The rapid evolution of agentic AI is reshaping enterprise operations, transitioning from human-driven to machine-speed processes. This shift places significant strain on existing technology infrastructures, as agents generate extensive internal communications and complex queries, potentially overwhelming traditional networks and databases.
To address these challenges, Google has introduced a cross-cloud infrastructure for the agentic enterprise at Google Cloud Next '26, featuring innovations across four key areas:
- Fluid compute: A collaboration between Google Compute Engine and Kubernetes services that enables high-speed AI agents and enterprise workloads through enhanced compute and orchestration capabilities.
- Secure cross-cloud connectivity: Tools like Agent Gateway and Cloud Armor create a secure networking foundation for AI agents, ensuring visibility and governance across clouds.
- Unified data layer: Innovations such as Smart Storage and Knowledge Catalog transform static data archives into dynamic reasoning engines, providing AI agents with essential context.
- Digital sovereignty: Features like Confidential External Key Management and enhancements in Google Distributed Cloud ensure that AI capabilities are available wherever data resides.
Fluid Compute
Fluid compute addresses the unpredictability of agentic workloads, allowing for real-time adaptation and resource allocation. This is crucial for both traditional enterprise applications and AI agents. New offerings include:
- Google C4N Series: Optimized VMs capable of processing up to 95 million packets per second, enhancing performance for demanding workloads.
- Google M4N Series with Hyperdisk Extreme: This series enhances data throughput and reduces costs, particularly for Oracle workloads.
- GKE Agent Sandbox: A secure environment for agents that can scale rapidly, launching up to 300 sandboxes per second.
Secure Cross-Cloud Connectivity
The transition to agentic AI introduces new network challenges, particularly regarding security and predictability. Key features include:
- Agent Gateway: Acts as a traffic controller for enterprise agent interactions, ensuring secure and compliant operations.
- Cloud Network Insights: Provides visibility across hybrid and multi-cloud infrastructures, enhancing troubleshooting and performance monitoring.
- Enhanced Cloud NGFW and Cloud Armor: Offers AI-powered protection against evolving cyber threats, ensuring the security of AI agent communications.
Unified Data Layer
A cohesive data strategy is essential for effective AI operations. Enhancements in this area include:
- Smart Storage: Converts unstructured data into actionable knowledge for AI agents, improving data accessibility.
- Knowledge Catalog: Maps data across the organization, allowing agents to access accurate and relevant information without needing data migration.
Digital Sovereignty
To meet the diverse needs of public sector and enterprise customers, Google offers solutions that ensure control over data and compliance:
- Confidential External Key Management: Provides organizations with complete control over their encryption keys within a secure environment.
- Gemini on Google Distributed Cloud: Allows for secure deployment of AI capabilities in sensitive environments, supporting various operational models.
Conclusion
The convergence of these innovations equips enterprises with a robust, secure, and adaptive infrastructure to navigate the complexities of the agentic era. This foundational shift not only enhances operational efficiency but also positions organizations to fully leverage the capabilities of AI.
