HAITONG INT'L: Investors hold optimistic attitude towards the future popularization of smart driving chips. Third-party chip suppliers shifting focus to Tier1 is key to competitiveness in the next stage.

HAITONG INT'L released a research report stating that, according to media reports, BYD COMPANY (01211) has adopted the BLACK SESAME (02533) car-grade autonomous driving computing chip, which will be used in vehicles under BYD Company Limited's Tengshi brand. The report believes that the news about Tengshi adopting BLACK SESAME is old news, and what truly boosted the stock price is the recent market enthusiasm surrounding AI and intelligent driving. The report mentions that from the popularity of DeepSeek, to the frequent exposure of humanoid Siasun Robot & Automation products, to BYD Company Limited's upcoming "Smart Driving Equality" strategy, the capital market is searching for high-quality targets that combine AI and hardware. In addition, Horizon Siasun Robot & Automation's recent surge in valuation has also triggered a revaluation of domestic intelligent driving chip companies, with BLACK SESAME being in the same category naturally benefiting from the upward trend. This increase in value is driven by emotions and reflects the market's optimistic outlook on the future popularization of intelligent driving chips. The report believes that algorithm iteration is still the top priority, and the integration of software and hardware is the key to breakthrough for chip manufacturers. As a leader in mass market sales, BYD Company Limited's smart driving equality strategy this year may accelerate the penetration of high-level intelligent driving throughout the industry, with L2+ functions expected to enter mainstream markets ahead of schedule, and the growth of the 100TOPS+ high computing power platform having a higher level of certainty. Although market expansion is positive for the upstream sector, the smart driving industry is deemed to be special: algorithmic and AGI convergence, transitioning from multi-stage/Rule-based architecture to single-stage/Learn-based architecture, with downstream resources still focusing on higher-level algorithms. Therefore, for downstream self-driving car manufacturers/algorithm suppliers, the powerful ecosystem support, higher computing power support, and higher development and deployment efficiency of NVIDIA's platform remains the first choice; for companies like Tesla and Huawei with a self-developed software and hardware ecosystem, their own hardware know-how provides significant advantages in terms of computing power utilization, software and hardware development synergy, and cost optimization, making it difficult for third-party chip suppliers to break through barriers. The report believes that the development trajectory for intelligent driving is still steep, and it is difficult to establish absolute barriers to leadership in stages. Staying in the game is still valuable compared to leading the industry, and for third-party chip suppliers, transitioning from providing only partial software support as a hardware Tier2 to a Tier1 position is crucial to maintaining their competitiveness in the next stage.