PWC has published in its report ‘Global semiconductor industry outlook 2026‘A projection: completely autonomous vehicles (level 5) will need five times more semiconductors than a traditional car. And its cost will multiply by ten.
A current basic car (level 0) carries between 200 and 300 chips valued at about $ 500. They serve mainly for engine control, safety and infotainment systems. Nothing especially sophisticated.
Why is it important. The semiconductor content of a car will multiply by ten when we reach full autonomy. From the current 500 dollars to more than $ 5,000 per vehicle. This jump will change the cost structure of the car industry and create a gigantic market for chips manufacturers.
The climbing. Each level of autonomy shoot the needs:
- Level 1 and 2: driver assistance and automatic distance maintenance. The cost rises to about 800 dollars in chips.
- Level 3: Autonomous driving on highways without constant supervision. More than 1,000 semiconductors and about $ 2,000.
- Level 4: Autonomous operation in the city, limited to specific areas. The cost reaches $ 3,200.
- Level 5: Total autonomy in any climatic and traffic condition. More than $ 5,000 per vehicle.
The context. The explosion in semiconductor content responds to the fact that autonomous vehicles must process large volumes of real -time data, and that translates into computing capacity, and more information collected from the environment.
A level 3 car already needs more than 1,000 semiconductors to capture information around it, high performance computer chips (HPC) to process it, advanced driver assistance systems (ADAS) and electronic control units to maneuver the vehicle.
- The most advanced designs will add chips for vehicle-a-all communication (V2X), exchanging data with the road and other cars.
- Meanwhile, the transition to the electric car continues to increase the demand for power semiconductors for investors and battery management.
In detail. The architecture of an autonomous car is divided into technological layers, each with its own semiconductors:
- Sensors: radars, lidar sensors, cameras and ultrasound that capture the environment. Each type requires specialized chips to process its signal.
- Computing: High performance CPUS and GPUS that execute AI algorithms to interpret data and make decisions.
- Electrification: semiconductors of silicon carbide (sic) and of Gallium Nitruro (Gan) to manage electric power with greater efficiency.
- Connectivity: Chips for 5G communication, V2X and remote updates.
- Control: electronic units that translate decisions into mechanical actions.
Image: Freepik, Xataka.
Turning point. Level 4 vehicles will start commercially climbing around 2030, mainly in Robotaxis services In delimited urban areas or transportation hub-to-hub of goods. Level 5, capable of operating without a steering wheel in any condition, will arrive much later, perhaps even in the next decade but something later.
This calendar gives time to the semiconductor industry to prepare, but also points out that the great ball is still seen.
Yes, but. This projection assumes mass production and that technology will fulfill its safety promises. Nothing guaranteed. Technical difficulties, regulatory frameworks and social acceptance can slow down the deployment.
Between the lines. Tesla, General Motors and Ford already design their own chips for central computers and ADAS systems. They seek control over critical technology and differentiation. Traditional semiconductor manufacturers (Infineon, NXP, STMICROELECTRONICS) will share market with these new competitors.
In Xataka | I have tried a totally autonomous taxi. This is traveling without driver
Outstanding image | GibblesMash Asdf
GIPHY App Key not set. Please check settings