SMIC: China’s Foundry Between Rapid Expansion and Strategic Dependence

Administrateur
2 days ago
3 min read

Between 2019 and 2025 Semiconductor Manufacturing International Corporation (SMIC) rose to third place among global foundries. Its trajectory embodies China’s quest for technological self-sufficiency: more fabs, surging revenue, but also geopolitical pressure and fragile margins. This synthesis reviews SMIC’s activities, markets, ownership, financial performance and investment strategy to position the company in the global chip race.

Activities and Technological Focus

A pure-play foundry, SMIC manufactures chips for third-party designers on 200 mm and 300 mm wafers. Its core output still comes from mature nodes (90 nm to 28 nm), augmented since 2019 by 14 nm FinFET and, more recently, a home-grown 7 nm process developed without EUV equipment. Joint ventures add specialist flows—high-voltage, RF and image sensors—targeting automotive, IoT and AR/VR. Fabs in Shanghai, Beijing, Shenzhen and Tianjin now run at nearly 950 000 eight-inch-equivalent wafers per month, a rise of over 35 % in five years.

Key Markets and Geographic Dependence

Stricter U.S. sanctions pushed SMIC inward: domestic customers accounted for roughly 85 % of revenue in 2024, up from about 70 % in 2019. Smartphones, once dominant, now bring less than a quarter of sales. Connected consumer devices—TVs, wearables, smart appliances—supply close to 40 %. PCs/tablets, industrial IoT and, to a smaller extent, automotive fill the rest. Heavy reliance on China guarantees volume but ties fortunes to local demand cycles.

Ownership Structure and Implications

Listed in Hong Kong and Shanghai, SMIC is tightly state-linked: Datang Telecom, China’s “Big Fund” and the SASAC together own more than a quarter of the capital. Public backing provides rich subsidies, soft loans and provincial partnerships, crucial for capital-intensive expansion. It also attracts U.S. export restrictions. While the model secures funding and political support, it limits strategic autonomy.

Competitive Positioning

TSMC retains clear leadership in cutting-edge nodes (5 nm, 3 nm) with net margins above 40 %. SMIC, confined to ≥7 nm, positions itself as China’s import-substitution foundry rather than a direct rival at the high end.
Samsung Foundry combines external clients and in-house chipmaking, owns EUV tools and invests overseas. SMIC trails technologically and geographically, though it matches Samsung on mature-node volume inside China.
Intel Foundry Services is opening advanced U.S. capacity to external customers; overlap with SMIC is indirect, each serving its geopolitical bloc.
GlobalFoundries focuses on mature nodes in the U.S., Singapore and Europe. Both GF and SMIC share a mid-range technology profile, but GF operates free of embargoes while SMIC enjoys a vast captive market.

SMIC is therefore a regional champion in intermediate processes, striving to narrow the gap on 5 nm through in-house innovation.

Financial Performance and Profitability

Revenue reached 8 billion USD in 2024 (+27 %), yet net profit fell 45 % to 493 million USD, slicing the net margin below 6 %. Three factors weigh on results:

Lower-value technology mix—nodes ≥28 nm command lower selling prices.
Rising domestic competition—Hua Hong, Nexchip and others press mature-node pricing.
High fixed costs—R&D at 12-16 % of sales plus heavy fab depreciation.

Even so, EBITDA continues to climb and cash flow stays healthy, supported by record utilisation and government support.

Investment Strategy and Capacity Buildup

Since 2019 SMIC has launched megaprojects in Shenzhen (40 000 wpm), Beijing Jingcheng (up to 100 000 wpm) and expanded Shanghai lines, sustaining annual capex above 7 billion USD. Goals:

Satisfy domestic demand for mature nodes vital to autos, 5 G and consumer electronics.
Advance non-EUV processes—7 nm DUV multipatterning is already shipping for Huawei; 5 nm is targeted for 2025.
Strengthen national sovereignty—each fab pairs with a provincial government, mirroring Beijing’s plan to cover 75 % of local chip needs over time.

The spending suppresses short-term margins but readies China’s industrial base for the late 2020s.

Lessons for China’s Chip Industry

SMIC’s path shows:

Spectacular catch-up at intermediate nodes—China now controls nearly one-third of global 28-65 nm capacity.
Persistent gap at the leading edge—without EUV, the 3-5 nm divide remains wide.
State-market hybrid model—domestic demand fuels volume, the state funds capacity, but capital efficiency is under scrutiny.
Partial geopolitical resilience—sanctions spur domestic innovation, yet supply chains (EDA, materials, critical tools) remain vulnerable.
Emerging multibloc competition—U.S., EU, Korea and Taiwan counter with their own subsidies, setting the stage for redundant global capacity.

Conclusion

From 2019 to 2025 SMIC evolved from ambitious national player to cornerstone of China’s microelectronics drive. Backed by a vast home market and state strategy, it now rivals GlobalFoundries in volume and edges closer to Samsung in the mature-node arena. Yet the lack of EUV tools, thin margins and reliance on subsidies expose the model’s fragility. Over the next five years SMIC is set to cement dominance in mature nodes and extend FinFET offerings; its success at an indigenous 5 nm will gauge China’s progress toward closing the technology gap. Beyond its own future, SMIC remains the best barometer of China’s ability to convert capital and talent into lasting semiconductor leadership.