Bank of China Securities·2025-09-23 00:35Investment Rating - The industry investment rating is "Outperform the Market," indicating that the industry index is expected to perform better than the benchmark index over the next 6-12 months [33]. Core Insights - The launch of the SAIC Roewe M7 DMH, which features the Doubao deep thinking model, marks a significant advancement in smart cockpit technology, redefining human-vehicle interaction from command-based to natural interaction [11][12]. - Figure AI has raised over $1 billion in Series C funding, achieving a post-money valuation of $39 billion, highlighting the growing interest in humanoid robotics and its potential applications in both domestic and commercial settings [13][15]. - NVIDIA is pushing for the development of a new microchannel liquid cooling plate (MLCP) technology due to the high power consumption of new AI platforms, which could become a strategic material in the industry [16][18]. Summary by Sections Smart Cockpit - The SAIC Roewe M7 DMH integrates the Doubao deep thinking model, enhancing its ability to understand ambiguous commands and execute complex vehicle control tasks, thus improving user experience significantly compared to conventional models [11][12]. - The smart cockpit is evolving from a simple command execution system to a more intuitive and emotionally engaging partner for users, indicating a shift in the automotive industry's competitive landscape towards data, computing power, and algorithm capabilities [12][11]. Humanoid Robotics - Figure AI's recent funding round reflects strong market confidence in humanoid robots, with plans to expand their production and deployment capabilities for both household and commercial applications [13][15]. - Despite the high valuation, concerns remain regarding Figure AI's ability to convert demonstration capabilities into actual delivery and operational success in real-world scenarios [14][15]. Cooling Technology - NVIDIA's MLCP technology is expected to address the cooling challenges posed by high-power AI platforms, with costs projected to be 3 to 5 times higher than traditional cooling solutions, indicating a shift in the cooling technology landscape [16][18]. - The introduction of MLCP is seen as a necessary evolution in the industry, as traditional cooling methods are becoming inadequate for the increasing power demands of AI hardware [17][18].