From Electrons to AI: IREN acquiring Mirantis to strengthen its AI cloud platform

IREN is joining forces with Mirantis, a provider of cloud infrastructure, Kubernetes-based orchestration and enterprise support services—and the team behind the k0rdent AI platform. The acquisition of Mirantis builds on IREN’s existing software, engineering and customer support capabilities, enhancing how compute is deployed, managed and operated for customers. 

Mirantis brings more than a decade of experience and a track record of over 1,500 enterprise customers globally, along with founding Independent Software Vendor status in the NVIDIA AI Cloud Ready Initiative. It is expected to operate as a standalone subsidiary, serving its existing customer base while supporting IREN Cloud.

Power, land, and time to compute 

AI infrastructure is, before anything else, an industrial business. The constraint limiting AI today is time-to-compute: electrons, land, procurement, and execution.

Meanwhile, models are getting larger. Reasoning workloads burn through tokens at exponential rates. Today’s AI builders need providers who deploy compute quickly, at scale, and across a broad range of requirements.

That is the foundation IREN is built on. Vertically integrated ownership. Purpose-built data centers designed for high-density GPU deployments. Grid-connected power that compresses the time between development and operational compute. Liquid-cooled facilities sized for the next generation of accelerators.

Dan Roberts, Co-Founder and Co-CEO of IREN, in today's announcement:

“IREN’s core advantage is execution — from securing power to building data centers, deploying GPUs and bringing compute online at scale. Mirantis builds on our existing capabilities and strengthens how that compute is deployed, managed and operated for customers.”

What Mirantis brings to IREN 

Mirantis brings the software, engineering and customer support expertise that turn deployed capacity into reliable, production-grade compute. The acquisition strengthens IREN’s AI Cloud platform across the four areas highlighted in today's announcement:

Deployment capability

The k0rdent AI platform supports deployment and operation of customer workloads on IREN's existing bare metal GPU infrastructure, with consistent automation across bare metal, virtual machines, and Kubernetes environments. It is hardware-agnostic across the major accelerator families and engineered for hard multi-tenancy from the ground up—the foundation for running a large, heterogeneous GPU estate as a multi-tenant cloud.

Operational visibility

k0rdent AI ships with full-stack observability spanning the network, bare metal, GPUs, and inference workloads, plus AIOps-driven reconciliation that catches issues before they become incidents and automates remediation where possible. The result is faster deployment of new capacity and tighter visibility into how that capacity is performing in production.

Customer support

Mirantis brings a global enterprise support and service delivery organization built over more than a decade of work with 1,500+ customers. For IREN, this means AI Cloud customers across a wider range of workload profiles can be onboarded with the operational expertise required to run production AI workloads at scale.

Market access

The acquisition expands the ways IREN can put deployed capacity to work across AI natives and developing enterprise markets. By combining IREN's GPU footprint with k0rdent AI's roadmap for tenant-level governance and workload-level metering, IREN can serve a broader spectrum of customer requirements from a single infrastructure base.

A standalone business, a continuing roadmap

The k0rdent AI roadmap continues‚ including Mirantis’ founding ISV status with the NVIDIA AI Cloud Ready Initiative. For Mirantis customers, the acquisition enables continuity.

Alex Freedland, Founder and CEO of Mirantis, in today's announcement:

“Mirantis has spent more than a decade helping enterprises deploy and manage cloud infrastructure. AI is creating a new set of customer requirements, and customers need platforms that are open, flexible and built for scale. IREN brings infrastructure at scale and proven delivery capability. Mirantis adds software and operational expertise that strengthens how customers deploy and use that infrastructure. Together, we will bring AI infrastructure online faster, while continuing to support existing customers and advance the k0rdent AI platform.”

How k0rdent AI works 

Mirantis k0rdent AI is an integrated platform designed to help service providers and enterprises provision multi-tenanted, AI-ready infrastructure and core services. It leverages Kubernetes as a universal control plane across clouds, edge, and on-prem—automating policy, reducing toil with self-service, and enabling AI/ML and other modern applications at scale.

The k0rdent AI stack covers four integrated layers: 

Layer 1: Infrastructure as a Service (IaaS) 

Bare metal server discovery, provisioning, and lifecycle management via the k0rdent API and UI, supporting iLO, iDRAC, and BMC (Redfish) interfaces across Ubuntu, Rocky Linux, and RHEL, with NVIDIA NCX Infra Controller integrated for physical layer hardware control. AI-optimized virtualisation enables workload partitioning and efficient bin packing across the compute estate. GPU provisioning includes MIG-based slicing, driver and operator deployment, and optimized VM configuration. Network and storage are both template-based, with tenant isolation, monitoring, and observability built in; storage supports pluggable third-party frameworks.

Layer 2: Container Services 

Provisioning and lifecycle management of Kubernetes clusters providing consistent, secure deployment across all layers of the solution.

Layer 3: AI PaaS Services 

On-demand Kubernetes clusters preconfigured with SLURM for HPC workloads, MLflow for ML lifecycle management, and Kubeflow for ML pipelines. Includes managed Training as a Service, a Model Catalog with versioned local storage and remote retrieval from services such as Hugging Face, and turnkey Inference as a Service with selectable operators for shared or isolated deployments.

Layer 4: Billing, Reporting, and Cloud Portal 

Comprehensive metering and invoicing APIs cover all infrastructure and AI services, with external payment gateway integration and detailed cost, usage, and compliance reporting. Operators manage the platform via an administrative UI; tenants access a self-service cloud console with marketplace-style service selection and invoice management. Turnkey dashboards provide real-time AI infrastructure visibility including GPU and network telemetry with automated alerting.

What's next

Customers can access GPU capacity on IREN Cloud today, with additional k0rdent AI capabilities rolling out across the platform soon.

Links:

[Link: IREN AI Cloud, documentation] 

[Link: Mirantis k0rdent AI‚ reference architecture] 

Press Release

*The acquisition was announced on May 4, 2026. Closing remains subject to customary conditions, including required regulatory approvals.*