China’s electric vehicle industry is being pulled in two directions at once. On one side, a sudden escalation in Middle East tensions has exposed how vulnerable Chinese automakers remain to shipping chokepoints, raw-material concentration, and aftersales logistics. On the other, the industry is accelerating into an AI- and semiconductor-led future, with NIO, BYD, Huawei, Tencent, Nvidia, and a growing list of Chinese suppliers pushing deeper into autonomous driving, smart cockpits, and vehicle chips. Taken together, the latest developments from March 11 show that the next phase of the Chinese EV market will be defined not just by sales growth, but by resilience, compute, and control over the supply chain.
Middle East Tensions Expose a New Weak Point for Chinese EV Exports
China’s auto exports topped 7 million units in 2025 for the first time, and the Middle East contributed roughly 1.39 million units, or nearly 20% of that increase, according to D1EV-cited industry reporting. That made the region one of the most important overseas growth engines for Chinese carmakers, especially as domestic price competition intensified.
But the late-February 2026 escalation around the Strait of Hormuz has sharply changed the risk calculus.
The impact goes beyond slower deliveries. It touches four critical areas:
- Vehicle shipment delays or cancellations as shipping lines suspend or reroute services
- Freight and insurance cost spikes, with VLCC rates on China–Middle East routes reportedly hitting $400,000 per day in extreme cases
- Aftersales parts shortages, which can damage brand reputation faster than delayed new-car deliveries
- Financial and insurance exposure in high-risk markets
A useful framework cited by Gasgoo CEO Zhou Xiaoying suggests that exposure depends on:
Impact ≈ Middle East export dependence × reliance on Gulf ports/transshipment hubs × local aftersales parts readiness × financial/insurance exposure
That equation helps explain why some brands are more vulnerable than others.
Which Chinese Carmakers Are Most Exposed?
Among major exporters to the Middle East, Chery, BYD, SAIC, Changan, and Geely stand out. Gasgoo data cited by D1EV shows that in January–November 2025, Chery led Chinese passenger car exports to the Middle East with 263,679 units.
Chery is seen as one of the most exposed players because of its scale and concentration in Iran, where it relies heavily on KD kit assembly with local partners. If foreign-exchange approvals tighten and shipping through Hormuz remains disrupted, both parts supply and complete-vehicle exports could be hit.
Other automakers face different risks:
- SAIC MG is highly dependent on regional logistics order and major hubs such as Jebel Ali Port in Dubai
- Great Wall Motor is especially sensitive to rising freight and insurance costs because SUVs and pickups are larger, heavier, and more expensive to ship
- BYD may face less immediate volume damage, but its strategic expansion timetable in Saudi Arabia could be delayed
- Changan, SAIC, and Geely may be somewhat cushioned in Saudi Arabia because some shipments can be rerouted via the Red Sea and Jeddah Port, albeit with higher inland transport costs
Exposure Snapshot: Key Chinese Automakers in the Middle East
| Automaker | Main Risk in Current Crisis | Strategic Weakness | Potential Mitigation |
|---|---|---|---|
| Chery | KD kit and vehicle shipment disruption | High exposure to Iran | Diversify assembly and logistics routes |
| BYD | Slower dealer buildout and new-model rollout | Strategic timing risk in Saudi/GCC | Accelerate local distribution infrastructure |
| SAIC MG | Delivery delays and inventory imbalance | Heavy reliance on Gulf transshipment hubs | Build buffer inventory, diversify ports |
| Great Wall Motor | Margin pressure from freight/insurance spikes | High shipping cost sensitivity for SUVs/pickups | Regional warehousing, local service stock |
| Changan/Geely | Route disruption but lower direct Gulf chokepoint dependency | Cost increases on alternate routing | More use of Red Sea/Jeddah channels |
The Supply Chain Risk Is Bigger Than Logistics
The shipping crisis is only part of the story. The deeper concern is upstream materials.
Methanol: A Hidden Risk to Auto Materials
Iran is the world’s second-largest methanol producer, with annual capacity of 17.16 million tons, about 9.2% of global capacity. China imported more than 7.92 million tons of methanol from Iran in 2025, accounting for over 55% of total methanol imports cited in the report.
That matters because methanol is a building block for a wide range of automotive materials, including:
- interior plastics
- seat foam
- adhesives
- coatings
- engineering polymers
A supply disruption here would not just affect fuel markets; it would feed directly into auto manufacturing costs.
