Workflow
类脑计算
icon
Search documents
高密低耗“生物—电子”存储设备研发成功
Ke Ji Ri Bao· 2026-02-26 23:04
团队表示,随着人工智能技术对算力和能效的要求不断攀升,像这样兼具低功耗与高存储性能的新型设 备将变得愈发关键。未来他们将继续优化这一技术路径,探索更多生物启发电子应用的新可能。 【总编辑圈点】 这项成果最重要的启示,在于它揭示了生物材料与半导体融合的可能方向。通过在分子层面实现DNA 与钙钛矿的精准耦合,科学家正在搭建起一条"生物+电子"的混合通道,让信息存储技术从传统半导体 工艺,向生物启发式设计演进。未来,我们使用的存储器或处理器,可能不再纯粹由硅基材料构成,而 是融入经过工程改造的生物分子结构,从而在纳米尺度上实现更高效、更智能的信息存储与处理。 美国宾夕法尼亚州立大学研究团队将合成DNA与半导体材料钙钛矿结合,研发出一种全新的"生物—电 子"存储设备,将DNA海量信息存储能力与钙钛矿出色的电子性能合二为一,存储密度更高,功耗仅为 传统存储设备的百分之一。这项突破有望颠覆低功耗存储设备的传统设计思路,相关成果发表于新一期 《先进功能材料》杂志。 DNA作为自然界最强大的信息存储系统,每克可存储高达2.15亿GB的数据。如果能将这一潜力引入电 子领域,将带来能效更高的数据中心、处理能力更强的计算设备。而这 ...
绿色算力 优刻得:构筑数字世界的"新基建"
通过与类脑产业生态的深度协作,成为连接"芯片—算法—应用—场景"的重要基础设施节点。 放眼更 长周期,随着人工智能、大模型及类脑计算等技术持续演进,算力已不再只是支撑单一应用的工具,正 逐步演变为数字经济时代的关键生产要素。在新一轮以绿色、高效、可持续为导向的数字"新基建"进程 中,优刻得这类以算力底座为核心的云服务商,正在成为支撑"数字世界"稳步运行的基础性力量。 (调研小组成员:谭飞、刘礼文、赵平) ◎记者 刘礼文 一手依托上海青浦与内蒙古乌兰察布两大自建智算中心,构建起超过万P规模的算力资 源池;一手基于智能算力平台"孔明",实现了算力资源的高效统一调度与精细化管理——科创板"云计 算第一股"优刻得在数字基建领域深耕多年,"算力底气"十足。 近日,优刻得董事长兼CEO季昕华在接 受上海证券报记者采访时表示,公司将持续发挥云计算与AI基础设施优势,携手产业各方在数字中国 建设方面作出更大贡献。 一朵"中立云"赋能千行百业 走进优刻得位于上海杨浦区的办公楼展厅,从游 戏、电商、在线教育等新兴产业,到AIGC、大模型、脑机接口等未来产业,一面贴有近百个知名品牌 logo的合作伙伴展示墙,诠释出"AI+智算云" ...
中科院AI芯片新路径登Science!铁电材料新结构突破存储密度极限
量子位· 2026-01-24 07:33
Core Viewpoint - The research from the Institute of Physics, Chinese Academy of Sciences, reveals a significant breakthrough in ferroelectric materials, specifically in the atomic-level "one-dimensional charged domain walls" within zirconia, laying a new physical foundation for next-generation artificial intelligence devices [1][4]. Group 1: Breakthrough in Ferroelectric Materials - The research team confirmed that the width and thickness of these domain walls are only the size of a single crystal cell, confined within a two-dimensional polar layer, achieving the physical limit of size [3][10]. - This discovery unveils the charge screening mechanism of oxygen ions' "self-balancing," breaking through the traditional storage density bottleneck of two-dimensional domain walls [3][22]. - The unique "polarization-ion" coupling transport characteristics of this one-dimensional structure open new physical pathways for constructing high-energy-efficient brain-like computing chips and AI devices [4][24]. Group 2: Characteristics of Ferroelectric Materials - Ferroelectric materials are defined as a class of crystalline materials with spontaneous polarization, where the polarization direction can be reversed by an external electric field [6]. - These materials can be visualized as filled with tiny "electrical compasses" that indicate the direction of charge separation rather than geographical north and south [6][7]. - The concept of ferroelectric domains is introduced, where these "compasses" align in groups to minimize energy, forming domain walls that separate different polarization regions [8][9]. Group 3: Unique Structure of Domain Walls - The research team discovered that in zirconia, the originally broad two-dimensional "walls" are compressed into atomic-scale one-dimensional "lines" due to the material's unique sub-cell layered structure [11][12]. - These one-dimensional structures are not ordinary "walls" but special charged domain walls, categorized as "head-to-head" and "tail-to-tail" [12][13]. - The stability of these high-energy structures, which are typically unstable, is maintained through the introduction of high concentrations of point defects acting as "charge glue" [29][30]. Group 4: Implications for Data Storage and Ion Transport - The theoretical data storage density using these atomic-level one-dimensional domain walls can reach 20TB per square centimeter, equivalent to storing 10,000 HD movies on a device the size of a postage stamp [24]. - The material exhibits superior ionic conductivity at room temperature, outperforming traditional solid electrolytes like yttria-stabilized zirconia (YSZ), transforming it into a "highway" for ion transport [22][23]. - The research highlights a precise "charge compensation mechanism" that allows the one-dimensional domain walls to exist stably while facilitating efficient ionic conduction [36].
