Mirantis Powering the Next Generation of AI Factories with NVIDIA DSX OS

The AI infrastructure market is at an inflection point: raw compute is commodifying, agentic AI is burying manual operations under massive token volumes, and new hardware generations are arriving faster than architectures can absorb them. In this environment, software is the differentiator. NVIDIA AI Clouds are building the next generation of AI factories to operate GPU infrastructure efficiently at scale, deliver differentiated AI services, and evolve their architectures continuously without replatforming from scratch with each hardware cycle. 

To help service providers meet these demands, Mirantis is collaborating with NVIDIA to provide a validated, production-ready software stack, built on open source and open standards. The solution combines NVIDIA DSX OS and Mirantis k0rdent AI to deliver a turnkey, open foundation for building and operating lasting, profitable, AI factories, with the ingrained flexibility to evolve and iterate with rapidly changing technologies and customer requirements.

An Ecosystem-First OS for the AI Factory

NVIDIA DSX OS is open-source, modular software for building, operating, and scaling AI factory infrastructure. As part of the NVIDIA DSX platform, DSX OS delivers composable components for lifecycle management, runtime consistency, health automation, resiliency, multi-tenant operations, and AI platform services – designed to integrate into partner control planes, infrastructure platforms, and AI cloud service stacks. DSX OS is designed to deliver on three outcomes that drive AI factory economics: faster time to revenue, better efficiency, and higher reliability and resiliency. 

NVIDIA has strategically made DSX OS open source and ecosystem-first, offering service providers flexibility and choice. NVIDIA AI Clouds can adopt components incrementally, integrate their own solutions, and build differentiated services on top of a validated, open foundation.

Mirantis is a proud inaugural NVIDIA AI Cloud-Ready ISV partner, and we offer k0rdent AI, a complete, full-stack AI cloud infrastructure platform that eliminates management complexity and enables service providers to focus on delivering value to their customers. Mirantis has integrated NVIDIA Infra Controller (NICo) into k0rdent AI to help AI-Native Clouds and NCPs streamline end-to-end AI infrastructure operations, from hardware lifecycle management through to AI services delivery.  Validated against Hopper and Blackwell architectures, the integration enables API-driven bare-metal lifecycle automation and hardware-enforced tenant isolation through NVIDIA BlueField DPUs and NVIDIA DOCA. By embedding NICo into our core, k0rdent AI provides a unified control plane across the full NVIDIA reference architecture portfolio so service providers can simply plug in NVIDIA-certified hardware, and let the platform manage it end-to-end. 

For intelligent AI workload scheduling and platform services, Mirantis has integrated NVIDIA Run:ai into k0rdent AI to help service providers accelerate the delivery of AI services while also maximizing GPU utilization. The certified integration enables on-demand AI factory provisioning, allowing operators to spin up a full Run:ai environment when needed and tear it down when workloads complete. By automating the AI factory infrastructure stack, from bare metal provisioning through platform deployment, k0rdent AI helps bring NVIDIA Run:ai and the workloads it supports online as quickly as possible.

Mirantis k0rdent AI is purpose-built to accelerate the delivery of production AI factories, integrating NVIDIA DSX OS components to offer live, revenue-generating AI cloud service as fast as possible. It delivers multi-tenanted GPU clouds, full-stack automation from the bare metal up, dynamic workload management across hardware generations, and the governance and observability capabilities that enterprise customers increasingly require. As NVIDIA reference architectures evolves, the k0rdent AI stack adapts with it, rather than requiring re-platforming with each hardware cycle.  This partnership scales customer optionality alongside k0rdent’s validated integrations with Dell, Supermicro, BE Networks, Netris, Saturn Cloud, VAST Data, and others for a fully open, end-to-end GPU ecosystem.

The Open Standard for AI Infrastructure Is Being Established Now

Establishing an open standard is not just a matter of good architecture, it requires validation at real-world scale. Mirantis is working with IREN, one of the world's leading vertically-integrated AI cloud providers, to build a major reference implementation of open, standards-based AI infrastructure, across thousands of GPUs. The result will be a production-scale deployment of a modern AI factory, built on validated, open-source foundations, and available as a referenceable architecture for the broader NCP ecosystem.

This is the same dynamic that has driven every prior technology wave. Open standards, built at scale and available to everyone, accelerate the whole market. We saw it with operating systems, with mobile, and with cloud, where Kubernetes emerged as the open standard against proprietary hyperscaler APIs. AI infrastructure is the next wave, and that same convergence is already underway. The open standard for AI infrastructure has yet to be established, and that is precisely what IREN and Mirantis intend to build together.

To learn more about how Mirantis k0rdent AI and NVIDIA DSX OS can accelerate your AI factory deployment, contact us.