Group 1: Industry Developments - Nvidia CEO Jensen Huang visited China for the third time this year, emphasizing the importance of the Chinese market and the transformative impact of AI across various industries [2] - Meta CEO Mark Zuckerberg announced plans to invest several hundred billion dollars in AI, aiming to establish industry-leading computational capabilities with the launch of the Prometheus supercluster by 2026 [3] - Huawei plans to initiate commercial trials for L3 autonomous driving by 2025 and aims for large-scale commercialization by 2027, potentially outpacing Tesla [4] Group 2: Company News - Tesla will officially enter the Indian market by opening its first showroom in Mumbai on July 15, facing high import tariffs of approximately 70% [8] - Xiaoma Zhixing has begun mass production and road testing of its seventh-generation autonomous Robotaxi, aiming to expand its fleet to 1,000 vehicles by the end of 2025 [9] - Synopsys received conditional approval from China's market regulator for its acquisition of Ansys, with stipulations to divest certain business segments to prevent anti-competitive effects [10] Group 3: Financial Performance - Xinyisheng expects a significant increase in net profit for the first half of the year, projecting a growth of 327.68% to 385.47%, driven by AI-related investments [17] - Wentai Technology anticipates a net profit increase of 178% to 317% for the first half of 2025, benefiting from a recovering semiconductor market [18] - Silan Microelectronics forecasts a turnaround with a net profit of 235 million to 275 million yuan for the first half of 2025, supported by full production at its chip manufacturing lines [19]

访问36氪出海网站 letschuhai.com ，获取更多全球商业相关资讯。 阿联酋 迪拜道路与交通管理局与小马智行签署谅解备忘录，将启动自动驾驶汽车试点测试 7月6日消息，迪拜道路与交通管理局 (RTA) 与小马智行签署谅解备忘录，将启动自动驾驶汽车试点测试。今年5月小马智行宣布与 RTA 达成战略合作，将在 当地落地 Robotaxi 车队。根据规划，此次合作将分阶段实施：2025年启动测试运营，2026年推进全无人商业化运营，并实现与迪拜当地地铁、有轨电车以 及海上航线等多元交通网络的无缝衔接。（IT之家） 7月7日消息，据阿布扎比报道，联邦身份、公民、海关和港口安全局（ICP）、证券和商品管理局（SCA）和虚拟资产监管局（VARA）发表联合声明，否 认了有关阿联酋向数字货币投资者发放黄金签证的不实报道。ICP 澄清黄金签证不包括数字货币投资者，而适用于房地产投资者、企业家等符合条件的类 别。SCA 强调其致力于规范金融业和证券服务，VARA 否认了向迪拜虚拟资产投资者发放黄金签证的说法，并警告投资者应从可靠官方渠道获取信息，避 免误传或欺诈。三家监管机构共同敦促公众和投资者谨慎行事，并参考政府官方网 ...

A公司主力方案，依然采用多感知+预测两阶段方案（map、od、occ等）。作为主打中低成本的供应 商，虽然A公司一直在想办法控制成本量产端到端和后续大模型方案，但高阶研发迭代贵，量产规模 上不来，风险还是挺大，这和新势力自产自销还是有所不同。 A公司因为早期设定了低成本方案和稳步切入的路线，一直没有大规模投入端到端方案。许多已经稳 定的场景，其实规则方案已经满足要求了，A公司就是这样。加上前面的和主机厂的合作方案多基于 模块化，供应链的稳定比较重要，主机厂会尽可能选择兼容现有生态的方案。端到端需要海量高质 量数据训练，而模块化方案可通过规则+小数据量调整适配区域差异，这也是其他tier1没有完全更改 量产方案的原因。加上早期积累了大量相关的训练数据，貌似模块化方案，是成本最低的。除非端 到端能验证，多个领域吊打模块化方案，否则许多tier1短期内还是会坚持已有量产方案。 深耕智驾职场圈，了解行业二三事，欢迎添加自动驾驶之心专属职位分析师，微信：Remix-clover， 和我们做进一步交流！ 许多公司和A公司面临一样的窘境，这也造就了技术研发团队的格局与不同...... ...

