Rebeca Moen
Apr 11, 2025 17:09
NVIDIA partners with Google Cloud to introduce agentic AI for enterprises, enhancing data security and performance through their Blackwell architecture and Confidential Computing.
NVIDIA has teamed up with Google Cloud to deliver agentic AI capabilities to enterprises, utilizing the Google Gemini models on the NVIDIA Blackwell infrastructure. This collaboration aims to enhance enterprises’ ability to harness AI while maintaining data security and compliance, according to NVIDIA.
Enhanced Security and Performance
Enterprises will benefit from NVIDIA’s Blackwell architecture, which integrates with Google Distributed Cloud to ensure data sovereignty and regulatory compliance. The NVIDIA Confidential Computing framework further secures sensitive data, protecting it from unauthorized access and potential leaks. This dual-layer protection enables enterprises to innovate securely with the Gemini models while ensuring data privacy.
Sachin Gupta, vice president and general manager of infrastructure and solutions at Google Cloud, highlighted the importance of this collaboration in enabling secure innovation without compromising performance. The integration of Gemini models with NVIDIA’s infrastructure is set to unlock the potential of agentic AI for business applications.
Transformative Potential of Agentic AI
Agentic AI is revolutionizing enterprise technology by offering advanced problem-solving capabilities. Unlike traditional AI models, which rely on learned knowledge, agentic AI systems can reason, adapt, and make decisions autonomously. This ability is particularly valuable in fields like IT support and finance, where AI can diagnose issues, execute fixes, and prevent fraudulent activities proactively.
Addressing On-Premises Needs
Many enterprises face challenges in deploying AI models on-premises due to stringent security or data sovereignty requirements. The new solution from NVIDIA and Google Cloud addresses these concerns by offering confidential computing capabilities that secure agentic AI workloads in all environments, whether cloud-based or hybrid.
The NVIDIA HGX B200 platform, powered by Blackwell GPUs, enhances energy efficiency and performance while safeguarding AI models and data integrity. This advancement allows enterprises to leverage cutting-edge AI without sacrificing security.
AI Observability and Security Enhancements
To scale agentic AI effectively, robust observability and security measures are necessary. Google Cloud has introduced a new GKE Inference Gateway, which optimizes AI inference workloads through advanced routing and scalability features. Integrating with NVIDIA Triton Inference Server and NeMo Guardrails, it ensures centralized security and governance of AI models.
Looking forward, Google Cloud plans to enhance observability for agentic AI workloads by incorporating NVIDIA Dynamo, an open-source library designed to scale reasoning AI models. This initiative will further support enterprises in deploying and managing AI applications efficiently.
Image source: Shutterstock