一间实验室,何以攀登“脑机接口”高峰?
Xin Lang Cai Jing· 2026-01-17 07:28
这款由脑机海河实验室团队研发的手指穿戴设备,被称为"神工—灵犀指"。据介绍,用意念控制该设备,可以配合人手完成辅助抓握,唤醒患 者的功能障碍中枢与外周神经系统,助力手部运动功能康复。脑机接口通过采集脑部神经信号,转换成特定指令。该技术能够在大脑与外部设 备之间创建连接,实现"脑"与"机"之间的信息交换。 天津大学副校长、脑机海河实验室主任明东说,实验室团队聚焦无创脑机接口领域,持续推动科技创新与产业化落地,在航空航天领域,设计 开发五代空间站在轨脑机交互系统,完成"太空脑机接口实验"。 "现在想象,右手腕长出了一根'第六指',用它去抓握眼前的小球。"在天津大学脑机交互与人机共融海河实验室(以下称脑机海河实验室) 里,科研人员王壮使用"读心术",操作机械指稳稳抓住桌子上的小球。 据了解,脑机海河实验室将聚焦基础元件、航空航天、康复医疗等多个领域继续深挖技术潜力,推动相关技术尽快落地应用。 在培养室中,小小的一片人工培育"大脑"不断接收并反馈信号,"指挥"着机械小车在沙盘上躲避障碍物。 作为天津市7家聚焦前沿科技领域的海河实验室之一,由天津大学牵头建设的脑机海河实验室2023年3月27日正式揭牌。"实验室聚焦脑机 ...
四中全会精神在基层丨一间实验室,何以攀登“脑机接口”高峰?
Xin Hua Wang· 2026-01-16 12:17
新华社天津1月16日电 题:一间实验室,何以攀登"脑机接口"高峰? 新华社记者郭方达、梁姊、杨文 "现在想象,右手腕长出了一根'第六指',用它去抓握眼前的小球。"在天津大学脑机交互与人机共融海河实验室(以下称脑机海河实验室)里, 科研人员王壮使用"读心术",操作机械指稳稳抓住桌子上的小球。 这款由脑机海河实验室团队研发的手指穿戴设备,被称为"神工—灵犀指"。据介绍,用意念控制该设备,可以配合人手完成辅助抓握,唤醒患者 的功能障碍中枢与外周神经系统,助力手部运动功能康复。脑机接口通过采集脑部神经信号,转换成特定指令。该技术能够在大脑与外部设备之 间创建连接,实现"脑"与"机"之间的信息交换。 天津大学副校长、脑机海河实验室主任明东说,实验室团队聚焦无创脑机接口领域,持续推动科技创新与产业化落地,在航空航天领域,设计开 发五代空间站在轨脑机交互系统,完成"太空脑机接口实验"。 在培养室中,小小的一片人工培育"大脑"不断接收并反馈信号,"指挥"着机械小车在沙盘上躲避障碍物。 有别于过去使用生物脑作为研究对象,这项名为片上脑机接口的技术,作为脑机接口领域的又一新兴技术分支,在脑机海河实验室走出了关键一 步。 作为片上 ...
四中全会精神在基层|一间实验室,何以攀登“脑机接口”高峰?
Xin Hua She· 2026-01-16 11:15
新华社天津1月16日电 题:一间实验室,何以攀登"脑机接口"高峰? 新华社记者郭方达、梁姊、杨文 "现在想象,右手腕长出了一根'第六指',用它去抓握眼前的小球。"在天津大学脑机交互与人机共融海 河实验室(以下称脑机海河实验室)里,科研人员王壮使用"读心术",操作机械指稳稳抓住桌子上的小 球。 作为片上脑机接口技术团队负责人,天津大学教授李晓红说,多领域的跨界交叉正不断催生片上脑机接 口技术等前沿成果。如今,李晓红团队正将精力转向技术优化和落地应用。她表示,未来片上脑机接口 技术有望推动混合智能、类脑计算等前沿科技领域的发展。 始于创新,不止于创新,更多产业化领域的实践正在稳步推进。脑机海河实验室推动成立目标规模百亿 元的脑机接口产业发展专项基金,实验室孵化的天开燧世(天津)智能科技有限公司目前已通过国家高 新技术企业认定,并获国家科技型中小企业称号。 "'十五五'规划建议中明确提出'推动科技创新和产业创新深度融合',我们也将科技成果转化落地作为实 验室工作的重要一环。"倪广健认为,未来实验室将继续发挥新型研发机构的黏合剂作用,以创新策源 带动产业"蝶变"。(完) 为推动技术创新成果落地转化,脑机海河实验室在建 ...