2025上半年已经过去了，上半年自动驾驶发展的怎么样？业内落地量产在做什么？学术界在关心什么技 术方向？转行具身还是深耕智驾？全都是问题！ 为此，自动驾驶之心采访了国内外数十位学术界和工业界的诸位大佬，和他们一起聊聊对这些问题的看 法，希望对大家有帮助~ 1. 当下前沿技术的发展路线是否已经成熟并适合量产？有哪些难点需要攻克？ 2. 如何看待新兴的技术方向：VLA/VLM、扩散模型、闭环仿真、强化学习、端到端自动驾驶、世界模 型等等 3. 未来自动驾驶技术还有哪些值得探索的新方向？ 4. 深耕自动驾驶 or 投身具身智能，怎么选？ 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 SJTU 张志鹏 BEV，Occ， E2E， VLA， 这几个词是从2022年到2025年自动驾驶中热词的一个变迁。 BEV和Occ的引入某种程度上都是因为特斯拉推崇纯视觉方案，然后借鉴Lidar和机器人的一些建模方式 去表征整个驾驶环境。近期又出现了一些用3DGS建模3D环境的方式。但是整体来看，自动驾驶中感知 的场景表征的方式还是比较成熟了，各家厂商也验证了这种模式在实际中是走 ...

Market Overview - The Hang Seng Index rose by 0.9% last week, closing at 24,139 points, while the Hang Seng Tech Index increased by 0.6% to 5,248 points[1] - Weekly market turnover decreased by 1.7% to HKD 242.5 billion, with a net inflow of HKD 20.7 billion through the Stock Connect[1] - The market is experiencing significant rotation among sectors, with strong inflows into stablecoin concept stocks, brokerages, and biopharmaceuticals, indicating a positive investment cycle[1] Macroeconomic Insights - The latest FOMC minutes revealed that only a few Fed officials support a rate cut in July, with most concerned about inflation risks from new tariffs[2] - Initial jobless claims in the U.S. unexpectedly fell to 227,000, the lowest in seven weeks, indicating a resilient labor market[2] Sector Performance - The consumer sector saw declines, with stocks like Maogeping and Lao Pu Gold dropping between 5% and 9% last week[3] - The vocational education sector performed well, with New Higher Education and China Spring rising by 24.2% and 14.6%, respectively[3] Healthcare Sector - The Hang Seng Healthcare Index slightly declined by 0.05% after two weeks of gains, despite WuXi AppTec's stock surging by 10.5% following a positive earnings announcement[4] - WuXi AppTec expects a revenue of approximately RMB 20.8 billion for H1 2025, reflecting a year-on-year growth of about 20.6%[4] Energy and Utilities - The renewable energy and utilities sectors experienced slight fluctuations, with notable increases in stocks like Xinyi Solar and GCL-Poly Energy, rising by 2.4% and 9.8%, respectively[5] - Environmental stocks also saw gains, with Everbright Environment and Beijing Enterprises Water rising by 3.8% and 5.7%[5] Company-Specific Updates - Tianlun Gas is expected to return to profit growth starting FY25, with a projected CAGR of 12.0% for net profit from FY24 to FY27[6] - The company aims to increase its dividend payout ratio from 30.0% in FY23 to 35.0% by FY25, indicating a commitment to shareholder returns[11]