Celestite and Strontium: A Direct EV Motor Risk
An even more EV-specific issue is celestite, the main source of strontium, which is used in high-performance magnetic ferrites found in:
- electric motors
- sensors
- braking systems
Iran reportedly holds around 85% of global high-grade celestite reserves, and China sources roughly 60%–70% of its celestite imports from Iran. Domestic inventories may be sufficient for only about three months, according to analysis cited by D1EV.
This is not theoretical. Following the April 2025 Bandar Abbas port blast and related disruptions, strontium carbonate prices doubled from RMB 8,000/ton in September 2024 to RMB 16,000/ton in June 2025.
For Chinese EV manufacturers, that raises the risk of:
- motor supply bottlenecks
- higher component costs
- delayed EV production schedules
- reduced flexibility in export planning
Oil Still Matters, Even for EVs
China imported an estimated 1.5 million barrels per day of Iranian crude in 2025, equivalent to about 75 million tons annually. When conflict intensified, Brent crude briefly jumped nearly 13% to around $82 per barrel.
Even in an EV-heavy market, higher oil prices matter because they raise:
- shipping costs
- petrochemical feedstock costs
- plastics and synthetic material prices
- operating costs across the automotive supply chain
In other words, geopolitics can still hit EV economics through the back door.
AI Is Becoming the New Core of Auto Competition
While logistics and raw materials dominate the risk side of the story, the growth side is increasingly about AI infrastructure and semiconductors.
Nvidia CEO Jensen Huang described AI as a “five-layer cake”:
- Energy
- Chips
- AI infrastructure
- Models
- Applications
His central argument is highly relevant for the auto industry: AI is no longer just software. It is an industrial system built on electricity, semiconductors, data centers, networking, and deployment hardware. Huang said the world has already invested hundreds of billions of dollars in AI, but full-scale adoption will require trillions more.
For EVs, that framework maps neatly onto the next battlefield:
- autonomous driving compute
- smart cockpit intelligence
- in-car AI agents
- cloud training infrastructure
- energy-intensive data center support for model development
Nvidia also showcased a Mercedes-Benz CLA equipped with its hands-off automated driving system, which reportedly completed a 22-minute urban drive in California without manual takeover, including construction zones and illegally parked vehicles.
That demonstration reinforces a larger trend: Chinese EV makers are no longer simply competing on battery size and price. They are competing on compute stacks.
Chinese Tech and EV Players Push Deeper Into Chips
Several developments on March 11 underline how quickly that transition is happening.
NIO Expands Beyond In-House Chips
NIO founder William Li said the company hopes to offer its in-house Shenji high-compute inference chip to more partners, including the auto industry and embodied AI sectors. NIO had also disclosed that its second smart chip, aimed at a broader customer base, has successfully taped out and is moving toward mass production.
This is significant because it suggests NIO sees semiconductor capability not just as a cost center for its own vehicles, but as a platform business.
China’s First Domestic High-Performance Automotive MCU Reaches Production
A separate milestone came with the DF30, described as China’s first domestic high-performance automotive-grade MCU to enter mass production vehicles in 2026. Built on a domestic 40 nm automotive-grade process and based on a RISC-V multi-core architecture, the chip has completed 295 tests, supports domestic AutoSAR software, and reaches ASIL-D functional safety.
It is already being used in models including:
- Dongfeng Haohan
- Mengshi 817
- eπ 007
MCUs may be less glamorous than AI accelerators, but they are foundational for:
- sensor data handling
- powertrain control
- lighting control
- smart cockpit systems
- safety-critical vehicle functions
At a time of geopolitical uncertainty, this kind of localization matters enormously.
Dreame-Linked Chip Startup Targets L4 Compute
At the AWE 2026 Chip Industry Summit, Dreame ecosystem company Xinji Chuanyue unveiled plans across smartphones, robotics, super AI PCs, and automotive chips. For smart vehicles, it previewed an integrated cockpit-driving chip built on a 2 nm process, with claimed single-chip compute of 2,000 TOPS—around three times the average compute of current high-end intelligent driving chips, according to the company.
If such a chip reaches production anywhere near those specifications, it could materially shift the economics of high-level assisted driving and potentially even L4 development.
