The Road to Domestic Substitution of Automotive-grade Chips

Lead: Reclaiming the Lifeline of Chips, Self-Redemption of China's Automotive Industry

On August 17, 2021, Bosch China Executive Vice President Xu Daquan still remembers this sleepless night.

That day, he posted a message on WeChat Moments: A semiconductor chip supplier's Muar plant in Malaysia closed several production lines, directly affecting Bosch ESP, IPB, VCU, TCU and other types of chips. He also posted a thought-provoking image: The building, 6 stories; Jump, or don't? Take the boss, or not?

Soon, the comment section of this Moments post gathered the mental state of auto industry professionals at that time: Better to be dead than alive, too sad to cry, September even worse... The culprit that made the entire industry overwhelmed was chips.

To ensure normal production, finding chips became the top priority for car companies.

That year, almost all car company CEOs squatted in Shanghai (Bosch China headquarters) to get chips. It was too crowded downstairs at Bosch, so procurement staff went to squat in front of chip foundries and manufacturers.

But the effect of this method was average, and the supply disruption crisis still affected the production frontline.

NIO became the first Chinese new energy car company forced to stop production due to chip shortage. The JAC-NIO Hefei factory stopped production for 5 working days from March 29, 2021. FAW-Volkswagen had production affected for 5 consecutive months, with single-month production capacity affected by over 50%.

Setting up checkpoints became the collective memory of complete vehicle manufacturers that year;

Anxiety was the general mood of practitioners at that time;

The shortage of chips, especially the hidden pain of reliance on imported automotive-grade chips, was torn open again by the wave of supply cuts and production reductions.

From then on, automotive-grade chips embarked on a road of self-redemption for domestic substitution.

Chip Crisis

For car companies, the worst scenario of chip shortage is production stoppage. Stoppage means the entire industrial chain is interrupted, bleeding money every day. To avoid stoppage, a roaring panic buying wave pushed automotive-grade chip prices up to hundreds of times. A certain chip model normally costs around 1500 yuan, with black market prices rising as high as 6 times, and this was considered a reasonable price in the black market transactions at that time.

And the cost increase brought by chip price hikes put car companies in a dilemma: If costs are not passed on to the sales end, car companies with declining profit margins, or even losing money on every sale, will accelerate bleeding. If passed on to consumers, they will lose price advantages in the already intensified price war market.

From later market feedback, it is clear that most car companies chose to pass costs to consumers at the end of the industrial chain.

Tesla, known for its image as a "price butcher", raised prices for multiple models one after another. Domestically, terminal discounts for best-selling models like VW Magotan and Honda Accord generally reversed, waiting one to two months for delivery, some even waiting half a year.

A more hidden cost is the reduction of specifications. Some new force car companies launched a millimeter-wave radar solution of "delivery first, installation later", some car companies dropped in-car navigation on certain models.

It can be said that every penny of cost from chip supply cuts will eventually land in consumers' wallets and experiences.

This chaos soon alarmed regulatory bodies. In August 2021, the State Administration for Market Regulation launched an investigation into automotive chip distributors suspected of price gouging. However, administrative means can only patch up after the sheep are lost.

When there are too many monks and not enough porridge for chips, supply is less than demand, prices must rise. A more fundamental question is, why is there no domestic supply chain to buffer the automotive chips that are vital to car companies?

Making Chips Is Not Hard

In 2021, China could make automotive-grade chips.

Automotive-grade chips mostly use mature processes, avoiding the pain points of being "strangled" by advanced processes like 3nm, 5nm used in mobile SoC chips, falling exactly within the range where domestic manufacturing capabilities exist. The mature process of domestic chips can fully meet car company needs and supply for a long time. Clearly a strong suit, why couldn't it relieve the automotive industry's thirst?

The key lies in these three words "automotive-grade".

Who has used your chips? Are they installed in vehicles? What if there are quality issues in mass production? Corechip Technology Founder and Chairman Qiu Yuqing called these three questions the "Three Soul Questions" from customers to domestic chip start-ups.

Around 2021, these three questions were enough to keep the vast majority of domestic chip companies outside the door.

It is not that car companies don't want to use domestic products, nor is it that domestic products cannot be used or fail to meet requirements. It is mainly because the testing cost is there. Once problems arise, it affects an entire production line or a batch of models. Domestic chips need to prove their quality and safety can withstand tests during a long verification cycle.

At that time, verifying whether a chip meets the strict requirements for reliability and functional safety, certification systems and standard organizations were initiated and led by international giants. So, domestic car companies naturally tended to choose foreign manufacturers for support.

For example, AEC (Automotive Electronics Council) was established by Chrysler, Ford, and GM. The AEC-Q series standards, after 30 years of development, have become the recognized general test standard for automotive components, containing 41 tests in 7 major categories, with at least 28 experiments required to be completed.

There is also IATF 16949 quality management system, a quality management standard for the automotive industry based on ISO 9001, emphasizing zero defects and full-process control. It requires ensuring the stability of the production process. Relying on this system, it ensures that every chip participating in later manufacturing reaches the same level as the initial sample certification.

