Here are the contents translated into English:

This article is a summary of Microsoft Build 2025 conference. On this occasion, Microsoft announced a series of major updates and new features, including:

  1. Microsoft Discovery: A platform that uses Agentic AI to revolutionize the scientific discovery process.
  2. NLWeb: An open-source project that makes it easier for websites to integrate with AI chatbots.
  3. Copilot in Power BI: A new feature that allows users to interact with Power BI reports through natural language dialogue.
  4. SQL Server 2025: A major upgraded version that comes with built-in scalable AI capabilities and integration with Microsoft Azure and Microsoft Fabric.
  5. Fabric: A continuously evolving platform that combines Azure Cosmos DB, Azure Databricks, and digital twin builder.
  6. Foundry Local: A service that allows users to run the same AI capabilities on-premises as they do in the cloud, enabling the development of cross-platform AI applications that can run offline.

Additionally, the conference also announced a series of infrastructure updates, including:

  1. Per-dollar-per-watt processing token count as a technical metric for measuring progress.
  2. NVIDIA GB200 Grace Blackwell chip first large-scale deployment on Azure.
  3. Arm architecture processor Cobalt widespread adoption in internal products.

In summary, these updates and new features demonstrate Microsoft’s efforts to advance AI technology development, particularly in scientific research, data analysis, and office applications.

Translation

这篇文章是微软 Build 2025 大会的总结。大会上，微软宣布了一系列重大更新和新功能，包括：

  1. Microsoft Discovery：一个利用Agentic AI改变科学发现流程的平台。
  2. NLWeb：一个开源项目，让网站更容易接入AI聊天机器人。
  3. Copilot in Power BI：一个新功能，允许用户直接通过对话的形式与Power BI报告进行交互。
  4. SQL Server 2025：一种重大升级版本，内置可扩展AI功能和集成Microsoft Azure和Microsoft Fabric。
  5. Fabric：一个持续演进的平台，集成了Azure Cosmos DB、Azure Databricks和数字孪生构建器。
  6. Foundry Local：一个服务，让用户在本地也能运行和云上相同的AI能力，从而构建可以离线运行的跨平台AI应用。

此外，大会上还宣布了一系列基础设施更新，包括：

  1. 每美元每瓦特处理token数量作为衡量指标的技术跃迁。
  2. NVIDIA GB200 Grace Blackwell芯片在Azure上的首个大规模上线。
  3. Arm架构的处理器Cobalt在内部产品中的广泛使用。

总之，这些更新和新功能表明微软正在努力推进AI技术的发展，特别是在科学研究、数据分析和办公应用方面。

Reference:

https://www.youtube.com/watch?v=ceV3RsG946s

<
Previous Post
Anthropic Developer Conference 2025
>
Next Post
Wall Street view of Google