Core Viewpoint - Multiple automotive manufacturers are increasingly adopting open-source strategies for their self-developed automotive operating systems, aiming to enhance efficiency, reduce costs, and foster innovation in the industry [2][3][4]. Group 1: Industry Trends - The automotive industry is transitioning from being merely a "transportation tool" to an "intelligent terminal," necessitating specialized automotive operating systems [3]. - The new generation of automotive operating systems serves as the "nerve center" connecting hardware computing power with AI algorithms, which is crucial for the efficient and safe operation of smart vehicles [3]. Group 2: Open Source Benefits - Open-sourcing automotive operating systems is seen as a way to improve efficiency by addressing inconsistencies in technology standards and slow update cycles in closed-source systems [3]. - The establishment of an open and efficient automotive operating system ecosystem is essential for better resource utilization and innovation within the industry [3][4]. - Open-source platforms can gather technical expertise from the industry, optimize innovation resource allocation, and accelerate the development and validation of new technologies [3][4]. Group 3: Impact on Manufacturers - The launch of the Li Auto Star Ring OS, which can complete chip adaptation in as little as four weeks, significantly reduces dependency on single chip suppliers and saves up to five months compared to traditional methods [4]. - The self-developed virtualization technology in Li Auto Star Ring OS minimizes performance loss and storage resource usage, helping manufacturers cut production and R&D costs [4]. Group 4: Consumer Implications - Open-source systems are expected to lower R&D costs for manufacturers, leading to more affordable vehicle prices for consumers [4]. - The construction of an open ecosystem is likely to enhance vehicle performance, providing consumers with improved driving and riding experiences [4]. - For instance, the Li Auto self-developed system can double response speed, increase response stability by five times, and shorten braking distance by seven meters, thereby enhancing driving safety [4]. Group 5: Future Industry Collaboration - Experts suggest that in the "second half" of the smart automotive era, the technical advantages of individual companies will be limited, and an open technological foundation will better promote collaborative innovation and development within the industry [5].

Group 1: Core Insights on New Quality Productive Forces - The development of new quality productive forces is being driven by technological innovation, showcasing significant high-quality development outcomes across various regions [1][2] - Key trends include the acceleration of traditional industry upgrades through technological empowerment, strong growth in emerging and future industries, and enhanced innovation capabilities among enterprises [1] - Challenges such as insufficient originality and a mismatch between talent training structures and industry needs have been identified, necessitating targeted recommendations [1] Group 2: Recommendations for Enhancing Technological Innovation - A call for the development of major application scenarios to support new quality productive forces, particularly in artificial intelligence and low-altitude economy sectors [1] - Emphasis on strengthening the role of enterprises in major technological tasks and increasing their involvement in project leadership and evaluation processes [1] - The establishment of a talent cultivation system that aligns with the needs of cutting-edge technology development and national strategic requirements [2] Group 3: Advancements in Artificial Intelligence and Open Source Innovation - Recommendations to accelerate the integration of artificial intelligence in scientific research, focusing on key fields such as chemistry, high-energy physics, and biotechnology [3] - The promotion of open-source innovation through support for native Chinese open-source projects and the establishment of sustainable funding mechanisms for open-source communities [4][5] Group 4: Enhancing National Strategic Technological Forces - The importance of leveraging national strategic technological forces to foster deep integration of technological and industrial innovation, enhancing economic and national competitiveness [6][7] - Suggestions to build a robust research team capable of bridging the gap between academic research and practical application, thereby facilitating the transformation of technological achievements into industrial applications [6][7] Group 5: Safety and Innovation in New Quality Productive Forces - The need for a dual focus on innovation and safety, particularly in critical sectors such as finance and healthcare, to ensure the secure application of AI technologies [8] - Recommendations for establishing minimum safety investment standards for key units and enhancing financial support mechanisms for technological innovation [8] Group 6: Strengthening Basic Research and Talent Development - A call for improved organization and collaboration in basic research to meet national strategic needs, emphasizing the role of various stakeholders including government, universities, and enterprises [9][10] - The establishment of a coordinated mechanism for education and technology talent development, ensuring alignment with the evolving demands of new quality productive forces [20][21] Group 7: Promoting Low-altitude Economy and Cross-industry Integration - The low-altitude economy is recognized as a vibrant sector with significant potential, necessitating reforms in airspace management and infrastructure development [11] - Encouragement for cross-industry integration, particularly between the automotive and information sectors, to enhance the digitalization and safety of vehicles [18][19] Group 8: Financial Innovations Supporting Technological Development - Recommendations for commercial banks to innovate financial products and services tailored to technology enterprises, enhancing their digital capabilities and risk management [16][17] - The establishment of collaborative platforms among government, universities, and investment institutions to support technology financing and risk compensation [17]

