《科创板日报》20日讯，据港交所披露，PPIO提交港交所上市申请，联席保荐人为工银国际、申万宏 源香港。招股书显示，PPIO是中国独立分布式云计算服务商，按计算节点数计算，运营着中国最大的 算力网络，截至2024年12月31日，PPIO 的算力网络覆盖全球1200多个县市，包含4000多个计算节点。 （记者 黄心怡） PPIO提交港交所上市申请 ...

The stock of Amazon (AMZN -0.94%) has delivered life-changing gains for investors who got in early enough. It's up nearly 200,000% over its lifetime and 900% over the past 10 years.Everyone is talking about what it's doing in artificial intelligence (AI) today and how that could jump-start sales growth. There's a huge long-term opportunity, but can Amazon stock double over the next five years? Growth drivers everywhereWith all of the AI hype, let's not forget that Amazon's main business, for now at least, i ...

Amazon's cloud business is planning a major revamp of its AI application strategy, with a new bundled service, Business Insider has learned. Internally called Q Business Suite (QBS), the new service is described as a "unified workspace" that combines business intelligence and generative AI capabilities from existing Amazon tools such as QuickSight, Q Business, and Q Apps, according to internal documents obtained by BI.QuickSight is Amazon Web Services's data visualization software, and Q Business is the co ...

金十数据6月20日讯，在6月20日的华为开发者大会2025上，华为常务董事、华为云计算CEO张平安发布 了CloudRobo具身智能平台。他强调，华为云不做机器人本体，把机器人本体交给伙伴。华为云的目标 是让一切联网的本体都成为具身智能机器人。 华为云发布CloudRobo具身智能平台 ...

从产品层面来看，Snowflake 简单易用且价值主张明确，这促使其在各类企业中快速普及。其最新产品 Cortex 因简洁性而脱颖而出，能助力客户迅速启动项目并取得实际成效，优于 Amazon(AMZN.US) Bedrock 等竞品。Snowflake 还积极拥抱人工智能技术，通过旗下 Agents 产品，使客户在数据查询等任 务上实现从数周到数小时的效率飞跃。例如，有金融顾问利用 AI 在 45 分钟内完成原本需一周的客户 请求响应。 此外，Snowflake 在数据跨部门共享方面的优势，使其在当下数据驱动决策的时代更具竞争力。其产品 体系日臻完善，从一年前存在诸多功能缺失到如今努力打造 360 度全方位平台，以减少客户对其他工具 的依赖。 根据小摩研报，Snowflake 近期财务数据表现出色。在 2025 财年，其收入预计为 36.26 亿美元，调整后 息税折旧摊销前利润EBITDA为 5.67 亿美元。2026 财年，收入有望增长至 45.15 亿美元，EBITDA 达 7.58 亿美元。而 2027 财年，收入预期进一步提升至 54.19 亿美元，EBITDA 为 9.5 亿美元。 小摩对 Sno ...

Core Viewpoint - Huawei announced the release of the Pangu Model 5.5 at the Huawei Developer Conference 2025, highlighting significant upgrades in predictive capabilities through a novel triplet transformer architecture [1] Group 1: Model Features - The Pangu Model 5.5 utilizes a unique triplet transformer architecture for unified pre-training, allowing for the integration of diverse data types from various industries [1] - The model can efficiently process and pre-train tabular data, time-series data from equipment logs, and image data from product inspections within the same framework [1] - This approach is expected to significantly enhance the accuracy of predictive models and improve generalization across different industries and scenarios [1]

Group 1 - Major domestic companies such as Meituan, Kingsoft Office, Ximalaya, and 360 have integrated with Huawei's cloud ModelArts Studio, a large model as a service platform (MaaS) [1]

谭待转战字节，火山秀出肌肉 2020年底，字节跳动完成一笔声名不显但影响深远的行业并购。 成立一年半，聚焦"互联网医疗"服务的幺零贰四科技有限公司被字节跳动并收入囊中。 比幺零贰四公司本身更出名的，是它异常豪华的创始团队。原百度副总裁吴海锋，前百度执行总监孙雯玉，前百度搜索首席架构师谭待........一众百度搜索 技术大拿，赫然在列。 明面业务并购，实际暗俘人才与技术。以这种极为讨巧的人才并购方式，字节将一众百度搜索技术大拿揽入麾下。 在一众大佬光环下，谭待并不是最出彩的那一位。尽管早早干到了百度最高的T11级工程师，但他依旧只能算是一名技术骨干。 没有掌舵过业务，没有管理过大团队经验的谭待，继续干老本行。最初，他被委任主导火山引擎技术架构设计，重点布局云计算与AI基础设施。 虽背靠字节这棵大树，但彼时同行对火山云业务并不看好。无他，字节云计算起步太晚了。整整十年的市场与技术代差，增加了谭待的闯关难度。 但字节硬是不信邪，谭待也偏向虎山行。一年后的2021年11月，火山引擎被字节跳动拔升为六大核心业务板块之一，谭待也升任火山引擎总经理。 压力山大的谭待，以更具挑战性的目标规划直面外界质疑：未来8-10年火山 ...

Core Insights - Huawei Cloud has released the Pangu Model 5.5, which includes a deep thinking model with 718 billion parameters based on MoE architecture, a world model based on the Pangu multimodal model, and a visual model with 30 billion parameters using MoE architecture [1] Group 1 - The Pangu Model 5.5 features a deep thinking model with 718 billion parameters [1] - The release includes a world model based on the Pangu multimodal model [1] - A visual model with 30 billion parameters is also part of the Pangu Model 5.5 [1]

订阅人工智能动态 +订阅 订阅华为动态 +订阅 华为新一代昇腾AI云服务上线 金十数据6月20日讯，6月20日举行的华为开发者大会2025（HDC 2025）上，华为常务董事、华为云计 算CEO张平安宣布基于CloudMatrix384超节点的新一代昇腾AI云服务上线。据了解，华为云新一代昇腾 AI云服务基于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"AI服务器"，单卡推理吞吐量跃升到2300Tokens/s。 ...