Enterprises can now deploy agentic AI locally with NVIDIA Blackwell and Google Cloud’s Gemini models.
Google Cloud and NVIDIA have announced a significant collaboration that aims to reshape how enterprises deploy and scale artificial intelligence. The companies unveiled plans to bring agentic AI capabilities to highly regulated industries by allowing enterprises to run Google’s Gemini language models locally on NVIDIA-powered infrastructure.
The announcement, made at Google Cloud Next 2025, centers on enabling enterprises to securely deploy AI models on-premises using Google Distributed Cloud and NVIDIA’s latest Blackwell architecture. The move comes as businesses increasingly seek to harness AI for automation, decision-making, and real-time responsiveness—without compromising on data privacy or regulatory compliance.Local deployment for sensitive dataAt the heart of this partnership is a growing demand for localized AI infrastructure. Sectors like healthcare, finance, and telecommunications are under pressure to innovate with AI while ensuring compliance with strict data residency and sovereignty laws. The integration of Gemini models with NVIDIA’s DGX and HGX platforms—equipped with the new Blackwell GPUs—enables organizations to keep sensitive data within their own environments while tapping into advanced AI capabilities.NVIDIA’s Confidential Computing framework plays a crucial role by securing both data and model integrity. Enterprise users can protect proprietary prompts and training data, while model owners gain safeguards against unauthorized access and tampering. “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,” said Sachin Gupta, vice president and general manager of infrastructure and solutions at Google Cloud. “This collaboration helps ensure customers can innovate securely without compromising on performance or operational ease.”A new era of AI agentsUnlike traditional AI systems that retrieve information or generate responses based on static inputs, agentic AI can reason, adapt, and take actions autonomously. The technology has applications that go far beyond standard use cases. In IT operations, for instance, an agentic AI model can identify problems, apply fixes, and escalate issues—all without human intervention. In the financial sector, these agents could block potentially fraudulent transactions in real-time and adjust fraud detection algorithms dynamically.The collaboration represents a broader trend toward enterprise AI systems that not only understand but act—ushering in an era where business processes can become self-correcting and self-improving.Infrastructure to govern AI workloadsGoogle Cloud also introduced new tools to manage and scale agentic AI workloads. The GKE Inference Gateway, announced alongside the Gemini deployment capabilities, helps optimize inference routing and load balancing. It integrates with NVIDIA Triton Inference Server and NeMo Guardrails, giving enterprises a streamlined way to secure and govern AI deployments.Looking ahead, Google plans to expand observability for AI environments with NVIDIA Dynamo—an open-source tool for scaling reasoning models across production-grade systems.Together, Google Cloud and NVIDIA are laying the groundwork for secure, scalable, and intelligent AI systems that meet the complex demands of today’s enterprise environments.
AI Compliance AI Infrastructure AI Security Blackwell GPU Confidential Computing Data Sovereignty Enterprise AI Fraud Detection Gemini Models Google Cloud IT Automation Nvidia On-Prem AI
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Google Just Showed How To Sell Agentic AI — Will Rivals Pay Attention?Based on its product marketing for agentic AI, Google seems to grasp the need to speak to solving business problems rather than trotting out platitudes about AI.
Read more »
Isaac GR00T N1: World’s first foundation model for humanoid robotics unveiled by NVIDIANVIDIA also announced a collaboration with Google DeepMind and Disney Research to develop Newton.
Read more »
AI Tech Giant Nvidia Partners With Disney and Google on Next-Generation Robotics SoftwareNvidia CEO Jensen Huang announced the partnership at his company's developer conference Tuesday.
Read more »
Nvidia Dynamo — Next-Gen AI Inference Server For EnterprisesAs AI reasoning models become mainstream, Dynamo represents a critical infrastructure layer for enterprises looking to deploy these capabilities efficiently.
Read more »
Shopsense AI Launches Industry-first Multi-Agent SystemShopsense’s Agentic AI platform transforms content monetization by integrating sponsorable shopping experiences across programs seamlessly.
Read more »
Palo Alto Networks CEO describes cybersecurity risk with agentic AIPalo Alto Networks CEO Nikesh Arora detailed how agentic AI technology brings new threats from cybercriminals.
Read more »