AI 资讯 · 2026-07-21

作者｜包永刚编辑｜林觉民三年前"百模大战"时，一边是英伟达GPU一卡难求、价格一路水涨船高，另一边却是许多国产AI芯片很难被大模型公司充分利用。看到AI产业巨大的想象力，也看到国产算力没有被充分释放的商业价值，夏立雪意识到，团队过去十多年的技术积累终于迎来了产业机会。在导师汪玉的鼓励下，他与一群同样来自清华大学电子工程系的师兄弟共同创立无问芯穹，目标是成为打通M个模型和N款芯片的新技术范式平台，让每一块芯片都能发挥最大的商业价值。“凭借多年的经验，创业之初我们就不只看训练需求。我们预判模型很快会进入大规模应用阶段，于是先从异构训练切入，同时积累面向大规模推理的基础设施能力。”无问芯穹创始人兼CEO夏立雪告诉雷峰网，“虽然有所预判，但AI落地的速度还是比我们预想得更快。”随着推理需求快速爆发，以及Agent开始走向落地，夏立雪和团队在2024年底看清楚趋势之后就迅速明晰了商业策略。外界看到这家成立仅三年的公司，对外的产业表达在变化：从连接M种模型和N种芯片，到Token工厂，再到今年WAIC提出的“前店后厂一中心”的Agentic Infra战略布局。看起来，无问芯穹几乎每年都在变化，但

36氪获悉，恒指收跌0.04%，恒生科技指数涨1.32%；半导体、硬件设备、软件服务板块领涨，智谱涨近37%，华虹宏力涨超17%，MINIMAX涨超15%，立讯精密涨超11%；石油石化、煤炭、银行板块跌幅居前，山东墨龙跌超3%，农业银行、中煤能源跌超2%；南向资金净买入71.67亿港元。

“如果全部算力都用于推理，曙光8000将是中国最大的Token工厂，能支撑当前全国5%～10%的Token需求。”在一次采访中，中科曙光公司高级副总裁李斌的换算，让现场嘉宾对国产十万卡集群的体量有了最直观的感受。7月，光合组织2026智能计算应用大会期间，首台全国产十万卡量级AI超集群系统——曙光8000（登峰）正式落成，一周后真机在世界人工智能大会（WAIC）正式亮相。据了解，这也是国内首台明确采用“超智融合”技术路线的系统。根据曙光8000（登峰）公布的最新运行数据，集群正式上线市场需求火爆，首周即实现应用满载运行，目前日均处理作业数突破15万个，单日处理作业峰值超过50万个。落成即投用，投用即满载。从万亿级参数大模型的训练和高通量推理，到超300余个重点科学工程计算应用，再到AI4S（AI for Science）的前沿探索，曙光8000承担的任务量被迅速拉满的同时，还为下一代十万亿参数大模型预留了空间。 但挑战随之而来。当建设成本与运营复杂度呈指数级攀升，这套在2024年底完成研制并在接下来一年多时间内完成批量部署的系统，究竟如何撑起十万卡的野心？ 十万卡集群的工程硬仗故事的开

7月19日，上海仪电在2026世界人工智能大会（WAIC）上正式发布“智算云赋能·标杆应用场景成果”，系统展示了其自主研发的仪电智算云YiCloud在金融、制造、医疗三大国家人工智能应用中试基地的规模化落地成效，全方位呈现“算力+模型+场景”全栈服务能力。金融领域|深度支撑金融大模型生态构建助力普惠金融风控新标杆在中国银联承建的金融领域国家人工智能应用中试基地建设过程中，仪电智算云YiCloud凭借垂类模型训推一体化能力，为中国银联联合17家机构共建的百亿级金融语料及50万条微调数据提供稳定算力支撑。基于全国产化训练的双基座模型，已形成L1垂域矩阵，幻觉率降低30%，安全合规能力提升25%，在国内主流模型中位列第一，复杂任务成功率超85%。目前，该平台已成功支撑智能银行卡、跨境支付、全域信贷风控及云闪付AI版等应用落地。在金融风控实战中，仪电智算云平台助力拦截可疑交易超2亿元，其中“票根经济”场景累计处理22万张票根，拉动消费规模达2.5亿元，并赋能8000家商户“Nihao China”项目，注册用户达23万人，切实推动AI技术普惠实体经济。制造领域|赋能工业智能体落地全栈智能体重塑

