By: - Rob Parrish, Head of Product, AI Agent Security, Check Point
At GTC Taipei during COMPUTEX 2026, NVIDIA is highlighting the growing adoption of its NVIDIA Vera BlueField-4 STX architecture and introducing new NVIDIA DOCA-powered innovations designed to secure the next generation of enterprise AI infrastructure.
As organizations continue scaling AI factories, private LLM environments, distributed inference systems, and increasingly autonomous AI operations, enterprise infrastructure requirements are rapidly evolving.
Modern AI environments combine high-performance compute, distributed storage systems, inference pipelines, Kubernetes clusters, APIs, GPU server farms, and sensitive enterprise data operating continuously at enormous scale. At the same time, AI-driven environments are introducing increasingly dynamic machine-to-machine interactions across infrastructure, applications, and workloads.
This evolution is driving demand for AI security architectures capable of protecting modern AI infrastructure environments while supporting increasingly automated AI operations.
AI Infrastructure Introduces New Security Requirements
Traditional security architectures were designed for conventional enterprise applications operating across relatively predictable environments.
AI infrastructure environments operate very differently.
AI factories introduce highly distributed workloads, east-west traffic flows, large-scale inference environments, multi-tenant architectures, continuously evolving traffic patterns, and increasingly automated AI operations spanning infrastructure and applications.
AI factories are also increasingly supporting autonomous AI systems capable of continuously interacting with applications, APIs, data sources, storage environments, and other services across enterprise infrastructure.
As organizations scale AI deployments, infrastructure visibility, workload isolation, runtime protections, segmentation, and policy consistency become increasingly important across the AI environment itself.
This shift is accelerating interest in infrastructure-native protection models capable of operating closer to the infrastructure layer itself and more directly within modern AI runtime environments.
NVIDIA BlueField and Infrastructure-Native Protection
NVIDIA Vera BlueField-4 STX represents an important advancement in how AI infrastructure environments can be secured and operated.
Powered by NVIDIA DOCA, BlueField-4 introduces infrastructure-native capabilities designed to support visibility, isolation, runtime protections, and policy enforcement directly within high-performance AI environments.
By operating within the DPU and closer to the data path itself, organizations can extend protections across distributed AI infrastructure environments while supporting the operational requirements of enterprise-scale training and inference workloads.
NVIDIA DOCA innovations such as DOCA Argus, DOCA Flow and DOCA Vault further extend these capabilities through runtime workload visibility, secure tenant segmentation, data access controls, and infrastructure-level protections designed specifically for AI factory environments.
Together, these technologies help establish a more resilient and scalable foundation for modern AI infrastructure operations.
Extending Security Across AI Factory Environments
This is where the collaboration between Check Point and NVIDIA becomes especially important.
Check Point customers can already leverage AI Factory Firewall integrated with NVIDIA BlueField and NVIDIA DOCA architectures to help secure distributed AI infrastructure environments, Kubernetes-based AI deployments, private LLM environments, east-west traffic flows, inference workloads, and increasingly autonomous AI operations running across AI factories.
As organizations adopt AI agents and automated AI workflows, infrastructure environments are becoming increasingly dynamic. AI agents continuously interact with APIs, enterprise applications, storage environments, data pipelines, and other services across distributed infrastructure environments. This creates additional operational and security requirements around visibility, segmentation, workload protections, and infrastructure-level policy enforcement.
This integration allows organizations to extend security visibility and enforcement capabilities closer to where AI workloads and AI-driven operations execute, supporting layered protections across modern AI infrastructure architectures.
This includes support for:
- AI factory segmentation
- Runtime workload visibility
- East-west traffic protections
- Infrastructure-level policy enforcement
- Distributed inference environments
- Multi-tenant AI deployments
- Kubernetes-based AI environments
- AI agent and automated workflow environments
As AI infrastructure environments continue expanding, organizations increasingly require security architectures capable of scaling alongside modern AI operations.
Building the Secure Foundation for Enterprise AI Infrastructure
With NVIDIA BlueField and NVIDIA DOCA providing the infrastructure foundation for secure AI factories, and Check Point AI Factory Firewall extending security protections across those environments, organizations gain a path toward building more resilient AI infrastructure architectures designed for next-generation enterprise AI operations.
As AI systems become increasingly autonomous, infrastructure security is becoming foundational to how organizations maintain visibility, resilience, operational consistency, and protection across rapidly evolving AI environments.
