中央纪委国家监委网站 陆丽环 高校承担着培养社会主义建设者和接班人的重大使命，事关党和国家前途命运。各中管高校纪检监察机 构坚决贯彻落实中央纪委全会精神，坚持党性党风党纪一起抓，以严实举措强化正风肃纪，推动严的氛 围在中管高校不断形成和持续巩固。 深入贯彻中央八项规定精神，重拳纠治"象牙塔"内不正之风 2025年以来，清华大学纪委、监察专员办结合查办案件拍摄警示教育片《迷失的初心》《守住第一道防 线》，深入院系单位播放，覆盖"关键少数"、重点岗位3400余人，推动引导学校党员干部教师懂法纪、 明规矩、知敬畏、守底线。 落实联合审查调查机制，系统整治高校领域突出问题 高校是培养社会主义建设者和接班人的重要阵地，绝容不得腐败。中央纪委国家监委印发《关于加强中 管高校纪检监察机构与地方纪委监委联合审查调查工作的指导意见》。各中管高校纪检监察机构深入落 实意见要求，持续紧盯高校重点领域深化系统治理，坚决查处高校权力集中、资金密集、资源富集的重 点领域违纪违法问题，推动校园政治生态不断净化。 上海交通大学纪委、监察专员办在上海市纪委监委统一调度指挥下，建立联合审查调查组，通过健全地 校纪委书记碰头会、组长阶段沟通会、组 ...

Group 1 - The meeting between Suzhou's Party Secretary Fan Bo and Tsinghua University's Party Secretary Qiu Yong emphasized the importance of collaboration for high-tech innovation and quality development in Suzhou [1][2] - Suzhou aims to integrate education, technology, and talent development, creating a conducive environment for various talents to thrive [1] - Tsinghua University has a strong tradition of nurturing outstanding talents and has established a solid foundation for cooperation with Suzhou in areas such as innovation and talent cultivation [2] Group 2 - Both parties expressed a commitment to deepen practical cooperation in key areas such as core technology breakthroughs, talent training, and major platform construction [2] - The collaboration aims to provide strong support for achieving high-level technological self-reliance and innovation [2]

Investment Rating - The report does not explicitly state an investment rating for the industry Core Insights - The artificial intelligence (AI) sector continues to dominate investment, accounting for 88.1% of total financing in the technology field in 2025, an increase of 5.3 percentage points from 2024 [21] - The humanoid robot sector has emerged as a significant area of investment, capturing 16.8% of the total technology financing [28] - The report highlights the rapid evolution of AI technologies, particularly in reasoning models and multimodal models, which are reshaping the landscape of AI applications [36][38] Summary by Sections 2025 Technology Market Review - In 2025, the technology sector saw a total financing of 738.4 billion yuan across 644 events, with AI leading the way [21] - Key developments included the rise of humanoid robots and advancements in AI models, such as the release of various reasoning models by major companies [19][36] 2026 Technology Trend Outlook - The report anticipates continued growth in AI and humanoid robotics, with significant investments expected in these areas [28] - The integration of AI into various applications is expected to accelerate, driven by advancements in multimodal models and reasoning capabilities [36][38] Financing Events and Trends - The AI sector experienced 212 financing events exceeding 100 million yuan, representing 88.7% of all such events in the technology field [29] - The report notes a shift towards more substantial financing rounds, with seed and A rounds making up 67.7% of total financing events [31] Key Companies and Innovations - Notable companies in the AI sector include Zhiyuan AI, Minimax, and Yuzhi Technology, which have raised significant funds in recent rounds [32] - The report highlights the emergence of new technologies and models, such as DeepSeek and various multimodal models, which are expected to drive future growth [36][38]

