The Tipping Point
In June, the price hike of automotive-grade chips once again became the focus of the industry chain. Chip manufacturers like Infineon and STMicroelectronics have issued price increase notices for the second time this year, with significant price hikes in categories such as automotive-grade DRAM and NAND flash memory. Industry insiders state that some automotive storage specifications have seen an increase of approximately 180% over the past three months.
If one only interprets this as 'automotive chips are getting more expensive again,' it is easy to underestimate the true impact of this event. More importantly, AI data centers are changing the allocation logic of semiconductor capacity.
In the past, when the automotive industry discussed chips, the core question was 'is there a shortage?' Now, the question has become 'who is more qualified to secure capacity?' When AI data centers enter the same supply pool with higher prices, larger orders, and longer lock-in periods, the automotive industry may be forced to bear higher costs and lower supply priority, even if its demand remains stable.
This is not an ordinary price hike, but a typical example of the pressure from AI infrastructure construction being transmitted to the physical industry chain.
I. AI Orders Are More Profitable, Capacity Shifts to Data Centers
The superficial reason for the price hike in automotive chips is tight supply and demand, but the deeper reason is that chip manufacturers are recalculating their return on capacity.
The same production line and the same type of storage and power device capacity can serve data centers, as well as automotive, industrial control, and consumer electronics sectors. In the past, automotive customers had the advantage of long cycles and stable demand; however, during the expansion phase of AI infrastructure, data center customers have more obvious advantages: larger order scales, longer lock-in periods, and a stronger willingness to pay for supply certainty.
This means that chip capacity is no longer allocated just based on 'who needs it,' but reordered based on 'who can bring higher returns.' The automotive industry's demand has not weakened, but when competing for the same batch of critical resources with AI data centers, its price tolerance and profit elasticity are not advantageous.

Source: Official announcements from STMicroelectronics, Reuters, TrendForce, AEC-Q100 Rev. J, compiled by Insight Research.
This change is already reflected in the business expectations of chip manufacturers. STMicroelectronics has raised its 2026 data center business revenue target to approximately $1 billion; Infineon expects revenue from AI data center applications to be approximately €1.5 billion in fiscal year 2026, with the potential to reach €2.5 billion in the following fiscal year.
For chip manufacturers, AI data centers are not just ordinary new demand, but a new business that can significantly improve their revenue structure. As long as demand for AI servers, enterprise SSDs, and power management continues to rise, upstream manufacturers will be more inclined to allocate incremental resources to data center-related applications. The price pressure faced by the automotive industry essentially stems from this 'capacity opportunity cost' being raised.
II. Automakers Struggle to Raise Prices, Gross Margins Under Pressure
The real difficulty for the automotive side is not the rise in chip procurement prices, but the inability to smoothly pass costs downstream.
New energy vehicles are still in a stage of high-intensity competition, where terminal prices, promotional policies, and configuration benefits all influence user decisions. If automakers raise prices directly, they may lose orders; if they do not, upstream chip costs will compress gross margins. For automakers already bearing pressure from batteries, marketing, channels, and R&D investment, the rise in automotive chip prices will further amplify operating leverage pressure.
More importantly, intelligent configurations have become a major selling point for new energy models. Cockpit chips, intelligent driving domain controllers, storage, and power devices are not optional peripheral parts, but core components that support product differentiation. Automakers cannot easily cut intelligent configurations, nor can they casually reduce functional experiences that are obvious to users; they can only rebalance between price, configuration, optional packages, and delivery rhythm.

Source: TrendForce, AEC-Q100 Rev. J, 21st Century Business Herald, compiled by Insight Research.
Therefore, the impact of automotive chip price hikes on automakers is often not manifested at once, but slowly released through multiple links: in the short term, gross margins are under pressure; in the medium term, optional package prices, model benefits, and configuration combinations may be adjusted; in the long term, automakers will re-evaluate which intelligent capabilities must be standard and which can be provided in tiers.
This is also a signal that competition in the new energy vehicle industry has entered a new stage. In the past, automakers competed more on product definition and sales price; moving forward, supply chain cost management capabilities will more directly affect product strategy.
III. The Higher the Intelligence, The Stronger the Sensitivity to Chips
The chip pressure on new energy vehicles will not fall equally on all models. The more intelligent, higher-end, and experience-focused a model is, the more sensitive it is to chip price hikes.
In the past, automotive electronics served more for power control, body control, and basic safety systems; while chip costs were important, they did not determine the selling point of the entire vehicle. Now, smart cockpits, high-level assisted driving, cockpit-driving integration, and central computing platforms are driving up the chip value per vehicle. The voice interaction, smoothness of the car system, assisted driving experience, and OTA iteration capabilities perceived by users all rely on higher-specification computing, storage, and power management systems.
Relevant data shows that the number of chips in a highly intelligent new energy vehicle has exceeded 1,000, nearly double that of traditional fuel vehicles. The scale of China's new energy vehicles is still expanding; in 2025, China's new energy vehicle production was 16.626 million units and sales were 16.49 million units, a year-on-year increase of 29% and 28.2%, respectively.