7月21日，Momenta（6880.HK）股价涨超10%。消息面上，公司近日宣布正式获得深圳市智能网联汽车道路测试许可，将在当地开展L4级Robotaxi实地测试。近期，在2026世界人工智能大会上，享道出行宣布启动上汽Robotaxi量产定制车项目，搭载Momenta L4级智驾系统，计划2027年亮相。该车型将是上海首款高级别自动驾驶前装量产定制车，由上汽集团生产、Momenta提供智驾方案、享道出行负责运营，三方形成从研发生产到落地运营的完整闭环。量产车之外，技术方案的迭代也在同步推进。今年4月，Momenta宣布R7世界模型量产上车，基于超120亿公里行驶里程及1亿段关键场景数据，通过预训练、闭环仿真和强化学习进行优化，旨在提升复杂路况下的决策能力。Momenta表示，未来将继续推进全球多个新城市的落地计划，其物理AI的基座平台，支撑乘用车、Robovan、Robotruck和Robotaxi的规模化落地，也将扩展至机器人等领域。7月，修订后的《深圳经济特区智能网联汽车管理条例》正式施行，细则在降低申请门槛、支持全无人测试等方面作出调整，将有序探索全域开放道路测试、示范应用。

世界人工智能大会（WAIC）2026上，超节点、万卡集群是算力领域最吸引人的焦点。更大的集群、更高的带宽、更强的供给能力，构成了大模型时代最直观的基础设施叙事。但安谋科技在WAIC期间的AI前瞻会上把讨论拉向了另一端：当AI从云端进入手机、AI PC、智能家居、汽车、机器人、工业设备和边缘节点，算力问题不再只是“够不够大”，而是“合不合适”。“我们认为，边端侧AI不是云端AI的缩小版，而是由功耗、实时性、软件、数据、可靠性共同定义的新计算市场。”安谋科技Arm China NPU产品线负责人、首席架构师舒浩博士在安谋科技世界人工智能大会（WAIC）2026 AI前瞻会上的这句话，比具体产品参数更值得被记住。安谋科技在AI前瞻会上没有把CPU、NPU、VPU和AIOS拆成四个孤立产品来讲，而是给出一套端侧AI的系统答案：硬件层面做异构，软件层面建生态，最终把自己从芯片IP供应商的位置，推向边端侧AI基础设施的组织者。如何在资源受限的情况下让端侧稳定运行AI？端侧AI的机会，首先来自模型能力的下沉。安谋科技CPU产品总监朱晓鸣在演讲中提到，2022年、2023年，要达到商用门槛的模型规模往

7月19日，全球领先的专业咨询服务机构——安永（EY）与火山引擎签署战略伙伴合作备忘录，双方将在人工智能技术落地、专业服务交付与企业 AI 转型等领域深化协同，携手助力企业客户以 AI 为核心重塑业务能力，推动运营模式向 AI 原生架构演进。在安永大中华区首席运营官兼金融服务主管合伙人忻怡、火山引擎总裁谭待见证下，安永大中华区咨询服务主管合伙人吴晓颖与火山引擎副总裁赵文婕代表双方签署协议。安永中国主席陈凯表示：“生成式AI正深刻重塑商业世界，安永始终致力于将前沿技术转化为客户价值。此次与火山引擎强强联合，是安永践行AI战略的关键一步，我们将携手助力客户在AI时代赢得先机。”火山引擎总裁谭待表示：“火山引擎正在助力千行百业探索AI创新，安永在企业转型方面拥有卓越交付力。希望通过此次合作，双方将‘技术+专业服务’紧密结合，让AI能够真正走进企业生产流程，创造更多业务价值。”进入人工智能时代，多数企业的系统与流程仍在沿用信息化时代的底层逻辑，数据孤岛、人工决策滞后等问题普遍存在，带来极大的效率损耗。破解之道在于以AI为核心重构企业的运转基础。面向企业客户服务，双方将围绕企业核心业务场景，共建