曲钧天介绍，两款机器人均体现了"仿生智能"的前沿理念。另一款磁吸式柔性机器人模仿的是海洋深处 常见的蠕虫，适用环境是复杂的管道。"管道是能源运输与城市运维的核心基础设施，其隐蔽性、复杂 性给检测维护带来了挑战。传统刚性管道机器人结构复杂、控制难度大，在小管径、破损管道或动态水 流等复杂工况下适应性受限；现有软体机器人虽具备灵活性优势，但普遍存在流阻大、负载能力弱、难 以应对真实管道故障等问题。" 磁吸式柔性蠕虫机器人 团队创新性地采用星形三向锚定足与纤维增强波纹管主干设计，让机器人兼具柔性与高负载能力。"它 的重量仅120克，却能在垂直管道中推动750克、拉动309.5克的物体，还能适应0度至90度的管道倾 角。"他说，搭载摄像头后，机器人可在复杂管道内部完成巡检任务。 目前，团队正在对两款机器人开展持续的技术优化，推动它们在海洋环境探测、基础设施巡检等领域应 用。 拥有四只仿生鳍肢的海龟机器人，能在水陆交界的复杂区域自如漫步；外表像蠕虫的磁吸式柔性机器 人，可以推动重量是自己体重6倍的物体，在复杂管道内自由穿梭。清华大学海洋软体机器人与智能传 感实验室近日发布了两款海洋仿生机器人，为海陆两栖跨介质探测与水下 ...

摘要 ：近日，清华大学发布多批政府采购意向，仪器信息网特对其中的仪器设备品目进行梳理，统计出 28项仪器设备采购意向，预算总额达1.98亿元。 特别提示 微信机制调整，点击顶部"仪器信息网" → 右上方"…" → 设为 ★ 星标，否则很可能无法看到我 们的推送。 近日，清华大学发2 5项仪器设备采购意向，预算总额达1 . 9 8亿元，涉及高带宽实时示波器、等 温滴定微量热仪、多组分红外分析仪表、扫描探针显微镜控制器系统、芯片及微系统电子负载 测试仪等，预计采购时间为2 0 2 6年1~3月。 详细采购清单： | | | | 1月 | | --- | --- | --- | --- | | 光刻胶 | 光刻胶。 | 200 | 2026年 | | | | | 1月 | | 晶圆衬底 | 晶圆衬底。 | 200 | 2026年 | | | | | 1月 | | 微系统集成综合FPGA 微系统集成综合FPGA开 | | 250 | 2026年 | | 开发验证系统 | 发验证系统。 | | 1月 | | 微系统设计硬件加速器 微系统设计硬件加速器。 | | | Comments of the United on t ...

格隆汇2月4日丨开创电气(301448.SZ)近日接受特定对象调研时表示，燕谷机器人公司跟清华大学合作 正常进展中，第一笔研发经费 1500 万元按照合作协议已支付清华大学。2026 年1 月初已召开联合研究 中心管委会第一次会议，通过了联合机构章程和管委会相关机制，同时决定批准了14 个项目立项，涵 盖具身智能感知、决策、控制、硬件设计及大模型基础研究等多个关键领域。 ...

据了解，机器学习与人工智能应用通常依赖高算力、高能耗的计算平台，而可穿戴传感器、脑植入器等 极端边缘设备需基于柔性基底、低成本元件完成本地基础计算。此前柔性中央处理器虽能实现通用计 算，但存在速度慢、功耗高的瓶颈，无法满足数据分类、机器学习等AI任务需求。 科技日报讯 （记者华凌）记者2月2日从清华大学获悉，该校集成电路学院教授任天令团队及合作者的 研究成果"FLEXI柔性数字存算芯片"，在柔性电子与边缘人工智能硬件领域取得突破，填补了当今高性 能柔性AI计算芯片技术空白。相关研究成果发表在国际期刊《自然》上。 任天令介绍，新研发的柔性AI芯片可直接在柔性基底上制造，兼具低功耗、低成本和高集成度优势。 研究团队通过工艺革新增加金属层数，突破了传统柔性电子难以支持复杂芯片互联的瓶颈；创新采用数 字"存内计算"架构，在存储器内部完成数据处理，既消除了数据搬运的时间与能耗开销，又突破了"存 储墙"性能限制，在能效与可靠性等关键指标上的表现优于传统模拟存内计算方案。 实测数据显示，该芯片在折叠、卷曲状态下可稳定工作，经4万次反复折叠后计算能力仍保持稳定，并 具备良好的耐温、耐湿和抗光照老化能力。该芯片能够集成至可穿戴 ...

