NVIDIA has updated its Enterprise AI Factory validated design to incorporate NVIDIA BlueField cybersecurity and infrastructure acceleration capabilities, marking a series of announcements aimed at enhancing enterprise AI operations. The expansion integrates leading software platforms from ecosystem partners—including Armis, Check Point, F5, Fortinet, Palo Alto Networks, Rafay, Red Hat, Spectro Cloud, and Trend Micro—validating them for deployment on NVIDIA RTX PRO Servers with BlueField acceleration.
BlueField processors offload essential services such as networking, storage, security, and orchestration, allowing CPUs and GPUs to remain dedicated to AI workloads. By isolating security and infrastructure tasks on dedicated hardware, enterprises can scale AI operations while maintaining real-time, zero-trust security across the AI pipeline. NVIDIA’s DOCA Argus framework provides runtime visibility and threat detection, offering continuous monitoring of AI data from ingestion to inference.
Integrated partner solutions now validated under this framework address multiple operational needs. Armis Centrix delivers continuous cyber exposure management, Check Point Infinity AI Cloud Protect provides network and host security with telemetry support, and F5 BIG-IP Next for Kubernetes enhances workload isolation and policy enforcement. Fortinet FortiGate VM extends zero-trust firewalling to AI factories, while Palo Alto Networks Prisma AIRS enforces runtime protection at the infrastructure layer. Rafay and Red Hat OpenShift improve production AI orchestration and security, and Spectro Cloud’s PaletteAI supports regulated environments with scalable, efficient AI operations. Trend Vision One contributes real-time monitoring and global threat intelligence integration.
BlueField-4 Powers AI-Native Storage Infrastructure
In parallel, NVIDIA introduced the BlueField-4 data processor as part of the NVIDIA Inference Context Memory Storage Platform. This infrastructure addresses the demands of AI models with trillions of parameters, which generate large context datasets essential for multi-turn inference. By extending GPU memory and enabling high-speed context sharing across nodes, the platform increases tokens per second by up to five times while improving power efficiency compared with traditional storage. Hardware-accelerated cache placement reduces metadata overhead, isolates data from GPU nodes, and integrates with NVIDIA Spectrum-X Ethernet for RDMA-based access. Storage providers including AIC, Cloudian, DDN, Dell Technologies, HPE, Hitachi Vantara, IBM, Nutanix, Pure Storage, Supermicro, VAST Data, and WEKA are among the first to adopt BlueField-4 for next-generation AI storage platforms, with availability expected in the second half of 2026.
Open Models and Tools for Enterprise and Physical AI
NVIDIA has also released new open models, datasets, and frameworks to support AI across industries. The updates include the Nemotron family for speech, multimodal intelligence, and safety; Cosmos for physical AI; Alpamayo for autonomous vehicles; Isaac GR00T for robotics; and Clara for biomedical applications. These models are supported by extensive open-source datasets, including language tokens, robotics trajectories, protein structures, and vehicle sensor data, providing enterprises with scalable resources for AI development.
Adoption by leading companies illustrates operational relevance: Bosch leverages Nemotron Speech for in-vehicle AI, ServiceNow applies it for multimodal enterprise workloads, Palantir integrates Nemotron into ontology-driven AI frameworks, and Franka Robotics uses Isaac GR00T for robotics training and validation. Alpamayo and Cosmos models support reasoning-based autonomous systems, while Clara models accelerate drug discovery, protein design, and RNA modeling. NVIDIA’s open-source datasets and training resources are accessible via GitHub, Hugging Face, and cloud-based NVIDIA platforms, with deployment supported on NVIDIA NIM microservices.
Implications for Enterprise AI
Collectively, these announcements underscore a broad enterprise AI strategy: delivering secure, scalable, and high-performance infrastructure, extending AI-native storage, and providing open models and data to accelerate development across industries. For enterprises, the combination of validated partner software, BlueField acceleration, and open AI resources provides a framework to deploy large-scale, reliable, and governable AI operations while improving efficiency and operational oversight.