从北京市经信局获悉，上半年，全市数字经济增加值增速7.8%，其中数字经济核心产业增加值增速9.8%，对全市GDP增长的带动作用显著。下半年，北京将推进人工智能全域赋能，推动智能翻译、智能同传等在文旅消费领域的应用。制定Token经济发展政策，围绕Token生产、分发和应用等关键环节，布局建设Token工厂、Token分发平台，推动在工业、教育、文旅等重点领域创新应用。举办“创赢未来”OPC专场路演活动，激发超级个体的人工智能应用与创作活力。开发面向OPC的AIGC培训精品课程，释放AIGC技术价值。依托开源芯片研究院、北京通明湖信息技术应用创新中心，打造“RISC-V+AI OS”开源开放生态，构建从芯片指令集到操作系统到智能体应用的全栈自主技术体系。进一步提升智能算力供给能力。打造“超级节点+行业节点”支撑体系，力争下半年新增智能算力5万P，年内算力总规模突破13万P。（财联社）

Google has released Gemini 3.6 Flash and 3.5 Flash-Lite as new workhorses designed to cut latency and token costs for enterprise AI agents. The economics of running autonomous software agents inside a production environment come down to a fixed equation few vendors advertise directly. A model needs to reason through a multi-step task competently, but […] The post Google’s Gemini 3.6 Flash targets enterprise agent token costs appeared first on AI News.

There are new 3.6 and 3.5 models today, but Google is already training Gemini 4.

OpenAI has come forward to claim responsibility for the Hugging Face breach, saying it was the result of internal testing gone awry.

OpenAI and Hugging Face share early findings from a security incident during AI model evaluation, highlighting advanced cyber capabilities and lessons for defenders.

A federal judge has signed off on Anthropic's $1.5 billion class action settlement with authors who accused the company of training its AI models on copyrighted books, as reported earlier by Reuters. In an order on Monday, Judge Araceli Martínez-Olguín writes that the settlement will provide "meaningful relief," offering authors around $3,000 for each book […]

OpenAI launches the ChatGPT for Small Businesses program, helping entrepreneurs build AI skills, automate work, and grow with ChatGPT Work.

OpenAI says its AI models mistakenly breached open-source AI platform Hugging Face during internal testing. In a blog post on Tuesday, OpenAI writes that GPT-5.6 Sol and "an even more capable pre-release model" discovered vulnerabilities within their sandboxed testing environment, allowing them to gain access to the internet and target Hugging Face. On July 16th, […]

The cybersecurity-focused models, including GPT-5.6 Sol, broke out of a testing sandbox, exploited a zero-day, and gained access to the open internet to pull off the attack.

Last week, two Chinese AI companies unveiled models they say can credibly compete with the best systems from OpenAI and Anthropic. The response was swift and predictable. Markets wobbled, commentators declared Silicon Valley shooketh, and policymakers reached for the familiar language of arms races and wake-up calls. In one headline, The Associated Press said a […]

Google is launching Gemini 3.6 Flash alongside a new security model dedicated to quickly finding and patching security vulnerabilities. In a blog post on Tuesday, Google describes Gemini 3.5 Flash Cyber as a "cost-efficient and highly capable alternative" to larger, more expensive AI systems, such as the one offered by Anthropic's Mythos. The cybersecurity model […]

Bristol Myers Squibb is purchasing an Nvidia DGX SuperPOD built on the chipmaker’s Vera Rubin architecture to support artificial intelligence use across its drug discovery and development operations. The pharmaceutical company said it will be the first life sciences group to acquire a DGX SuperPOD based on Vera Rubin. Nvidia introduced the architecture earlier this […] The post Bristol Myers Squibb buys Nvidia AI system for drug discovery appeared first on AI News.

Enterprises evaluating Chinese open-weight models this month face a question that has nothing to do with benchmarks: whether using one will still be straightforward in a year. Moonshot AI’s Kimi K3 arrived on July 16 as the largest open-weight model yet released, and within days it had reopened a policy argument in Washington that had […] The post Chinese open-weight models are cheap. Washington is deciding what that costs. appeared first on AI News.

You open a generative AI tool expecting a quick boost. Ten minutes later, you’re still there, refining a prompt for the fourth time. The task you started with has drifted off to the side somewhere. Sound familiar? Knowledge workers in 2026 are running into this more and more. It makes sense once you look at […] The post The AI Slot Machine Effect: Why Generative Feeds Disrupt Deep Work And How to Reclaim Focus appeared first on AI News.

Buzz is a group chat platform for the workplace that puts humans and their AI agents in the same conversation.