通信:小米SU7发布,算力与智能化融合成智能汽车趋势
Huajin Securities·2024-03-28 16:00

Investment Rating - The report maintains an investment rating of "Outperform" for the industry, indicating a projected return exceeding the Shanghai and Shenzhen 300 Index by more than 10% over the next six months [3]. Core Insights - The launch of the Xiaomi SU7 marks a significant step in the integration of computing power and intelligence in smart vehicles, with a starting price of 215,900 CNY and various configurations available [1][2]. - The report highlights the transition from L2 to L3+ autonomous driving capabilities in China's passenger vehicles, with an expected penetration rate of 51% for L2 features in 2023 [1]. - The smart cockpit market in China is projected to grow from approximately 73.9 billion CNY in 2022 to over 100 billion CNY by 2025, with a compound annual growth rate (CAGR) of 12.7% [2]. Summary by Sections Investment Rating - Preferred stocks include: - 301191.SZ (Feiling Kesi) - "Accumulate" [1] - 600105.SH (Yongding Co.) - "Accumulate" [1] - 002881.SZ (Meige Intelligent) - "Accumulate" [1] Industry Performance - The report notes that the smart electric vehicle segment is experiencing rapid growth, driven by advancements in autonomous driving technology and smart cockpit features [1]. Market Trends - The report emphasizes the shift towards Ethernet-based vehicle communication systems, with the market for Ethernet physical layer chips in mainland China expected to exceed 12 billion CNY by 2025, reflecting a CAGR of over 30% [2]. - The demand for high-bandwidth vehicle communication is pushing upgrades in vehicle architecture, transitioning to a centralized architecture with Ethernet as the primary communication method [2]. Technology Developments - The Xiaomi SU7 features advanced smart configurations, including the Xiaomi Surge OS and a self-developed large model for voice interaction and vehicle control [1]. - The report indicates that the integration of laser radar in smart vehicles is expected to increase significantly, with the number of sensors per vehicle projected to rise from approximately 8 to around 24 as the technology matures [1][2].