来源：环球网 【环球网智驾综合报道】7月14日消息，近日，第十七届国际交通技术与设备展览会盛大举行，滴滴自 动驾驶携新一代L4级前装量产车型及最新技术成果惊艳亮相，成为展会焦点。此次展出的新一代L4级 前装量产车型，搭载了滴滴自动驾驶全新一代硬件平台，传感器总数达33个，强大的硬件配置使其在展 会现场备受关注。滴滴自动驾驶明确表示，将持续加大研发投入，与产业伙伴携手推进技术创新，致力 于为乘客提供更便捷多样化的出行服务。这一动作不仅展现了其在自动驾驶领域的最新突破，更释放出 加速规模化落地的强烈信号。 回顾滴滴自动驾驶的发展历程，其布局起步较早。早在2016年，滴滴便启动了相关技术布局，由联合创 始人张博亲自带队组建团队，专攻L4级自动驾驶系统。2019年8月5日，滴滴将自动驾驶业务升级为独 立公司，聚焦核心技术研发与商业拓展，由此进入专业化运营阶段。2020年，滴滴在上海首次面向公众 开放自动驾驶服务，实现了从实验室到真实道路的关键跨越，为Robotaxi商业模式验证积累了宝贵经 验。 为加快量产和落地，行业逐渐形成"主机厂 + 智能驾驶技术公司 + 出行平台"的深度协同模式，滴滴自 动驾驶也积极投身其中。 ...

点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 自驾领域的许多公司25年校招陆续开启了！比如地平线、滴滴、元戎启行等。整体来看，岗位依然充裕， 感知、规控、端到端、大模型需求还是蛮多的。许多公司也增加了一些具身智能相关岗位，自驾+具身概 念，逐渐融合！ 大家在投递的时候，一定在简历上突出自己的优势，面试准备做足。不要冒然面试，机会损失掉就太可惜 了。许多公司预计在7月底和8月初开始逐步安排技术面试和HR面试。 | 【2026届校招】智能驾驶域控制器硬件开发工程师 硬件序列 | | | --- | --- | | 北京地平线信息技术有限公司 上海市 | | | 发布时间:2025-07-04 招聘人数:若干 | 展开 √ | | 【2026届校招】感知后处理工程师（人机交互方向） 软件序列 | | | 北京地平线信息技术有限公司 北京市、上海市 | | | 发布时间: 2025-07-04 招聘人数: 若干 | 展开 √ | | 【2026届校招】智能驾驶中间件软件工程师 软件序列 | | | 北京地平线信息技术有限公司 北京市、上海市 | | | 发布时间 ...

Core Viewpoint - Pony.ai, known as the "global Robotaxi first stock," is facing scrutiny from short-seller Grizzly Research, which claims the company has fabricated data and is the worst performer in the autonomous taxi sector [1][3]. Company Overview - Pony.ai was founded in 2016 and is headquartered in Nansha District, Guangzhou. It launched its Robotaxi service in 2018 and is one of the first companies in China to obtain autonomous driving service permits in major cities like Beijing, Shanghai, Guangzhou, and Shenzhen. It successfully went public on NASDAQ on November 27, 2024, becoming the "global Robotaxi first stock" [3]. Allegations and Financial Performance - Grizzly Research's report cites serious allegations from apparent insiders, claiming that Pony.ai has actively falsified data related to its autonomous driving software algorithms, with management allegedly aware of and attempting to cover up these actions. This behavior is deemed reckless and dangerous for a company developing autonomous driving technology [3][4]. - The financial situation of Pony.ai appears dire, with worsening conditions post-IPO. The company heavily relies on related-party transactions. Revenue figures from 2022 to 2024 show a growth trend: $68.39 million, $71.90 million, $75.03 million, and $13.98 million for the first quarter of 2024. However, net losses have increased: $148 million, $125 million, $275 million, and $37.4 million for the first quarter of 2024 [3][4]. Research and Development Costs - The continuous losses are attributed to high costs, particularly in technology, product development, and marketing, with R&D expenses consistently exceeding $100 million annually. Specific R&D expenditures from 2022 to 2024 were $154 million, $122 million, and $240 million [4]. Competitive Position and Regulatory Challenges - Grizzly's report suggests that Pony.ai lags behind competitors like Baidu Apollo and WeRide, requiring significant time and funding to catch up. The company lost its permit for autonomous vehicle testing in California due to an incident a few years ago and faced a similar situation in China in May, leading to a temporary service disruption. Currently, Pony.ai is only permitted to conduct autonomous vehicle testing with a safety driver present in the U.S. [4]. Market Pressure - Following the expiration of the lock-up period for shares, Grizzly believes that Pony.ai's business and stock price will face increasing pressure. As of July 11, 2024, the stock price was $12.77 per share, reflecting a slight decline of 0.93% [6].