我科学家研发出多物理域全新计算系统
Huan Qiu Wang Zi Xun· 2026-01-15 01:21
Core Insights - The research team from Peking University has developed a novel multi-physical domain fusion computing system that enhances Fourier transform capabilities, achieving nearly a fourfold increase in computational power, which opens new possibilities in fields like embodied intelligence and communication systems [1][2] Group 1: System Performance - The new system integrates two distinct types of memristors, achieving remarkable parallel computing efficiency [2] - In brain-computer interface experiments, the system demonstrated low latency and high accuracy in EEG typing, with a single classification accuracy of 99.2% [2] - The processing throughput of the system increased from approximately 130 billion operations per second to 504.3 billion operations per second, reaching 96.98 times the performance of existing dedicated fast Fourier transform hardware [2] Group 2: Significance and Future Implications - This achievement signifies a shift from "algorithm-driven design" to "physics principle-driven" approaches, transforming mathematical operations into a more efficient process akin to natural evolution [2] - The new computing framework is expected to overcome the challenges of operator spectrum expansion for post-Moore devices, enabling support for multiple computing methods and laying a solid foundation for future advancements in edge intelligence, brain-like computing, and optoelectronic integration systems [2]
深圳脑机星链深度参与《类脑计算系统与训练算法评估指南》团体标准制定
Quan Jing Wang· 2026-01-12 12:42
Core Viewpoint - The release of the "Guidelines for Evaluation of Brain-like Computing Systems and Training Algorithms" marks a significant step in the standardization of brain-like computing technology in China, with Shenzhen Brain-Machine Starlink Technology Co., Ltd. playing a crucial role in its development alongside top universities and research institutions [1][2]. Group 1: Company Overview - Shenzhen Brain-Machine Starlink Technology Co., Ltd. focuses on artificial intelligence and aims to establish a comprehensive development system from basic research to technology transfer and industrial implementation in the brain-machine interface field [2]. - The company integrates brain-machine interface applications with neural modulation technologies to accelerate the development of diagnostic products for neurological diseases such as Alzheimer's, autism, and depression [2]. Group 2: Industry Context - The brain-machine interface technology is on the verge of explosive growth, with significant potential across various sectors including medical rehabilitation, consumer electronics, education, entertainment, and smart manufacturing [3]. - A market research report predicts that the global brain-machine interface market could exceed hundreds of billions of dollars by 2030, with the medical and health sector being the primary growth area [3]. Group 3: Collaborative Efforts - The collaboration between Shenzhen Brain-Machine Starlink and leading domestic universities and research institutions in standard formulation highlights the company's leading position and strong technical foundation in the brain-like computing field [4]. - The company has announced a partnership with leading immunodiagnostic firm Yihuilong to advance the clinical application and industrialization of brain-machine interface technology in various diagnostic fields [4].
新华财经早报:1月11日
Xin Hua Cai Jing· 2026-01-11 00:44
Group 1 - The National Internet Information Office has drafted the "Regulations on the Collection and Use of Personal Information by Internet Applications" and is seeking public opinion, emphasizing minimal impact on personal rights and the necessity of data collection for services [1] - The revised "Market Supervision Complaint and Reporting Handling Measures" will take effect on April 15, 2026, aiming to enhance the quality of complaint handling and regulate malicious claims [1] - The National Medical Insurance Administration has initiated a pilot program for the "Personal Medical Insurance Cloud," set to explore a new model for smart medical insurance management from February to December this year [1] Group 2 - The Supreme People's Procuratorate reported that from 2024 to November 2025, 191 individuals were prosecuted for financial fraud, with a 21% increase in prosecutions from January to November 2025 compared to the previous year [1] - The China Securities Regulatory Commission has imposed penalties on Jiang Wei for insider trading, resulting in the confiscation of illegal gains amounting to 4,709,741.41 yuan and a fine of 14,629,224.23 yuan [1] - The "China Ice and Snow Tourism Development Report (2026)" forecasts that the number of participants in ice and snow tourism will reach 360 million by the winter of 2025-2026, generating an expected revenue of 450 billion yuan [1] Group 3 - QDII funds are set to receive policy support, with adjustments required in the use of QDII quotas for public and private products, aiming for at least 50% of the adjustments to be completed by the end of 2026 [1] - The Beijing Electric Power Trading Center reported that inter-provincial electricity trading volume in the State Grid's operating area reached 1.67 trillion kilowatt-hours in 2025, a 10% year-on-year increase [1] - China has applied for frequency resources for over 200,000 satellites, indicating a strategic elevation of satellite frequency resource applications [1]
我国创出全新计算架构提升算力
Mei Ri Jing Ji Xin Wen· 2026-01-10 15:24
Core Insights - The research team from Peking University has developed a novel multi-physical domain fusion computing architecture that utilizes post-Moore new devices to support Fourier transforms, enhancing computing power by nearly 4 times [1] Group 1: Technology and Innovation - The new architecture allows for the translation of complex signals such as sound and images into frequency language, which is fundamental in scientific and engineering applications [1] - This advancement opens new possibilities in fields such as embodied intelligence, edge perception, brain-like computing, and communication systems [1] Group 2: Publication and Recognition - The research findings were published on the 9th in the journal Nature Electronics, highlighting the significance of the work in the academic community [1]