Source: CAAM, industry reports, compiled by Insight Research.
The key here is not the 'increase in the number of chips' itself, but that the cost structure of automotive products is changing.
In the era of fuel vehicles, mechanical systems such as engines, transmissions, and chassis determined brand capability; after electrification, batteries became the core of costs; as intelligence continues to advance, chips, sensors, software, and computing platforms have begun to determine the product experience. In other words, the cost center of new energy vehicles is extending from 'electrification hardware' to 'intelligent systems'.
This will change the way automakers compete. High-end models may continue to maintain brand premiums with stronger intelligent driving, greater computing power, and better cockpit experiences; mid-range models need to make trade-offs between cost and experience; entry-level models may emphasize the stability of basic configurations rather than blindly stacking intelligent functions. The chip price hike is not simply driving up costs, but is forcing automakers to re-divide the value hierarchy of intelligent configurations.
IV. Long Automotive Certification Cycles, Replacement Is Not Easy
The difficulty of the automotive supply chain also lies in the fact that automotive-grade chips cannot be replaced as quickly as consumer electronics chips.
Consumer electronics products update quickly, and supplier switching is relatively flexible; automotive-grade chips need to meet requirements for high temperature, low temperature, vibration, lifespan, and functional safety, and pass reliability verifications such as AEC-Q100. For vehicle manufacturers, a chip is not a simple procurement item, but a key component embedded in the vehicle platform, software system, safety system, and after-sales life cycle.

Source: AEC-Q100 automotive chip reliability testing standard, compiled by Insight Research.
This determines that when facing price hikes, automakers cannot simply adopt a 'switch to a cheaper supplier' strategy. The closer to core links such as power, intelligent driving, domain control, and safety control, the higher the threshold for substitution; the closer to mass-produced models, the greater the switching cost.
Therefore, the real reminder of this round of price hikes for automakers is that supply chain management must be proactive. In the past, procurement departments focused more on price negotiation and delivery guarantees; in the future, the chip supply chain needs to enter vehicle planning, platform design, and technical route selection earlier. Automakers need to judge not only 'which chip to buy now,' but also 'which chips might be squeezed out by AI data centers in the next three years,' 'which devices need dual-supplier verification,' and 'which domestic solutions are worth introducing in advance'.
This is also why supply chain capability is moving from a back-office function to a front-office strategy. Whoever can lock in key chips in advance, establish an alternative verification system, and manage cross-cycle price risks is more likely to maintain product rhythm in the next round of supply and demand fluctuations.
V. Domestic Substitution Has a Window, Verification Remains a Threshold
AI data centers squeezing out some capacity does not mean the automotive industry can only passively accept it. For domestic automotive-grade chips, this may instead become a window to enter the mainstream supply chain.
When overseas giants invest more resources in high-return AI data centers, domestic automakers will be more willing to evaluate alternative solutions such as domestic storage, power devices, MCUs, and automotive computing chips. Especially in scenarios such as cockpits, body control, low-to-mid-level intelligent driving, and power management, domestic suppliers have the opportunity to strive for verification opportunities based on response speed, local service, and customization capabilities.
However, the key to domestic substitution is not 'whether there is a product,' but 'whether it can be safely used in mass production by automakers.' The automotive market truly values not short-term low prices, but long-term reliability, mass production consistency, failure rate control, delivery stability, and after-sales traceability. Getting on the supplier list is only the first step; to truly form a substitution, it must go through small-batch verification, vehicle platform adaptation, mass production ramp-up, and long-term quality testing.
The global automotive semiconductor market is still in a growth channel. MarketsandMarkets expects the global automotive semiconductor market to grow from $77.42 billion in 2025 to $133.05 billion in 2030, with a compound annual growth rate of approximately 11.4%.

Source: MarketsandMarkets, compiled by Insight Research.
This set of data shows that the opportunity for domestic automotive-grade chips comes not only from short-term price hikes, but more from the long-term increase in the value of automotive electronics. Short-term supply tightness will make automakers more willing to open verification windows, while long-term market growth will determine whether domestic manufacturers can move from point-based substitution to platform-based supply.
For domestic chip companies, what is truly worth striving for is not one-off orders, but vehicle platform designations. Once they enter the mainstream automaker platform and pass multiple rounds of mass production verification, subsequent product iteration, customer stickiness, and bargaining power will significantly improve.
Conclusion: The AI Boom Is Rewriting the Cost Curve of the Physical Industry
The true impact of the automotive chip price hike lies not in short-term cost fluctuations, but in the fact that the automotive industry has begun to be included in the resource competition brought about by the expansion of AI infrastructure.
In the past, automakers focused on 'is there enough supply?'; now, they need to judge 'where will the capacity flow?' When AI data centers become a more powerful competitor for capacity, the automotive industry must re-examine the supply stability, cost elasticity, and substitution paths of key devices.
Moving forward, the focus of competition in the new energy vehicle supply chain will continue to shift forward. Price wars will still exist, but deeper capability competition will be reflected in key chip lock-in, domestic substitution verification, multi-supplier system construction, and cross-industry capacity analysis.
This also makes the value of industry research more prominent. Looking at the automotive industry alone, it is easy to interpret price hikes as a procurement problem; looking at the semiconductor industry alone, it is easy to underestimate its impact on vehicle configuration and profit models. Only by observing AI data centers, storage prices, automotive certification, vehicle platforms, and domestic substitution in the same industry chain can one judge whether this round of price hikes is a short-term disturbance or the beginning of a change in the rules of competition in the automotive supply chain.