如果您希望可以时常见面，欢迎标星收藏哦~ 该芯片采用7纳米多核异构架构，单颗芯片NPU高达512TOPS，通过多芯片协同方案可实现最高 2048TOPS的算力输出，LPDDR5内存带宽高达204 GB/s。同时，"星辰一号"具备独立的高性能安 全岛，能够在自动驾驶数据路径上实现强隔离、高可靠性和功能安全覆盖。该芯片还支持EVITA Full安全标准，集成了硬件安全模块，能够执行国家密码算法和国际加密协议。 尤为关键的是，为加速AI模型的开发与部署，"星辰一号"提供了高性能的AI工具链，对CNN和 Transformer架构进行了深度优化，为车企实现L2+至L4级智能驾驶提供了平台化的支撑能力。 此前，芯擎推出的智能座舱芯片"龍鷹一号"已成功搭载于领克系列、银河系列、一汽红旗系列、沃 尔沃、长安启源系列等数十款量产车型中。去年底，"龍鹰一号"成为首个获得德国大众总部海外车 型订单的国产座舱芯片，2024年稳居同类芯片装机量首位。目前，芯擎科技仍是国内目前唯一可 以大规模量产7纳米智能座舱SoC的供应商。 芯擎科技创始人兼CEO汪凯博士表示："'星辰一号'的量产是公司新的里程碑。智驾市场逐步成 熟，大家对高算力、 ...

Core Viewpoint - Chery Automobile is initiating a self-research chip program to develop its own vehicle MCU and intelligent driving SoC, aiming to enhance its competitiveness in the smart vehicle sector [2][4]. Group 1: Self-Research Chip Initiative - Chery's self-research chip plan is in its early stages, with recruitment for NPU design architects and senior chip design engineers underway [3][4]. - The company aims to transition from a traditional automaker to a modern intelligent industry cluster, with a focus on AI and smart technologies as a key growth area for the next 20 years [4][5]. - Chery's goal is to enter the top tier of intelligent driving capabilities by 2025, which necessitates the development of high-level intelligent driving chips [8][9]. Group 2: Cost Reduction Strategy - The shift towards self-research chips is driven by the need to reduce costs associated with purchasing high-end intelligent driving chips from suppliers [11][12]. - For instance, the cost of high-end driving hardware based on a 100 Tops computing platform for BYD is reported to be 4,000 RMB per unit, leading to a total cost of 16 billion RMB for 4 million vehicles [11]. - Chery's projected sales of over 2.6 million vehicles in 2024 indicate that its investment in intelligent driving will also reach a significant scale, potentially exceeding 10 billion RMB [11]. Group 3: Current Market Position and Challenges - Despite the self-research initiative, Chery currently relies heavily on external suppliers for its intelligent driving capabilities, which poses a challenge in achieving its 2025 goals [13][18]. - The company has established partnerships with leading suppliers like Huawei, NVIDIA, and Horizon Robotics to enhance its smart driving solutions [13][18]. - Chery's internal R&D team consists of over 5,500 personnel, indicating a strong foundation for developing autonomous driving technologies [18]. Group 4: Future Outlook - The ideal outcome of Chery's self-research efforts is to achieve technological independence and develop advanced features, while the worst-case scenario could result in merely evaluating supplier capabilities without significant technological advancement [19]. - The company is adjusting its strategy by integrating some self-research teams into its main R&D institute to ensure timely delivery of competitive products [17].