NVIDIA NCX Infra Controller
Type: Tool Tags: NVIDIA, NCX, infrastructure controller, bare metal, lifecycle management, Kubernetes, DPU, zero trust, AI cloud Related: NVIDIA-Cloud-Accelerator-NCX, NVIDIA-DOCA-Platform-Framework, NVIDIA-BlueField-DPU, NVIDIA-DOCA, NVIDIA-AI-Cluster-Runtime, NVIDIA-Run-ai Sources: https://docs.nvidia.com/ncx/index.html; https://github.com/NVIDIA/ncx-infra-controller-core Last Updated: 2026-04-29
Summary
NVIDIA NCX Infra Controller, also called NICo in its public repository, is an NCX component for bare-metal provisioning and secure lifecycle management in multi-tenant GPU infrastructure. NCX describes it as bare-metal provisioning and secure lifecycle management, while the public repository describes site-local, zero-trust bare-metal lifecycle management with DPU-enforced isolation. It is part of the operator-facing infrastructure layer for NVIDIA-powered AI clouds.
Detail
NICo targets the foundational layer below AI workload orchestration: preparing and managing physical systems in a way that cloud partners can automate. The public README describes it as an API-based microservice that automates bare-metal lifecycle complexity to accelerate next-generation AI cloud offerings.
The repository’s bare-metal cluster setup describes an NCX stack deployed onto Kubernetes in layers: common services, Carbide Core / NCX Infra Controller, and Carbide REST/API/site-agent services. That makes NCX Infra Controller adjacent to NVIDIA-AI-Cluster-Runtime, NVIDIA-DOCA-Platform-Framework, NVIDIA-BlueField-DPU, and zero-trust DPU-enabled isolation.
Connections
- NVIDIA-Cloud-Accelerator-NCX - NCX lists NCX Infra Controller as a core software component.
- NVIDIA-DOCA-Platform-Framework - both are infrastructure-level NCX components for DPU-accelerated cloud environments.
- NVIDIA-BlueField-DPU - NICo’s zero-trust positioning depends on DPU-enforced isolation.
- NVIDIA-DOCA - DOCA is the software substrate for DPU-accelerated infrastructure services.
- NVIDIA-AI-Cluster-Runtime - AICR defines reproducible GPU Kubernetes runtime recipes above the infrastructure lifecycle layer.
- NVIDIA-Run-ai - Run:ai schedules workloads on the AI cloud substrate that NICo helps provision.
Source Excerpts
- “Bare-metal provisioning and secure lifecycle management for multi-tenant GPU infrastructure.”
- “zero-trust, bare-metal lifecycle management with DPU-enforced isolation”