Mirantis is attempting to bridge the gap between raw hardware provisioning and sophisticated AI workload management by integrating open-source NVIDIA Run:ai into its k0rdent AI platform. This move aims to provide a unified operational model that spans from bare-metal infrastructure to GPU orchestration within the NVIDIA DSX AI Factory Platform. By combining these technologies, Mirantis seeks to offer cloud service providers and enterprises a method to maximize GPU utilization and workload throughput while maintaining centralized control across the entire AI factory software stack.
Automating the NVIDIA DSX AI Factory Stack
The integration focuses on extending the lifecycle management capabilities of k0rdent AI to include the orchestration functions provided by NVIDIA Run:ai. Because NVIDIA Run:ai is now open source, Mirantis is positioning its ability to integrate the tool more directly into the k0rdent AI environment. This connection allows operators to manage the installation of the Run:ai control plane and its supporting components automatically across Kubernetes clusters. The company is framing this as a way to provide a consistent operational model that covers infrastructure provisioning, cluster lifecycle management, and granular GPU orchestration. By automating these processes, Mirantis suggests that operators can maintain consistent software configurations and manage updates more effectively across distributed AI infrastructure. This technical alignment is intended to help organizations keep GPU capacity available for production workloads by reducing the manual overhead typically associated with managing complex, high-performance computing environments.
Orchestration Capabilities and Managed Services
Mirantis is positioning this integration as a component of a broader suite of NVIDIA-related technologies, including the NVIDIA Infra Controller (NICo), NVIDIA GPU Operator, NVIDIA Network Operator, and NVIDIA AI Cluster Runtime. The goal is to enable a production-ready AI factory stack that handles everything from bare-metal provisioning to network configuration and GPU scheduling. To support these complex deployments, Mirantis is offering two distinct service tiers: OpsCare, which provides 24x7 support, onboarding, and training, and OpsCare Plus, which offers fully-managed, around-the-clock operations. These services are designed to address the operational risks of AI infrastructure, such as configuration errors and service disruptions. The company claims that by standardizing deployment and maintenance, operators can perform predictable fleet-wide updates and accelerate recovery through centralized visibility. This approach is intended to help enterprises preserve infrastructure choice, allowing them to customize their stacks and select preferred monitoring technologies while still benefiting from automated lifecycle management.
Key Takeaways
- Mirantis integrated open-source NVIDIA Run:ai with k0rdent AI to provide orchestration from bare metal to AI workloads.
- The integration automates the installation of the Run:ai control plane and its supporting components across Kubernetes clusters.
- Mirantis offers OpsCare and OpsCare Plus service options to provide 24x7 support and fully-managed operations for AI factory availability.
TechInsyte's Take
In our view, Mirantis is making a strategic play to become the essential management layer for the "AI Factory" era. By linking k0rdent AI directly with NVIDIA’s Run:ai and the broader DSX OS portfolio, they are moving beyond simple Kubernetes management into the high-stakes realm of GPU resource optimization. This signals a shift in the enterprise IT landscape where the primary bottleneck is no longer just software deployment, but the efficient orchestration of expensive, scarce compute assets. If Mirantis can successfully deliver on its promise of "Metal-to-Model" automation, they will provide a critical solution for organizations struggling to turn raw GPU power into predictable, scalable production revenue.
Questions & Answers
How does the integration of NVIDIA Run:ai affect GPU management for enterprise operators?
The integration allows operators to extend lifecycle management from bare-metal infrastructure directly into GPU orchestration. By using k0rdent AI to automate the Run:ai control plane installation, companies can achieve more consistent software configurations and better visibility into workload throughput across distributed clusters.
What specific operational risks is Mirantis targeting with this announcement?
Mirantis is targeting risks such as configuration errors, avoidable service disruptions, and the complexity of fleet-wide updates. The integration aims to reduce these issues by standardizing deployment and automating tasks like TLS certificate renewal to maintain production uptime.
What are the differences between the Mirantis OpsCare service offerings?
OpsCare provides 24x7 support, training, and guidance for platform maintenance and updates. In contrast, OpsCare Plus is a more intensive tier that provides fully-managed, around-the-clock operations performed by Mirantis experts.
Does this integration lock enterprises into a specific monitoring or hardware setup?
No. Mirantis states that the integration is designed to preserve infrastructure choice, allowing operators to customize their AI factory stack, including the selection of their preferred monitoring technologies.
Source: Mirantis