责编 |梦依丹 出品丨AI 科技大本营（ID：rgznai100） 自从 GPT-3 横空出世，生成式 AI 彻底点燃了全球科技圈： 尽管 LLMs 如 GPT-4、Claude 等展现了惊人的能力，但闭源模型的闭源特性让研究者难以深入理解其运作机制，同时开源模型的开放程度有限： Moxin-7B：从预训练到强化学习，全面透明的 AI 革新 Moxin-7B 的诞生，正是为了解决这一问题！ 它由来自东北大学、哈佛、康奈尔等机构的研究团队联合开发，完全遵循"开源科学"原则，公开了从数据 清洗到强化学习的全流程细节，从预训练到 DeepSeek 同款强化学习，成为目前透明度最高的开源 LLM 之一。 2. 高性能低成本：小模型的大能量 零样本任务：在 ARC-C（AI2推理挑战）上达到 58.64%，超越 LLaMA 3.1-8B（53.67%）和 Qwen2-7B（50.09%）。 数学推理：经过 RL 微调后，在 MATH-500 上准确率 68%，超越 70B 参数的Llama-3-Instruct 模型（64.6%）。 长上下文支持：通过滑动窗口注意力（SWA）和分组查询注意力（GQA），高效处理 32K ...

Core Insights - The establishment of the Guangdong Open Harmony Intelligent IoT Innovation Center marks a new phase in the development of the open-source Harmony ecosystem in China [1] - The center will focus on the OpenHarmony operating system, emphasizing core technologies such as real-time computing and AI, and aims to develop the M-Robots OS, a distributed heterogeneous multi-machine collaborative robot operating system [1][2] - M-Robots OS is the first operating system based on open-source Harmony, which aims to enhance collaboration among various types of robots and facilitate intelligent interaction with humans and the environment [2] Group 1 - The Guangdong Open Harmony Intelligent IoT Innovation Center is the first provincial-level open-source Harmony manufacturing innovation center in China, led by Shenzhen Kaihong Digital Industry Development Co., Ltd. and Shenzhen Zhongsoft International Technology Service Co., Ltd. [1] - M-Robots OS was officially released in April this year and represents a new stage in China's robot operating systems, focusing on multi-machine collaboration and collective intelligence [1][2] - The system has been validated in industrial robotic arms and collaborative robots, with plans for commercial testing on industrial production lines by 2025 in collaboration with partners in Guangdong [2] Group 2 - The center will adopt an "open-source community" model to iteratively advance the system, with a 2.0 version expected by the end of 2025 and a complete development ecosystem within three years [2] - The establishment of the center is expected to accelerate the gathering of upstream and downstream enterprises in the open-source Harmony ecosystem, creating an innovative hub and industrial cluster for open-source Harmony application software [3]

要说全球开源大模型生态圈里，谁最让人"魂牵梦绕"？ 阿里，当仁不让。 就在上周，继DeepSeek和"AI教母"李飞飞之后，英伟达也相中阿里了。除了在最新的"混合推理模 型"千问3宣布开源当日，火速官宣接入适配后，5月9日，英伟达还开源了全新的代码推理模型Open Code Reasoning （后续简称：OCR），包括7B、14B、32B三种尺寸，基础模型用的都是通义千问。 在LiveCodeBench 评测中，成功超越Open AI 公司o3-mini和o1模型的英伟达OCR- Qwen-32B-Instruct模型，正是基于Qwen2.5-32B微调形成的。 在通义千问已经迭代至3.0版本，模型性能再度突破的当下，英伟达居然还基于上一代千问 模型做出了比肩全球一流水平的模型，让人不禁想问，千问到底还有多少隐藏潜力待各方解 锁？ DeepSeek、李飞飞后， 英伟达也相中了"通义千问" 目前，英伟达开源的OCR系列模型的代码及数据集，已公开分享至全球最大AI开源社区 Hugging Face平台上，供开发者们免费浏览学习。 其 中 ， 英 伟 达 OCR-Qwen-32B-Instruct 在 LiveC ...

Group 1 - The "Guangdong Open Source Harmony Smart IoT Innovation Center" has been officially unveiled, marking the first provincial-level open-source Harmony manufacturing innovation center in China [1] - The center is led by Shenzhen Kaihong Digital Industry Development Co., Ltd. and ChinaSoft International Information Technology Co., Ltd., focusing on core technologies such as real-time computing and AI [1] - The center aims to develop a highly secure, widely compatible, and real-time distributed heterogeneous multi-machine collaborative robot operating system called M-Robots OS [1] Group 2 - Harmony OS has achieved a breakthrough in domestic operating systems, expected to surpass Apple's operating system in the Chinese market by Q1 2024, becoming the second-largest operating system for mobile phones in China [1] - The development of the Harmony ecosystem has shown significant results, with rapid growth in the development and adaptation of Harmony ecosystem applications [1] - Qianhai has become a hub for the development of the Harmony ecosystem, with the first Harmony Ecosystem Conference scheduled for November 2024 [2]

