Nvidia to Launch NemoClaw AI Platform for Salesforce, Adobe with Multi-Hardware Support
Nvidia plans to launch NemoClaw, an open AI agent platform pitched to enterprise clients including Salesforce, Cisco, Google, Adobe and CrowdStrike, with support for non-Nvidia hardware. It emphasizes built-in security, supports open-source models versus CUDA, and will debut with a new inference chip at its San Jose developer conference.
1. NemoClaw Platform Overview
Nvidia’s NemoClaw is designed as an open platform for autonomous AI agents capable of executing enterprise tasks without constant human input. The company has pitched the software to major clients such as Salesforce, Cisco, Google, Adobe and CrowdStrike, and promises compatibility with non-Nvidia hardware.
2. Strategic Shift to Open-Source and Security Emphasis
By supporting open-source models rather than relying solely on its proprietary CUDA ecosystem, Nvidia aims to become a software layer across diverse infrastructures. The platform includes built-in security tools to mitigate risks of unpredictable agent behavior, addressing concerns raised by prior autonomous tools in corporate environments.
3. Developer Conference and Future Prospects
Nvidia plans to reveal NemoClaw details alongside a new inference chip at its San Jose developer conference next week. Key information on release timing, pricing and regional availability is expected at the event, which will highlight Nvidia’s strategy to expand beyond gaming and data-center GPUs into enterprise AI services.