland China, has a revenue scale of about one-tenth that of TSMC, and the total output value of the entire foundry industry in mainland China is far less than that of TSMC. The profit margin gap (approximately 39% for TSMC vs. about 10% for SMIC/Huahong) is even more significant than the revenue scale gap. This directly reflects the huge differences in technological barriers, product added value, customer structure, and global pricing power.

• SMIC: Like a “national team main attacker” aiming to challenge international top levels, its task is to break through on the cutting-edge track and solve the “existence” problem.
• Huahong: Like a “special forces” competing for gold and silver in advantageous projects, it does not pursue being comprehensive but aims to be world-class in specific projects.
• Hefei Integrated Circuit: Like a “dark horse” starting from a single project and quickly entering the first tier, its strategy is to focus on one point, deepen it, and then expand.

Currently, with the wave of artificial intelligence and domestic substitution becoming self-controllable, driven by emerging demands in AI computing power and smart vehicles, the global semiconductor cycle has entered a recovery phase. The technology industry is still in the early stages of industrial explosion, and these companies all possess investment value.

SMIC Investment Logic: As the only hope for mainland China to break through the “bottleneck” in high-end chips, this scarcity constitutes its deepest moat. More critically, under the ongoing external restrictions, domestic substitution has shifted from an “optional” to a “mandatory” choice. Orders from domestic high-end chip design companies (such as Huawei HiSilicon) will continue to tilt towards SMIC, providing strong internal momentum for its advanced process capacity utilization and technological iteration. Despite facing challenges in acquiring the most advanced EUV equipment, the company is effectively meeting domestic market demands for high-performance computing through innovative processes and a focus on special architectures of “quasi-advanced processes” (such as N+1, N+2). Looking ahead, SMIC is not only the ballast for mature processes but also the only outlet for China to advance in semiconductor self-reliance, and its value will be reassessed with each substantial breakthrough in technology nodes. Investing in SMIC is investing in the future ceiling of China’s semiconductor industry.

Huahong Investment Logic: The group relies on external advanced assets (such as the R&D platform that has successfully run 7nm, Huahong Micro’s fifth factory, etc.) to possess strong potential for transitioning to advanced processes. With the implementation of strategic synergies such as “Huawei’s acquisition of the fifth factory” and the clear expectation of future advanced process asset injection into the listed company, the company will break through the ceiling of mature processes and rise to become a scarce advanced logic process platform in China. At the same time, the company’s mature process base is solid, and the rapid expansion of the 12-inch wafer factory will drive a dual enhancement of performance and valuation, making it a core target with both certainty and high growth in the wave of semiconductor domestic substitution.

Hefei Integrated Circuit Investment Logic: Lies in its status as the most notable dark horse in the domestic wafer manufacturing field, successfully transitioning from “the leading foundry for display driver chips” to “a multi-platform semiconductor manufacturing platform.” The company leverages its deep manufacturing experience and cost control capabilities accumulated in the display driver field to strongly enter the huge market of CIS (image sensors) and actively layout future 3D DRAM and other cutting-edge storage technologies, demonstrating exceptional strategic foresight. Particularly crucial is its strategic cooperation with Changxin Storage, which not only locks in the manufacturing demand for the localization of future advanced storage technologies but also means the company has embedded itself in the core ecosystem of the national storage strategy, likely gaining an advantage in the next generation of storage technology transformation. This unique model of “high growth + platform expansion + binding national strategy” makes Hefei Integrated Circuit a high-quality target that transcends traditional cycles, enjoying the dual drivers of domestic substitution and technological upgrades, with significant performance elasticity and valuation reshaping potential.

3. Conclusion

All three companies possess clear investment value, but their driving logic varies, forming a gradient layout from strategic cornerstone to high elasticity.

SMIC, as the only domestic leader achieving large-scale mass production of advanced processes, still has a huge gap in revenue and market value compared to TSMC. However, it benefits from the accelerated promotion of domestic substitution and continuous breakthroughs in technology nodes, possessing the “ballast stone” value of scarcity and strategic position, with clear medium to long-term growth potential.Huahong currently has a valuation significantly lower than SMIC, with the core focus being on the potential asset injection of advanced processes that have been successfully run at the group level, such as 7nm. If advanced process capabilities are integrated into the listed company, it will open up a dual enhancement channel for performance and valuation, presenting considerable recovery space.Hefei Integrated Circuit exhibits the strongest growth potential, rapidly rising not only in the display driver chip field but also strongly entering CIS image sensors and proactively laying out in cutting-edge directions like 3D DRAM. Its high gross margin and rapid expansion of production capacity make it a “dark horse” target in the domestic foundry sector, combining high elasticity and high foresight. The three represent certainty, expectation difference, and growth potential, collectively forming an investment portfolio for the localization of semiconductor manufacturing.