Mirantis, a Kubernetes-native AI infrastructure company based in Campbell, California, announced the Mirantis AI Factory Reference Architecture, the industry’s first comprehensive framework for building, operating, and optimizing AI and ML infrastructure at scale. Built on Mirantis k0rdent AI, the architecture provides a secure, composable, scalable, and sovereign platform tailored to enterprise AI workloads.
Industry’s first reference architecture for AI infrastructure, launched June 17, 2025.
Built on Mirantis k0rdent AI, supporting NVIDIA, AMD, and Intel accelerators.
Enables AI workload deployment in days, with rapid prototyping and deployment.
Addresses high-performance computing challenges like RDMA networking and GPU slicing.
Supports multi-tenancy, data sovereignty, and compliance for AI/ML workloads.
Composable templates for compute, storage, GPU, and networking layers.
The Mirantis AI Factory Reference Architecture leverages Kubernetes to support diverse AI workloads—training, fine-tuning, and inference—across dedicated servers, virtualized environments (KubeVirt/OpenStack), public/hybrid clouds, and edge locations. Key features include:
Rapid Deployment: Deploys AI workloads within days using k0rdent AI’s templated, declarative provisioning model.
Accelerated Development: Shortens the AI/ML lifecycle with faster prototyping, iteration, and model deployment.
Curated Integrations: Offers a k0rdent Catalog with open-standard tools for AI/ML, observability, CI/CD, and security.
High-Performance Computing: Tackles complex requirements like RDMA networking, GPU allocation/slicing, and Kubernetes scaling.
Platform Services: Integrates with Gcore Everywhere Inference and NVIDIA AI Enterprise ecosystems.
Shaun O’Meara, Mirantis CTO, stated, “We’ve built and shared the reference architecture to help enterprises efficiently deploy and manage large-scale multi-tenant sovereign infrastructure solutions for AI and ML workloads.”
Unlike cloud-native workloads designed for multi-core scale-out, AI workloads often demand aggregating multiple GPU-based servers into a single supercomputer with ultra-high-performance networking (e.g., Infiniband, NVLink, RoCEv2). The architecture addresses these unique challenges:
Fine-Tuning and Configuration: Simplifies complex setup processes for AI systems.
Hard Multi-Tenancy: Ensures data security, isolation, and resource allocation for multiple tenants.
Data Sovereignty: Controls sensitive AI/ML data and intellectual property, critical for compliance with regional regulations like GDPR.
Scale and Sprawl Management: Manages distributed infrastructure for edge and large-scale compute systems.
Resource Sharing: Optimizes scarce GPUs and compute resources for cost efficiency.
Skills Accessibility: Designed for data scientists and developers, reducing reliance on IT infrastructure expertise.
The composable design allows users to assemble infrastructure from reusable templates across compute, storage, GPU, and networking layers, tailored to specific workloads with support for NVIDIA, AMD, and Intel AI accelerators.
The architecture integrates advanced technologies to meet AI demands:
Compute and GPU Layer: Supports fractional provisioning and GPU sharing for NVIDIA, AMD, and Intel accelerators.
Storage and Networking: Provides high-throughput NVMe-oF, multi-tiered storage, and AI-optimized networking (RDMA, Infiniband, SmartNICs/DPUs).
Security and Compliance: Implements zero-trust, hard multi-tenancy, confidential computing, and data sovereignty controls.
Automation: Uses Kubernetes for orchestration, with GitOps and policy-driven automation for self-healing and scalability.
Posts on X, such as from @ITOpsTimes, highlight the platform’s ability to deploy AI workloads rapidly, while @MirantisIT emphasized its integration with NVIDIA BlueField for secure GPU-as-a-Service platforms.
The AI infrastructure market, valued at $30 billion in 2024, is projected to grow at a 35% CAGR through 2030, driven by demand for GPU-accelerated computing and data-intensive applications. Mirantis competes with providers like Google Cloud’s Vertex AI and Azure’s AI platforms, but its open-source, Kubernetes-native approach avoids vendor lock-in, appealing to enterprises like Adobe, PayPal, and Societe Generale.
Mirantis aims to expand adoption through partnerships, such as with Gcore, and events like RAISE Summit 2025 in Paris. The architecture’s emphasis on open standards and sovereignty positions it to address enterprise concerns about data control and compliance. Challenges include integrating with legacy systems and competing with cloud hyperscalers’ proprietary AI stacks. Mirantis’ valuation, estimated at $500–700M, could rise with increased enterprise adoption.
The Mirantis AI Factory Reference Architecture, launched on June 17, 2025, sets a new standard for scalable, secure AI infrastructure. By addressing the unique demands of AI workloads, Mirantis empowers enterprises to accelerate innovation while maintaining control over their infrastructure strategy.
Mirantis is the Kubernetes-native AI infrastructure company, enabling organizations to build and operate scalable, secure, and sovereign infrastructure for modern AI, machine learning, and data-intensive applications. By combining open source innovation with deep expertise in Kubernetes orchestration, Mirantis empowers platform engineering teams to deliver composable, production-ready developer platforms across any environment - on-premises, in the cloud, at the edge, or in data centers. As enterprises navigate the growing complexity of AI-driven workloads, Mirantis delivers the automation, GPU orchestration, and policy-driven control needed to cost-effectively manage infrastructure with confidence and agility. Committed to open standards and freedom from lock-in, Mirantis ensures that customers retain full control of their infrastructure strategy.
Mirantis serves many of the world’s leading enterprises, including Adobe, Ericsson, Inmarsat, PayPal, and Societe Generale.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.