Google and NVIDIA Enable On-Premises Deployment of Gemini AI Models

The collaboration integrates NVIDIA’s Blackwell GPUs with Google Distributed Cloud to meet enterprise demands for secure, localized AI solutions.

Google Cloud has announced that enterprises can now deploy its multimodal Gemini AI models in their own data centers using Google Distributed Cloud.
The initiative, developed in partnership with NVIDIA, leverages Blackwell GPUs and confidential computing to ensure data security and compliance.
This deployment is designed to address the needs of highly regulated industries, including healthcare, finance, and government, by supporting strict data sovereignty requirements.
New infrastructure tools, such as GKE Inference Gateway and Cluster Director, enhance scalability and efficiency for large-scale AI workloads.
The Gemini models, capable of processing text, audio, and video in over 100 languages, introduce agentic AI capabilities for autonomous IT management and enterprise threat detection.