英伟达发布了运行AI智能体在固定边界内设计的方案，覆盖从企业基础设施到边缘设备。英伟达表示，Anthropic、OpenAI、Meta、微软、谷歌云和AWS等公司正在合作。
Nvidia has released a design for running AI agents inside fixed limits, from enterprise infrastructure to edge machines. Anthropic, OpenAI, Meta, Microsoft, Google Cloud and AWS are among the companies working with it, Nvidia said.
英伟达在周三的柏林GTC大会上发布了OpenShell参考架构，现已开放。上周，英伟达公布了20亿美元欧洲交易计划。
Nvidia announced the OpenShell Reference Architecture at its GTC Berlin conference on Wednesday. It is available now. Last week, Nvidia set out plans for $2B in European deals.
该设计的工作原理：该设计基于英伟达的开放源码运行时软件OpenShell，为智能体能执行的操作设定边界。黄仁勋在主题演讲中表示，该软件以Apache 2.0许可证发布。OpenShell也是英伟达于9月推出的智能体安全平台的一部分。模型运行在英伟达的Vera Rubin系统上。驱动智能体及其工具的软件运行在Vera CPU上。
How it works The design builds on OpenShell, Nvidia’s open source runtime software, which sets a boundary around what an agent can do. Huang said in his keynote that it is released under the Apache 2.0 licence. OpenShell is also part of the agent safety platform Nvidia launched in September. Models run on Nvidia’s Vera Rubin systems. The software that drives the agents and their tools runs on Vera CPUs.
BlueField-4数据处理器运行英伟达的Sentry软件，该软件从主系统外部监控智能体的行为。如果智能体试图超出其软件边界，Sentry会在毫秒内将其隔离并停止，英伟达表示。Sentry还能将智能体容量提升高达20%。
BlueField-4 data processing units run Nvidia Sentry, software that watches agent behaviour from outside the main system. If an agent tries to move outside its software boundary, Sentry quarantines and stops it in milliseconds, Nvidia said. It said Sentry also raises agent capacity by up to 20%.
该设计还覆盖边缘设备，包括RTX Spark、DGX Spark和DGX Station系统。上周，英伟达展示了专为本地运行智能体而设计的RTX Spark笔记本。
The design also covers edge machines, including RTX Spark, DGX Spark and DGX Station systems. Nvidia showed RTX Spark laptops built to run agents locally last week.
“AI能力和安全必须同步推进。随着企业准备部署数十亿个AI智能体，安全和控制必须嵌入计算架构中。”英伟达创始人兼首席执行官黄仁勋表示。黄仁勋上周表示，将切断那些构建不安全AI的客户。
“AI capability and safety must advance together. As enterprises prepare to deploy billions of AI agents, safety and control must be built into the computing architecture,” said Jensen Huang, Nvidia’s founder and chief executive. Huang said last week he would cut off customers who build unsafe AI.
参与合作的公司：Anthropic、Cohere、谷歌、Meta、OpenAI、Perplexity、Reflection和SpaceX Super Intelligence的智能体可在OpenShell基础设施上运行，英伟达表示。1Password、Check Point、CrowdStrike、思科、Fortinet、Okta、Palo Alto Networks和Zscaler等安全公司也在为其增加支持。
Who is working with it Agents from Anthropic, Cohere, Google, Meta, OpenAI, Perplexity, Reflection and SpaceX Super Intelligence can run on OpenShell infrastructure, Nvidia said. Security companies including 1Password, Check Point, CrowdStrike, Cisco, Fortinet, Okta, Palo Alto Networks and Zscaler are adding support for it.
在舞台上，黄仁勋指出了英伟达与Anthropic的合作伙伴关系。Anthropic当时表示，自9月以来，Anthropic已将其用于运行Claude代理的服务Claude Managed Agents与OpenShell配对，以便公司能够限制Claude代理的行为并记录每一个动作。
On stage, Huang pointed to Nvidia’s partnership with Anthropic. Since September, Anthropic has paired Claude Managed Agents, its service for running Claude agents, with OpenShell, so companies can limit what Claude agents do and log every action, Anthropic said at the time.
BBVA和桑坦德银行正在探索该技术，以加强代理的安全性和治理。SAP正将其构建到其Joule Studio运行时中。ServiceNow于周三正式推出了长期运行的个人代理Arc，其行为由OpenShell强制执行。
BBVA and Santander are exploring it to strengthen agent safety and governance. SAP is building it into its Joule Studio runtime. ServiceNow made Arc, a long-running personal agent, generally available on Wednesday, with its actions enforced by OpenShell.
该设计可在Canonical、SUSE和红帽的Linux和Kubernetes上运行，并通过Microsoft Execution Containers在Windows上运行。它还可在AWS、Google Cloud和Microsoft Azure上运行。戴尔、思科、惠普企业、联想和超微正在利用它构建基础设施。
The design runs on Linux and Kubernetes from Canonical, SUSE and Red Hat, and on Windows through Microsoft Execution Containers. It also runs on AWS, Google Cloud and Microsoft Azure. Dell, Cisco, HPE, Lenovo and Supermicro are building infrastructure with it.
黄仁勋表示：“有了OpenShell参考架构，代理可以在Rubin上思考，在Vera上行动，在BlueField上受管。”英伟达未给出价格。
“With the OpenShell Reference Architecture, agents can think on Rubin, act on Vera and be governed on BlueField,” Huang said. Nvidia did not give prices.