
Shanghai Auto Show Day 1: Goldman Sachs focuses on "Robotaxi, domestic self-driving chips, urban NOA, intelligent cockpit"

Goldman Sachs stated that the cost of Robotaxi has significantly decreased, with Pony AI's seventh-generation Robotaxi overall material costs dropping by 70% and LiDAR costs decreasing by 68%; the computing power of domestically produced autonomous driving chips has improved, and the NOA (Navigation on Autopilot) solutions for different price range models have also been upgraded
The 2025 Shanghai Auto Show officially opened on April 23. On the first day of the Shanghai Auto Show, Goldman Sachs focused on several automotive supply chain companies and found that technologies such as autonomous driving and intelligent cockpits are accelerating upgrades.
In a research report on April 23, Goldman Sachs stated that on the first day of the Shanghai Auto Show, it mainly observed five sub-sectors of automotive supply chain companies:
- Robotaxi costs have significantly decreased, and commercialization is accelerating; 2. Domestic autonomous driving chips have improved computing power; 3. Multi-price range models have upgraded NOA solutions; 4. Integration of cockpit and driving domain; 5. Automotive semiconductors and components.
According to the report, Goldman Sachs expects the Chinese ADAS/autonomous driving chip market to grow at a compound annual growth rate of 18%, reaching USD 5.5 billion by 2030; the shipment volume of onboard cameras is expected to grow at a compound annual growth rate of 22%, reaching 343 million units by 2030. This growth is primarily driven by the upgrade of high-end autonomous driving features and the demand for product differentiation.
Robotaxi Costs Significantly Decrease, Commercialization Accelerates
The report stated that Pony AI showcased its seventh-generation Robotaxi model. Compared to the sixth generation, the overall material cost of the seventh-generation Robotaxi has decreased by 70%, with the cost of the onboard computing unit reduced by 80% and the cost of LiDAR decreased by 68%. The seventh-generation Robotaxi will be based on various models from cooperative OEMs, including BAIC Arcfox αT5, Toyota BZ4X, and GAC Aion V.
This new generation of Robotaxi performs excellently in durability and reliability, with a designed lifespan of up to 600,000 kilometers or 10 years, and has upgraded redundant systems and software functions to enhance safety in the event of hardware failures.
According to Pony AI's management, the new domain controller is more powerful and efficient, capable of tracking over 500 pedestrians on the road simultaneously. Goldman Sachs is optimistic about its accelerated commercialization deployment and the technological advantages brought by continuous hardware and software iterations.
Domestic Autonomous Driving Chip Computing Power Improvement
Horizon Robotics released the Journey 6P product and SuperDrive solution, achieving a computing power of 560 TOPS, focusing on urban NOA functions, suitable for vehicles priced above 150,000 yuan, and will be mass-produced first in Chery models. Key features of this solution include:
Reliability: Highly flexible speed adjustment when encountering pedestrians or vehicles, and risk prediction under challenging road conditions;
Capability: Accurate identification of complex roads or navigation through narrow roads;
Intimacy: Improved interaction between the vehicle and the driver.
NOA Solution Specification Upgrade
Several companies showcased NOA (Advanced Driver Assistance) solutions for different price range models, including Deep Vision Technology, Desay SV, Zhixingzhe, Qingtong Intelligent Driving, and Lenovo.
These solutions are primarily developed based on the Horizon J6 series chip platform, providing consumers with more cost-effective autonomous driving features
Cockpit and Driving Domain Integration: AI-Driven User Experience Upgrade
With the development of smart cockpits and intelligent driving technologies, domain integration has become a new development direction. Companies such as iFlytek, ThunderSoft, and Chipone have showcased AI-based cockpit solutions that enhance users' multi-sensory and emotional interaction experiences.
The AI cockpit solution presented by iFlytek is supported by the Spark large model, providing AI interaction, AI agent platform, and iFlycode functionality. The smart solution 3.0 launched by Desay SV focuses on emotional and multi-sensory interaction, as well as health monitoring to improve user experience.
Automotive Semiconductors and Components
In addition to the aforementioned fields, companies like AAC Technologies have showcased solutions for automotive acoustics, optics, RF front-end modules, and in-vehicle chips, providing critical support for the overall upgrade of automotive electronics