Systems such as airbags, Anti-lock Braking Systems (ABS), and Electric Power Steering require the highest level of ASIL-D. ASIL-D level systems must have a fault detection coverage rate of over 99%.

Car companies dare not use domestic chips because there is no large-scale vehicle installation data, and domestic chips cannot produce large-scale vehicle installation data because no car companies use them. This forms a chicken-and-egg deadlock, firmly limiting the path for domestic chips to enter vehicles.

Difficulty in Vehicle Installation

Chips are not the problem at all. The real problem is whether car companies dare to use them. Domestic chips lack only an opportunity to enter vehicles. Being in a desperate situation is often the time to break the deadlock.

The outbreak of the overseas chip supply crisis gave domestic chips a critical window period for vehicle installation. Food and ammunition cut off, now they dare or have to dare.

A small car fired the starting gun for domestic substitution.

Wuling Hongguang MINI EV, claimed to be the people's commuter car, sold hot in the market, but its ABS system was supplied by two suppliers. Due to chip shortage, the other supplier simply gave up, and the originally promised products could not be delivered. At the critical moment, Lianchuang Automotive Electronics was given a dangerous mission, deciding to use pre-researched replacement chips to guarantee Wuling's production line would not stop.

On August 26, 2021, Wuling and Lianchuang jointly decided to complete all software development, calibration, and production line verification of the replacement solution within one month. At 4:21 AM on September 27, a Hongguang MINI EV equipped with a new ABS product went off the production line.

The successful cooperation between Wuling and Lianchuang Electronics in 2021 was not an isolated case.

As mentioned earlier, automotive installation falls within the manufacturing capabilities of the semiconductor industry.

Within this stage, domestic chips were densely installed on vehicles and took root in many models.

Among them, power semiconductors, as the core of the new energy vehicle three-electric system, are the fastest replacing component.

In 2022, the leader of domestic automotive-grade IGBT modules, StarPower Semiconductor, achieved the first mass installation of an 800V high-voltage SiC main drive platform in China, including models from XPeng, NIO, and BYD. BYD Semiconductor is a benchmark for vertical integration, with power device installations reaching 3.806 million sets in 2025.

Smart driving and cockpit SoC are the computational power highlands of automotive chips. Among cockpit chip manufacturers, Corechip Technology ranks fifth with 3.77% share, first among local manufacturers, Huawei HiSilicon ranks sixth with 3.39%. In the cockpit domain control chip market, the share of domestic manufacturers such as Huawei, SiEngine Technology, and Corechip Technology is also continuously rising.

MCU and security chips are recognized as the hardest bones to gnaw in automotive-grade chips. GigaDevice's automotive NOR Flash ranks third globally in market share, entering supply chains for Tesla, VW, etc. Guoxin Technology excels in self-controlled PowerPC architecture, and its automotive-grade airbag controller chip also broke Bosch's monopoly.

There are also many hidden champions in niche chip fields, such as Ingenic, whose automotive DRAM ranks among the top in the global market. The company's main customers include Tier 1 suppliers like Bosch and Visteon, as well as complete vehicle manufacturers like Tesla, BYD, and Geely. Novosens is the leader in domestic automotive isolation drivers and signal chain analog chips, filling the gap in domestic high-voltage isolation chips...

Dormestic chips finally began to break the deadlock of vehicle installation. And automotive-grade chips, this mature process market, also became the first battlefield torn open for semiconductor domestic substitution.

Temp Workers Officially Hired

Perhaps some will say, in 2021, due to urgent matters, domestic chips accepted a sudden pie from overseas supply chain fluctuations. Such unexpected events will not exist for long. Once the global supply chain returns to normal, the vehicle installation rate of domestic chips will recede.

But in fact, this is not a temporary official hiring of a spare tire. The road to domestication of automotive-grade chips has just begun. The ability of domestic chip companies to receive orders from car companies does not rely on coincidence. Behind this lies complex causes, which constitute the stable factors for domestic chips entering vehicles.

First is the right time. In the traditional fuel car era, Bosch, Continental, and Denso delivered system solutions, autonomous car companies bought black boxes, with underlying chips encapsulated inside the supply chain. For domestic chips to break into this industrial chain, the threshold was very high. In 2021, coinciding with the changes in automotive electrification and intelligence, the three-electric system became the new value center. The cockpit and vehicle control chips, the "brain" of the car, are core universal links, and also the part where car companies need the most stable supply chain. Chinese car companies had the first opportunity to define chips starting from complete vehicle requirements.

We also need to see the geographic advantage. The domestic semiconductor industry accumulated to a certain stage, forming a basically complete full-link industrial structure. Except for a few advanced chips, it can support the design, tape-out, and mass production of most automotive-grade chips. Meanwhile, compared to overseas supply chains, domestic supply chains have more obvious advantages in response speed, collaborative innovation, and communication efficiency, can do what car companies need to the extreme. It also established better geographic advantages for domestic substitution.