Core Insights - Tsinghua University has developed a fully flexible artificial intelligence chip, filling a gap in flexible electronics technology and providing dedicated, scalable, and low-power hardware support for AI applications [1][2] Group 1: Technology and Innovation - The flexible electronic devices are transforming applications in wearable medical technology, implantable neural recording, human-computer interaction, and the Internet of Things due to their ultra-thin, lightweight, customizable, and low-cost advantages [1] - The FLEXI chip, designed for edge intelligence acceleration, utilizes low-temperature polycrystalline silicon CMOS technology, combining lightweight, low-cost, and high energy efficiency [1] - FLEXI can integrate up to approximately 265,000 transistors and achieves stable, high-speed, and parallel dot product operations through a cross-layer collaborative optimization strategy [1] Group 2: Performance and Reliability - Experimental results indicate that the FLEXI chip can withstand over 40,000 bending cycles at a 1mm radius and 180° fold without significant performance degradation [2] - The chip demonstrates high reliability, excellent energy efficiency, high yield, and good scalability while maintaining stable, error-free operation under high-frequency computing, extreme mechanical stress, and accelerated aging conditions [2] - FLEXI exhibits over six months of long-term stability, laying a solid foundation for the application of flexible electronic devices in mobile healthcare, embedded intelligence, and other edge computing scenarios [2] Group 3: Academic Recognition - The research findings were published in the prestigious journal Nature on January 28, under the title "A Flexible Digital In-Memory Computing Chip for Edge Intelligence" [2]

来源：科技日报微信 来源：科技日报 图片由清华大学集成电路学院提供 记者 华凌 实测数据显示，该芯片在折叠、卷曲状态下可稳定工作，经4万次反复折叠后计算能力仍保持稳定，并 具备良好的耐温、耐湿和抗光照老化能力。其最小尺寸芯片制造成本仅0.016美元，能够集成至可穿戴 设备，利用心率、呼吸频率、体温等生理信号实现人体日常活动识别。 1月28日，清华大学集成电路学院任天令教授团队及合作者的研究成果"FLEXI柔性数字存算芯片"正式 发表于国际顶级期刊《自然》，标志着我国在柔性电子与边缘人工智能硬件领域取得重要突破，填补了 高性能柔性AI计算芯片的技术空白。 专家点评指出，该技术填补了柔性电子领域AI专用计算硬件的空白。未来通过新型半导体材料应用、 功率门控技术优化等，有望进一步提升性能。若能持续优化生产良率与芯片尺寸，将推动可穿戴健康设 备、物联网终端等领域的产业升级与技术革新。 新研发的柔性AI芯片采用CMOS低温多晶硅（LTPS）工艺，可直接在柔性基底上制造，兼具低功耗、 低成本和高集成度优势。研究团队通过工艺革新增加金属层数，突破了传统柔性电子难以支持复杂芯片 互联的瓶颈；创新采用数字"存内计算"架构，在存 ...

Core Insights - The 2026 Beijing Youth "Outstanding" Innovation Winter Camp has commenced, focusing on cultivating innovative talents in various fields such as artificial intelligence, quantum computing, life sciences, and aerospace [1][2] - The program aims to create a collaborative education model involving high schools, universities, and research institutions, facilitating early identification and continuous development of innovative students [1][2] Group 1: Program Structure - The winter camp consists of 11 thematic camps that will be held during the winter break, guided by experts and academicians [1] - The camps are designed to stimulate interest and potential in students, providing a sustainable and structured approach to talent development [1] Group 2: Educational Approach - Students will engage in project-based learning, enhancing their interdisciplinary innovation capabilities through real-world problem-solving [1] - In the "Extreme Intelligence Challenge" AI camp, students will learn about AI technology principles and applications, including practical development and safety measures [1] Group 3: Specific Themes and Activities - The "Mathematics Talent" camp will focus on the intersection of mathematics and finance, allowing students to experience the entire process from data organization to strategy backtesting [2] - The program includes opportunities for students to participate in investment application projects, applying mathematical knowledge to real-life scenarios [2] Group 4: Institutional Support and Future Plans - The Beijing Municipal Education Commission emphasizes the importance of the winter camp in promoting educational excellence and building connections between different educational levels [2] - There are plans to deepen collaboration with technology companies and explore a dual-teacher model involving scientists and engineers to enhance STEM education [2]