Workflow
AI原生思维
icon
Search documents
华为云CEO:384超节点每卡性能可达英伟达H20三倍
Guan Cha Zhe Wang· 2025-08-30 03:38
Core Viewpoint - The importance of chips is acknowledged, but the ability to provide the required computational results for customers is emphasized as more critical [1] Group 1: Huawei Cloud's Developments - Huawei Cloud is undergoing significant organizational restructuring, with a focus on enhancing computational power through its Ascend AI cloud services and Tokens services [1] - The CloudMatrix384 super node, launched in April, integrates 384 Ascend NPUs and 192 Kunpeng CPUs, achieving a computational scale of 300 PFlops [2] - The Tokens service has been integrated with the CloudMatrix384 super node, achieving a maximum of 2400 TPS and 50 ms latency, surpassing industry standards [2][3] Group 2: Performance and Growth Metrics - Huawei Cloud's overall computational capacity has increased by nearly 250% compared to the previous year, with the number of Ascend AI cloud service customers rising from 321 to 1714 [5] - The CloudMatrix384 super node can support training of large models, with the capability to connect 432 super nodes to form a 160,000-card AI cluster [2] - The deployment of over 40 CloudMatrix384 super nodes in Guizhou is part of a strategy to create a national computational network [5] Group 3: Market Position and Strategic Focus - Huawei Cloud ranks second in China's cloud service market with an 18% share, while Alibaba Cloud holds a 33% share [6] - The market demand is shifting from "cloud adoption" to "AI integration," prompting Huawei Cloud to streamline its operations and focus on maximizing the advantages of its Ascend AI and Pangu model combinations [6][7] - The organizational restructuring aims to concentrate resources on core areas that can leverage AI capabilities effectively [6][7]
华为云张平安:坚持打造“算力黑土地” 加速行业智能跃迁
Yang Guang Wang· 2025-08-28 13:52
央广网贵阳8月28日消息(记者殷雨婷)8月28日,中国国际大数据产业博览会在贵州省贵阳市举办,华为常务董事、华为云计算CEO张平安在大会开幕 式上表示,面对未来十年算力需求的数万倍增长,华为云将坚定不移地打造"算力黑土地",充分发挥公司多年积累的联接技术优势和数据中心资源布局,以 优的性能、好的服务、低的成本,通过昇腾AI云服务和Tokens服务满足客户的"最终计算结果",加速行业智能跃迁。 图为华为常务董事、华为云计算CEO张平安 发挥"大杂烩"优势,打造算力"黑土地" 算力是智能世界的基础设施,大模型催生了对大算力的需求,未来十年,算力的需求可能会有数万倍的增长。 张平安介绍道, 华为云一直坚定打造"算力黑土地",围绕贵安、乌兰察布、和林格尔、芜湖三大核心枢纽,打造 "全国算力一张网",中国的算力黑土 地正在成为支持全球客户的AI算力场。今天,华为云的整体算力规模相比去年同期增长率接近250%,使用昇腾AI云服务的客户从去年的321家增长到今年的 1714家。 在贵安,华为云部署了最大规模的CloudMatrix384超节点,服务全国客户,打造东数西算工程的标杆;在贵安、乌兰察布还建设了超大规模的灾备云中 ...
华为云张平安:坚持打造“算力黑土地”,加速行业智能跃迁
Jing Ji Wang· 2025-08-28 08:41
张平安介绍道, 华为云一直坚定打造"算力黑土地",围绕贵安、乌兰察布、和林格尔、芜湖三大核心枢纽,打造 "全 国算力一张网",中国的算力黑土地正在成为支持全球客户的AI算力场。今天,华为云的整体算力规模相比去年同期增长率 接近250%,使用昇腾AI云服务的客户从去年的的321家增长到今年的1714家。 在贵安,华为云部署了最大规模的CloudMatrix384超节点,服务全国客户,打造东数西算工程的标杆;在贵安、乌兰 察布还建设了超大规模的灾备云中心,为企业,尤其是央国企,提供高性能、稳定可靠的算力服务。 在目前芯片工艺受限情况下,如何提升算力的性能和规模?张平安表示,华为可以发挥"大杂烩"的优势——综合在光 通信、网络、供电等多方面的技术,以系统补单点,以空间换算力,以带宽换算力,以能源换算力,通过云上算力集群, 获得规模优势和提升性能。 今年4月,华为云在芜湖发布了CloudMatrix384超节点,将384颗昇腾NPU和192颗鲲鹏CPU,通过全新高速网络 MatrixLink全对等互联,形成一台超级"AI服务器",算力规模可以达到300 PFlops。此外,对于万亿、十万亿参数的大模型 训练任务,还可 ...
混沌AI教练营报名倒计时 | 以 AI 为光,做时代的「摆渡人」
混沌学园· 2025-08-27 11:58
Core Viewpoint - The article emphasizes the importance of AI as an essential component for business survival in the modern era, advocating for the role of AI coaches to guide enterprises in effectively utilizing AI technology [5][8][10]. Group 1: AI as a Necessity - AI is no longer optional but a fundamental requirement for businesses to thrive [8]. - The rapid advancement of large models and decreasing computing costs indicate that algorithms are evolving from mere tools to foundational infrastructure [9]. - The real opportunity lies in reconstructing business processes with AI rather than merely using it as an efficiency tool [10]. Group 2: AI Coach Program - The company has established the AI Coach Camp to empower individuals to master AI technology and lead enterprises through AI transformation [12]. - The program focuses on practical learning, combining theoretical knowledge with hands-on experience in real business scenarios [17]. - Participants will engage with industry experts and entrepreneurs to gain insights into AI trends and practical applications across various sectors [16]. Group 3: Learning and Collaboration - The learning approach is designed to foster collaboration among participants, allowing them to share experiences and insights [18]. - AI coaches serve as connectors between technical potential and business value, facilitating deep connections with enterprises [20]. - The program aims to transform complex AI concepts into actionable strategies that businesses can understand and implement [20]. Group 4: Enrollment Details - Key enrollment dates include the closing of the application on August 29, online interviews by August 31, and the start of the program on September 5 [27]. - The investment for participation is set at 6,800 yuan per person, with discounts available for group registrations [27]. - Participants will have opportunities for practical application and certification by January 10 [27].