Core Viewpoint - The Hong Kong Stock Exchange has launched a new policy called "Tech Company Special Line," providing a confidential listing channel and lowering the threshold for specialized technology and biotechnology companies, attracting more tech firms to consider listing in Hong Kong [1][2]. Group 1: Applicable Entities - The policy is aimed at specialized technology companies (e.g., AI, chips, new energy) and biotechnology companies (e.g., innovative drugs, medical devices), particularly those in early stages or with non-commercialized products [3]. - Core thresholds include industry attributes defined by the Hong Kong Stock Exchange under "Specialized Technology" (Chapter 18C) or "Biotechnology" (Chapter 18A) [4][6]. Group 2: Self-Assessment and Application Process - Companies must assess if they meet the criteria by checking the "Special Line" page on the Hong Kong Stock Exchange website and downloading the self-assessment form [8]. - If uncertain, companies can fill out the inquiry form and send it to the Hong Kong Stock Exchange for preliminary feedback within one week [9]. Group 3: Confidential Submission Process - The first step involves signing a Non-Disclosure Agreement (NDA) with the Hong Kong Stock Exchange to ensure confidentiality of submitted materials [11]. - Companies must submit a "confidential version" of their materials in a specified format [13]. - The review phase will take 30 days, focusing on technical feasibility and compliance [14]. Group 4: Exclusive Services of the "Tech Company Special Line" - Companies can receive one-on-one guidance from the Hong Kong Stock Exchange experts, including interpretations of listing rules and fundraising strategies [16]. - Eligible companies can benefit from a fast-track review process, reducing the review period to 30 days [17]. - Flexible equity design allows founders to retain control without additional proof of "innovation" [18]. Group 5: Common Pitfalls to Avoid - Companies should provide clear descriptions of their technology and avoid vague claims without supporting evidence [21]. - Transparency in related party transactions is crucial to avoid compliance issues [22]. - Establishing a diverse investor base is important to strengthen investor relations [25]. Group 6: Post-Listing Compliance - Continuous information disclosure is required, including updates on technology commercialization and major collaborations [27]. - Companies are encouraged to maintain market value by releasing quarterly research updates and engaging with analysts [28]. - A green channel for refinancing allows specialized companies to issue new shares through a simplified process [29]. Group 7: Comparison with Other Markets - The article compares the listing requirements and processes of the Hong Kong Stock Exchange with those of the A-share market and NASDAQ, highlighting differences in profitability requirements, review periods, and information disclosure levels [30].

以下文章来源于红色星际 ，作者红色星际科技 红色星际 . 某智驾公司从去年就开始了一轮融资，但是一直没融到钱。原因也不复杂，相比拿到手的量产项目来说，估值太高了，快接近头部智驾公司的估值了，但是手里的量产 项目却没几个，所以融资很艰难。 让更多人，更深入地了解自动驾驶行业！ 作者 | 钟声 来源 | 红色星际 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 本文只做学术分享，如有侵权，联系删文 据悉，某智驾公司和某头部车企探讨一个非常"风骚"的融资交易。 论论论论论论论论文文文文文文文文辅辅辅辅辅辅辅辅导导导导导导导导来来来来来来来来啦啦啦啦啦啦啦啦 知知知知知知知知识识识识识识识识星星星星星星星星球球球球球球球球交交交交交交交交流流流流流流流流社社社社社社社社区区区区区区区区 近4000人的交流社区，近300+自动驾驶公司与科研结构加入！ 涉及30+自动驾驶技术栈学习路线，从0到一带你入门 自动驾驶感知（大模型、端到端自动驾驶 、 世界模型 、 仿真闭环 、 3D检测 、 车道线 、 BEV感知 、 O ...