International Chip Giants Bet on China! GlobalFoundries Partners with Guangzhou Zengxin, Mature Process Becomes the Key to Breakthrough

“Previously, international giants viewed the Chinese market as a place to ‘sell chips for quick profits’, but now they are actively coming to ‘set up factories for localization’—this shift is happening faster than the iteration of chip processes!” Recently, GlobalFoundries announced its partnership with Guangzhou Zengxin at a technology summit in Shanghai, launching a “China production, China delivery” strategy, which has triggered a chain reaction in the semiconductor circle. As an engineer who deals with wafer foundries daily, I realized after reviewing the details of the cooperation and market data: this is not just a simple business collaboration, but a ‘survival choice’ for international giants amid supply chain turmoil, and the mature process is their ‘golden key’ to unlock the Chinese market.

1. Understanding First: GlobalFoundries’ ‘Chinese Calculation’ is Not Following the Trend, but a Necessity

Many people think GlobalFoundries’ bet on China is ‘following the trend’, but behind it lies precise market calculations—this ‘cake’ of China has long transformed from an ‘optional market’ to a ‘must-fight territory’, especially in the mature process field.

First, let’s look at a set of key data: China is the world’s largest producer of new energy vehicles, and by 2025, the demand for automotive electronic chips will account for over 40% of the global total; meanwhile, 25% of GlobalFoundries’ revenue comes from automotive electronics, with a growth rate exceeding 12% in the first half of the year, making it the most powerful ‘growth engine’ among all its businesses. This is akin to two advantageous companies ‘joining forces’: China has a massive demand for automotive-grade chips, GlobalFoundries has 40nm mature process technology, and Guangzhou Zengxin has an existing 12-inch production line—together, they perfectly fill the gap in ‘local supply of automotive-grade chips in China’.

More critically, there is a demand for supply chain security. Previously, GlobalFoundries’ chips had to be shipped from factories in Singapore and Germany to China, which not only had a delivery cycle of up to 8 weeks but also faced concerns about logistics disruptions due to geopolitical issues. Now, with the cooperation with Zengxin, chips are produced in Guangzhou and delivered locally, shortening the cycle to just 3 weeks and reducing costs by 15%. As Hu Weiduo, President of GlobalFoundries China, said: “Choosing Zengxin means we value its ‘production capacity and local implementation’, which can help us quickly respond to the ’emergency needs’ of Chinese customers.”

A colleague once calculated for me: in 2024, China’s dependence on imported automotive-grade chips will still reach 60%, with a shortfall of 12 million units per month for 40nm CMOS image sensors (used in automotive cameras). GlobalFoundries’ move is equivalent to preemptively seizing the ‘essential track of automotive electronics in China’, understanding better than competitors focused on high-end processes what the ‘Chinese market needs’.

2. Mature Process: The ‘Breakthrough Weapon’ for International Giants and the ‘Pain Point’ of the Chinese Market

GlobalFoundries’ choice to enter the Chinese market with a 40nm mature process may seem ‘conservative’, but in reality, it is a ‘precise strike’ that understands the industry’s pain points—after all, for the Chinese market, the importance of mature processes is no less than that of advanced processes, and perhaps even more so.

First, mature processes are the ‘industrial staple’, indispensable. Mature processes of 28nm and above cover 90% of the demand in automotive electronics, IoT, industrial control, and other fields. For example, automotive MCU chips and power semiconductors, as well as sensors for smart homes, mostly use 40nm and 55nm processes. These chips do not pursue ‘extreme performance’ like AI chips but rather ‘stability, reliability, and timely supply’. As the global manufacturing center, China has an annual demand for over 50 billion mature process chips, with an import dependence exceeding 50%, which is precisely GlobalFoundries’ opportunity.

Second, mature processes are less affected by geopolitical factors, providing stability. Advanced processes (below 7nm) face many restrictions due to equipment and technology blockades, making international giants hesitant to expand production in China; however, mature processes can be produced using DUV lithography machines, which have a relatively low technical threshold and a more stable supply chain. GlobalFoundries’ 40nm process technology has already gone through 5 generations of iteration, with a yield rate stable at over 97%, perfectly meeting Chinese customers’ demands for ‘high yield and stable supply’.

