NVIDIA and Google Cloud have forged a groundbreaking partnership to bring Google’s Gemini AI models to enterprise environments through NVIDIA’s Blackwell platform and Confidential Computing technology. The collaboration, announced on April 9, 2025, enables organisations to deploy advanced AI capabilities while maintaining strict security and data sovereignty requirements.

This strategic integration addresses a critical challenge in enterprise AI adoption by allowing companies to run sophisticated AI models on-premises without compromising security or performance. The partnership leverages NVIDIA’s Blackwell HGX and DGX platforms to power Google Distributed Cloud, creating a secure foundation for enterprise-grade AI deployment.

Sachin Gupta, Vice President and General Manager of Infrastructure and Solutions at Google Cloud, emphasised the significance of this development: “By bringing our Gemini models on premises with NVIDIA Blackwell’s breakthrough performance and confidential computing capabilities, we’re enabling enterprises to unlock the full potential of agentic AI.”

The collaboration introduces several technological innovations, including NVIDIA’s Confidential Computing technology, which creates protected enclaves for both API prompts and training data. This dual-layer protection system specifically addresses privacy concerns that have previously prevented regulated industries from adopting advanced AI solutions.

Google Cloud’s newly introduced GKE Inference Gateway works seamlessly with NVIDIA Triton Inference Server and NeMo Guardrails to optimise AI workload management. This integration enables intelligent routing and balancing of AI tasks while maintaining centralised governance and cost efficiency.

The partnership’s impact extends across multiple sectors. Healthcare organisations can now analyse medical records while maintaining patient data security within their firewalls. Financial institutions can implement advanced fraud detection systems without exposing sensitive customer information, and government agencies can process classified data while maintaining security clearances.

NVIDIA’s Confidential Computing technology serves as the security cornerstone of this collaboration, establishing a new standard for protected AI operations in enterprise environments. The technology creates secure enclaves that safeguard both API prompts and training data, addressing the primary security concerns of regulated industries.

The integration also includes future plans to incorporate NVIDIA Dynamo, designed to address the unique challenges of scaling reasoning models in enterprise environments. This addition will ensure smooth deployment and management of agentic AI workloads across organisations.

This development comes at a crucial time, following Google’s recent launch of Gemini 2.5 Pro and NVIDIA’s continued expansion of its AI hardware capabilities. The collaboration represents a significant step forward in making advanced AI accessible to organisations that require stringent security measures.

The partnership’s broader implications suggest a transformation in how enterprises approach AI implementation, particularly in regulated industries. By combining Google’s advanced AI models with NVIDIA’s secure computing infrastructure, organisations can now leverage sophisticated AI capabilities while maintaining compliance with regulatory requirements.

News Source: https://blogs.nvidia.com/blog/google-cloud-next-agentic-ai-reasoning/