Smart Driving and In-Car AI Are Converging
The line between vehicle software, consumer AI, and mobility services is getting thinner.
Tencent is reportedly developing a highly confidential AI agent for WeChat, with testing expected around mid-2026 and broader rollout potentially in the third quarter. The significance is not merely in messaging. WeChat connects to millions of mini programs across ride-hailing, food delivery, and services used by its 1.4 billion monthly active users.
Huawei, meanwhile, has introduced a HarmonyOS version of OpenClaw-style agent functionality via Xiaoyi Claw Beta, allowing smartphone users to create persistent AI agents that can execute multi-step tasks like:
- processing documents
- generating presentations
- replying to emails
- managing files
For the EV sector, this matters because China’s leading automakers increasingly position the car as part of a wider intelligent-device ecosystem. The same AI-agent logic moving into smartphones can ultimately flow into:
- voice assistants in smart cockpits
- in-car productivity tools
- service booking and charging orchestration
- multimodal personal mobility agents
This is also why companies like Xiaomi, Huawei, and Tencent matter to the Chinese EV market even when the immediate announcement is not a car launch.
Market Signals: EV Demand Is Holding, Even if the Mix Is Changing
According to CAAM data cited by D1EV, in February 2026 China produced 694,000 new energy vehicles and sold 765,000, down 21.8% and 14.2% year on year respectively. Even so, NEVs still accounted for 42.4% of new-car sales.
That headline needs context. Monthly volatility is common around seasonal effects, but a 42.4% share still shows that electrification remains structurally strong in the Chinese EV market.
Other notable market developments include:
- NIO shares rose 14% in Hong Kong trading on March 11
- XPeng gained nearly 4%
- CATL rose 9%
- CALB gained more than 8%
- Aion i60 cumulative sales surpassed 28,000 units, with both BEV and EREV variants available
Key Industry Data at a Glance
| Metric | Latest Figure | Why It Matters |
|---|---|---|
| China auto exports in 2025 | 7 million+ | Shows global scale of Chinese automakers |
| Middle East contribution | 1.39 million | Nearly 20% of export growth tied to region |
| Chery Middle East exports (Jan-Nov 2025) | 263,679 units | Highest among Chinese passenger car exporters there |
| NEV production, Feb 2026 | 694,000 | Reflects short-term market softness |
| NEV sales, Feb 2026 | 765,000 | NEVs still at mass-market scale |
| NEV share of new-car sales | 42.4% | Confirms deep penetration in China |
| Extreme Middle East freight rate | $400,000/day | Highlights logistics shock severity |
| DF30 process node | 40 nm | Domestic automotive MCU milestone |
| Xinji Chuanyue chip claim | 2,000 TOPS | Signals escalating smart-driving chip race |
Why This Matters Globally
The Chinese EV story is often told as a simple tale of low costs and fast product cycles. That is no longer enough.
What the latest news makes clear is that Chinese automakers now face two strategic tests at the same time:
-
Can they build a more resilient global operating model?
- diversified logistics
- regional distribution hubs
- local aftersales parts systems
- less dependence on single-source materials
-
Can they move up the value stack in intelligence?
- proprietary chips
- AI infrastructure
- autonomous driving compute
- operating-system level agent integration
These are not separate challenges. They are converging into a new definition of competitiveness.
A carmaker that controls more of its chip stack, software architecture, and supply base is better positioned to survive geopolitical shocks. And a carmaker with weak international logistics or aftersales support can quickly lose the advantages of advanced technology.
The Road Ahead
Expect Chinese EV brands to respond on two fronts over the coming quarters.
First, they are likely to rethink overseas logistics and regionalization, especially in the Middle East. That could mean:
- more local warehousing
- bigger aftersales parts buffers
- new shipping routes
- more CKD/KD diversification outside single-risk markets
Second, they will continue to invest aggressively in AI and chips, because smart driving, cockpit intelligence, and software-defined vehicle capability are increasingly central to margins and brand positioning.
BYD joining the International Automotive Task Force (IATF) is symbolic here. Chinese automakers are no longer just scaling exports; they are moving into standard-setting, core technologies, and ecosystem influence.
The takeaway is straightforward: the next era of the Chinese EV market will not be won solely by building more electric cars. It will be won by those that can ship reliably through crisis, secure critical materials, and deploy enough compute to make the car a truly intelligent machine.