I once spoke with a procurement manager from a domestic automotive company, and their biggest headache was the ‘supply interruption of automotive-grade chips’: in 2024, a certain international giant’s 55nm MCU chip delayed delivery, causing the automotive company to halt production for 3 days, resulting in losses exceeding 200 million yuan. Now, with GlobalFoundries achieving ‘China production, China delivery’, it is like giving automotive companies a ‘peace of mind’—this is also why, as soon as the cooperation news was released, many domestic automotive companies proactively approached to discuss orders.

3. From ‘Selling Chips’ to ‘Building Ecosystems’: The ‘Localization Advancement’ of International Giants

GlobalFoundries’ actions are not an isolated case but a reflection of the collective shift of international chip giants—from ‘simply selling chips’ to ‘deep localization’, even ‘co-building ecosystems’, which reflects a complete change in their understanding of the Chinese market.

Previously, international giants viewed the Chinese market with the attitude of ‘I have technology, you just buy it’, such as shipping chips produced overseas to China for sale, rarely considering local needs; now they are beginning to ‘adapt to local customs’: GlobalFoundries is not only cooperating with Zengxin for production but also setting up new offices in Beijing and Guangzhou to form local teams; another major player, Infineon, has even established a testing center for automotive-grade chips in Suzhou, specifically for reliability verification in the high-temperature and high-humidity environments of Chinese new energy vehicles.

The core of this transformation is shifting from ‘earning one-time sales money’ to ‘earning long-term ecosystem money’. Taking the cooperation between GlobalFoundries and Zengxin as an example, both parties are not only producing automotive-grade CMOS chips but also plan to jointly develop a 55nm low-power process for the Chinese IoT market—this means GlobalFoundries is no longer just a ‘pure technology licensor’ but a ‘deeply involved partner in the Chinese industrial chain’.

Industry analysts have joked: “Previously, international giants viewed the Chinese market as ‘fat meat’, but now they treat it as a ‘base’—after all, in the current global supply chain turmoil, the stable demand and complete industrial chain of the Chinese market are irreplaceable by other regions.”

4. Engineer’s Perspective: The ‘Technical Game’ Behind the Cooperation is Not One Side Dependent on the Other, but Mutual Benefit

Many people worry that ‘will the technology licensing from international giants lead to a bottleneck’, but from a technical detail perspective, this cooperation resembles ‘mutual benefit’ rather than ‘one-way output’.

For GlobalFoundries, by licensing the 40nm technology, it can quickly activate Zengxin’s idle production capacity (Zengxin’s 12-inch production line had previously only been utilized at 60%), while also leveraging Zengxin’s local channels to reach more Chinese customers—it’s worth noting that Zengxin has already collaborated with BYD and GAC in the automotive sensor field, allowing GlobalFoundries to quickly penetrate these core customers through this cooperation.

For Zengxin, obtaining GlobalFoundries’ 40nm technology license allows it to skip the ‘3-5 years of independent R&D’ phase and directly produce high-value-added automotive-grade chips. More critically, GlobalFoundries will also provide yield optimization support—I learned that GlobalFoundries has already sent a technical team of 10 people to Zengxin to help improve the yield of 40nm CMOS chips from 85% to 95%, which is ‘practical experience that money can’t buy’ for Zengxin.

From a technical perspective, while the 40nm process may not be the most advanced, it is the ‘most cost-effective training ground’. Through this cooperation, Zengxin can master the entire process of design, production, and testing of automotive-grade chips; meanwhile, GlobalFoundries can optimize the adaptability of its technology in local production lines through collaboration with Chinese factories—this kind of ‘technical interchange’ is more stable than a simple buying and selling relationship.

5. Comparing with SMIC: Is the Entry of International Giants Competition or Complementarity?

Seeing GlobalFoundries ramping up its investment in China’s mature process, many people will ask: “Will this impact SMIC?” In fact, from the market structure perspective, the two are more ‘complementary’ than ‘direct competitors’.