插图来源：CHARIS TSEVIS, PHOTOS COURTESY OF THE COMPANIES 美国引领创新，中国擅长迭代。这个说法（或者更加直白的版本："中国擅长模仿"）长期主导着关于世界第一大经济体和第二大经济体的相对实力 的诸多讨论。传统观点认为，美国始终矗立在技术前沿，用颠覆性的产品和服务引领全球经济；而中国往往会借鉴这些创意，迅速推出更廉价、或 许质量稍逊的产品版本。 在人工智能领域里，这一陈见多年来似乎都有据可依。面对财力雄厚、人才济济的美国科技巨头，中国企业始终疲于追赶。但在今年1月，一家中国新创 企业打破了这种说法。 总部位于中国浙江省杭州市的深度求索 （DeepSeek）——严格来说甚至算不上是一家科技公司，而是对冲基金幻方量化的衍生团队——发布了"推理 型"大语言模型（LLM）R1。其性能竟与美国的人工智能翘楚OpenAI公司在几个月之前发布的大语言模型o1旗鼓相当。R1感觉像是从天而降，不仅 创新性十足，成本也低得令人震惊。据DeepSeek透露，其前代模型V3最终的训练成本仅为600万美元。特斯拉（Tesla）的前人工智能科学家安德烈· 卡尔帕西（Andrej Karpath ...

Core Viewpoint - The establishment of the Guangdong Open Harmony Intelligent IoT Innovation Center marks a significant step in advancing the Open Harmony ecosystem and addressing challenges in the robotics industry through the introduction of the M-Robots OS, a distributed heterogeneous multi-machine collaborative robot operating system [1][2]. Group 1: Innovation Center Establishment - The Guangdong Open Harmony Intelligent IoT Innovation Center is the first provincial-level Open Harmony manufacturing innovation center in China, officially unveiled on May 13 [1]. - The center is led by Shenzhen Kaihong Digital Industry Development Co., Ltd. and Shenzhen Zhongsoft International Technology Service Co., Ltd. [1]. Group 2: M-Robots OS Development - M-Robots OS is the first distributed heterogeneous multi-machine collaborative robot operating system based on Open Harmony, officially released in April [1][2]. - The system aims to unify standards for efficient collaboration among various types of robots and enhance intelligent interaction between robots and their environments [2]. - Industrial applications of M-Robots OS are expected to undergo commercial testing by 2025 in collaboration with partners in Guangdong [2]. Group 3: Ecosystem Expansion - The Open Harmony ecosystem has over 70 community co-construction units and has launched various products, with over 1 billion devices equipped, facilitating advancements in multiple sectors including energy, finance, and healthcare [2]. - The establishment of the innovation center in Shenzhen is expected to attract more enterprises within the Open Harmony ecosystem, further developing the application software ecosystem [3].

要说全球开源大模型生态圈里，谁最让人"魂牵梦绕"？ 阿里，当仁不让。 就在上周，继DeepSeek和"AI教母"李飞飞之后，英伟达也相中阿里了。除了在最新的"混合推理模型"千 问3宣布开源当日，火速官宣接入适配后，5月9日，英伟达还 开源了全新的代码推理模型Open Code Reasoning （后续简称：OCR），包括7B、14B、32B三种尺寸， 基础模型用的都是通义千问。 在LiveCodeBench 评测中，成功超越Open AI 公司o3-mini和o1模型的英伟达OCR-Qwen-32B-Instruct模 型，正是基于Qwen2.5-32B微调形成的。 在通义千问已经迭代至3.0版本，模型性能再度突破的当下，英伟达居然还基于上一代千问模型做出了 比肩全球一流水平的模型，让人不禁想问，千问到底还有多少隐藏潜力待各方解锁？ DeepSeek、李飞飞后，英伟达也相中了"通义千问" 目前，英伟达开源的OCR系列模型的代码及数据集，已公开分享至全球最大AI开源社区Hugging Face平 台上，供开发者们免费浏览学习。 其中，英伟达OCR-Qwen-32B-Instruct在LiveCodeBench ...

新大陆 20250512 摘要 • 新大陆与中科软国际合作，推动开源鸿蒙在金融场景的适配，包括 POS 机 具等金融终端产品；与电科院北京分公司合作，推动开源鸿蒙在煤炭行业 的应用，实现设备和软件一体化应用。 • 新大陆自动识别完成自研解码芯片与开源鸿蒙系统的适配，实现国产化替 代；工业级读码器产品通过开源鸿蒙兼容性测试，实现从芯片、模组到设 备端的全面适配；与华为签署开源鸿蒙生态使能合作协议，开发自主可控 的工业领域解决方案。 • 新大陆在网证网号领域与公安部一所合作开发解密卡芯片及解密算法，是 目前唯一一家拥有公安部一所解密授权的加密芯片供应商，为其网络认证 号码业务提供坚实基础。 • 《国家网络身份认证公共服务管理办法（征求意见稿）》推动网证网号发 展，传统身份证识别设备替换为网证网号适配设备市场规模约 400 亿至 800 亿元，软件端市场规模达千亿元以上，新大陆作为核心受益公司，有 望占据 20%-30%的市场份额。 • 新大陆与公安部一所独家合作，为其提供二维码解决方案，已在景区、酒 店、机场等场景实现商业模式落地，但大规模推广仍需政策推动。 Q&A 新大陆公司在鸿蒙操作系统方面的布局和进展如何？ ...