More importantly, it is people, which is the reconstruction of the relationship between car companies and chip supply chains.

In the traditional model, car companies generally do not purchase chips directly. They are provided by Tier 1 suppliers like Bosch and Continental. After Tier 1 receives car company orders, they place orders with chip manufacturers. Chip manufacturers entrust wafer foundries to produce. Part manufacturers get chips to make parts, finally supplied to car factories.

On this industry chain with clear division of labor and extremely long links, chip factories cannot reach complete vehicle factories at all. By the time car companies feel supply disruption risks, industry-wide crises are often already quite serious. So after the chip shortage, some car companies began to bypass Tier 1 and place orders directly with chip design companies. NXP also publicly stated they were seeking to bypass Tier 1s like Bosch and establish supply relationships directly with complete vehicle factories.

And the linked co-creation between vehicle factories and chip factories also paved the way for domestic automotive-grade chips to enter vehicles. Once this co-creation path is started, it will not be easily terminated.

In 2026, self-developed smart driving chips from four top car companies, BYD, Li Auto, NIO, and XPeng, achieved mass vehicle installation in the same period.

Except for self-development, car companies can also invest in chip manufacturers to lock in long-term priority supply rights. The most representative is the deep binding between BYD and CXMT, from capital participation to R&D collaboration, locking in pre-installation demands for high-end brands like Denza and Yangwang in advance.

Some original equipment manufacturers also established joint innovation centers with chip companies, deep binding, and collaborating upstream and downstream to build a technology community, shortening development cycles.

This series of actions not only promotes the intelligent upgrade of the entire automotive industry, but also gradually forms an open and win-win industrial ecosystem. If the entry of domestic chips in 2021 was the official hiring of spare tires under a supply crisis, then coming to today, the domestic substitution of automotive-grade chips already has the superposition of right timing, geographic advantage, and people support.

After the Dawn Appears

Dormestic chips have really been mass-installed, car companies are really using them. The domestication of automotive-grade chips seems to have already shown a dawn.

But the Chinese industry has a familiar script: As long as a breakthrough occurs in one field, the other side of concentrating resources to handle big matters is a crowd pouring in.

From 2024 to 2026, on the track of automotive-grade chips, the temperature gauge of the capital market was quite hot. At the same time, the disadvantages of crowding in and fighting separately also began to appear.

Due to the lack of a unified national chip evaluation system, multiple top car companies built their own research centers to verify chips, and access requirements were "fighting separately". The result was chip manufacturers were exhausted responding to various certifications, Company A ran it once, Company B didn't recognize it, had to re-test, and a complete reliability test might take 6 to 9 months, with high costs.

Price wars also began to surface. Among the approximately 400 MCU manufacturers in China, over 30 focus on automotive-grade, frequent price wars. Many chip factories without automotive background took their industrial-grade chips, did AEC-Q100, then claimed to enter the automotive MCU track, resulting in varying quality of related chips, difficult to distinguish.

In this way, with no unified standard and low-quality supply appearing, the result is diluting the reputation of the entire "domestic" label.

For ordinary consumers, automotive-grade chips concern vehicle performance, safety, and reliability. Once domestic chips leave consumers with a brand perception of low price and low quality, this distrust will transmit to car companies, and the hard-won situation for domestic automotive-grade chips will face obstacles again.

The force to break the problem currently comes from two aspects, one is the national level, building a technical system standard that fosters industry consensus.

On August 17, 2026, five certification recognition industry standards such as "General Evaluation Requirements for Automotive Chip Institution Certification Review" were officially released, adopting a "1+4" architecture: the top-level 1 general evaluation requirement covers the whole picture, and 4 sub-special standards target the four core links of design, certification, testing, and computing power. It realizes full-process coverage from chip R&D to vehicle integration, systematically building an institutional capability evaluation system in the automotive chip field. This is equivalent to having a unified evaluation scale for the whole industry chain, helping to foster industry consensus and avoid resource waste from fighting separately.

Another force for change comes from the industry level. Some manufacturers are fighting price wars by trading price for volume. Many manufacturers give up the red ocean of low-end mass products and take a high-quality high-price route, creating high-reliability products for 800V high voltage, high-level smart driving, and AI scenarios, which are in short supply in the market.

Looking back from 2026, the sense of despair of having no chips to use has quietly gone away.

Car company executives who squatted in front of Bosch and foundry gates in 2021 probably wouldn't have thought that the sharp pain of a supply chain rupture would open the road for domestic substitution of automotive-grade chips. The domestication rate of automotive-grade chips rose from 4.7% in 2022 to 16.8% in 2025, and is expected to reach 25% to 30% in 2027.

This road is far from over, but the value has already begun to appear.

When complete vehicle factories no longer need to stop production lines for a chip shortage, then delivery cycles will shorten, and vehicle price fluctuations will also converge. This is a dividend that consumers cannot see but actually enjoy.

The premise of all this is holding the lifeline of chips in our own hands. This self-redemption originating from chips proves that China's automotive industry and semiconductor industry can really achieve each other.