SMIC’s advantage lies in its ‘full process coverage’, with layouts from 90nm to 14nm, especially the 28nm HKC+ process, which has over 30% market share in the domestic automotive-grade market; while GlobalFoundries’ advantage lies in its ‘automotive-grade specialized technology’, with its 40nm CMOS image sensor process having passed AEC-Q100 automotive-grade certification, offering unique advantages in low power consumption and high reliability, perfectly filling the gap in the domestic automotive-grade image sensor foundry field.

To put it simply: SMIC is like a ‘comprehensive supermarket’, offering everything to meet the basic needs of most customers; GlobalFoundries is like a ’boutique store’, focusing on high-value areas like automotive-grade chips, providing more specialized products. The combined efforts of both can actually reduce China’s dependence on imported mature process chips—by 2024, China’s imports of automotive-grade chips will still reach $80 billion, enough for both companies to grow together.

More importantly, the entry of international giants can drive the development of the local supply chain. The cooperation between GlobalFoundries and Zengxin will prioritize the procurement of domestic materials such as photoresists and specialty gases, providing a ‘rare validation opportunity’ for domestic semiconductor material manufacturers. For instance, a domestic photoresist company has already passed GlobalFoundries’ certification and started supplying materials for the cooperative production line—this ‘point-to-surface’ driving effect is more valuable than simple capacity expansion.

6. Future Battlefields: Localization is Not the ‘End Point’, but a ‘New Starting Point’

GlobalFoundries’ localization strategy is just the ‘first step’ for international giants in China. According to its plan, it will further introduce more mature process technologies such as 55nm and 65nm into China, and even plans to jointly develop SiC (silicon carbide) power devices for new energy vehicles with Zengxin—this means that the localization of international giants in China will extend from ‘mature process foundry’ to ‘high-end device R&D’.

For the Chinese semiconductor industry, this presents both opportunities and challenges: the opportunity lies in rapidly enhancing the technical level and supply chain capabilities of local enterprises through cooperation; the challenge is how to master core technologies in collaboration and avoid the predicament of ‘only doing foundry work without mastering the core’. For instance, Zengxin, while cooperating with GlobalFoundries, has also formed its own R&D team, planning to achieve independent optimization of the 40nm technology within 3 years—this ‘leveraging strength’ approach is the way for local enterprises to break through.

One thing is certain: as more international giants join the ‘localization camp’ in China, competition in the Chinese semiconductor market will become more intense, but it will also become more mature. The ‘hundred flowers blooming’ in the mature process field will ultimately benefit not only the enterprises but also the entire Chinese manufacturing industry—after all, only a stable and sufficient supply of chips can support the continuous development of industries such as new energy vehicles, smart homes, and industrial internet.

Conclusion: The Essence of Supply Chain Security is ‘You Have Me, I Have You’

The cooperation between GlobalFoundries and Guangzhou Zengxin may seem like an ordinary business marriage, but it is actually a microcosm of the global semiconductor supply chain restructuring. In the current geopolitical turmoil, the consensus has become to ‘put eggs in multiple baskets’, and the Chinese market, with its huge demand and complete industrial chain, has become one of the ‘safest baskets’ in the eyes of international giants.

For international giants, ‘China production, China delivery’ is not a ‘stopgap measure’, but a ‘long-term strategy’; for China, attracting international technology and investment through cooperation while enhancing local industrial capabilities is also the ‘inevitable path’ to achieving semiconductor self-sufficiency. This ‘mutual benefit’ model is more in line with the development laws of the global semiconductor industry than a ‘zero-sum game’.

In the future, we may see more international giants collaborating with Chinese enterprises to jointly advance in fields such as mature processes, automotive electronics, and IoT. The rise of the Chinese semiconductor industry will not be ‘self-reliant behind closed doors’, but will continuously enhance core competitiveness through open cooperation—after all, the highest realm of supply chain security is not ‘complete independence’, but ‘you have me, I have you’, where no one can do without the other.

What do you think about the impact of international giants’ localization strategies on the Chinese semiconductor industry? How should SMIC respond to the new competitive landscape? Feel free to share your thoughts in the comments section!