Mirantis Integrates with NVIDIA BlueField to Build Next-Generation AI Infrastructure 

/ AI, Application Development, Disaster Recovery, Kubernetes

Enabling secure multi-tenant Kubernetes for AI and accelerating compliant adoption through NVIDIA AI Factory for Government

The News

At NVIDIA GTC DC 2025, Mirantis announced its integration with NVIDIA BlueField Data Processing Units (DPUs) to power next-generation AI infrastructure that meets the stringent performance, governance, and agility requirements of modern workloads.

As one of the first design and integration partners to deliver secure, multi-tenant networking for AI with NVIDIA BlueField, Mirantis is combining its Kubernetes-native platform, k0rdent AI, with NVIDIA’s high-performance data processing capabilities. The result is Kubernetes clusters optimized for AI, offering real-time provisioning, fine-grained isolation, and multi-tenant data sovereignty.

In parallel, Mirantis was also named a collaborator in the NVIDIA AI Factory for Government reference design, extending its work in secure, compliant AI infrastructure for public sector and enterprise environments. Together with NVIDIA, Mirantis will help government agencies deploy mission-critical AI systems that meet federal data privacy, compliance, and security mandates.

Analyst Take

AI workloads have outgrown the limits of traditional compute environments. With the proliferation of agentic and generative AI, enterprises now need infrastructure that combines high-performance compute, data isolation, and governed networking without compromising speed or flexibility.

Mirantis k0rdent AI, integrated with NVIDIA BlueField, directly addresses these needs. The solution aims to let organizations deliver multi-tenant Kubernetes environments with secure, network-level isolation at the VLAN and interface layer, ensuring that tenants can only access their assigned resources. This configuration allows shared use of GPUs and data storage while maintaining strict compliance and privacy standards.

As Shaun O’Meara, CTO of Mirantis, explained, “AI workloads demand precision through specialized hardware, tightly-integrated networking, secure multi-tenancy, and real-time provisioning. With Mirantis k0rdent AI and NVIDIA BlueField, we’re delivering multi-tenant Kubernetes with strict isolation at all layers, while giving tenants full access to their clusters.”

The integration also introduces flexible, repeatable deployment templates to prevent configuration drift and automate infrastructure provisioning, effectively transforming AI infrastructure into a service model.

BlueField Integration Where Performance Meets Policy

Mirantis’s partnership with NVIDIA capitalizes on BlueField’s data processing and security capabilities, which are increasingly viewed as foundational for AI-native data centers.

The BlueField DPU offloads data plane tasks (such as encryption, networking, and telemetry) from CPUs, freeing system resources for AI workloads. When combined with Mirantis’s declarative, Kubernetes-native orchestration, this architecture supports data sovereignty, GPU sharing, and real-time risk detection which are features that matter for AI in regulated sectors.

Looking ahead, Mirantis is preparing for support of NVIDIA BlueField-4, which delivers six times more compute and doubles the throughput of BlueField-3. This next-generation platform will bring expanded automation for switch configuration, enabling self-healing, policy-driven network orchestration across AI clusters.

Accelerating Compliant AI Adoption in Government

In a related announcement, Mirantis revealed its selection as a collaborator in the NVIDIA AI Factory for Government reference design, a major initiative to standardize secure and compliant AI deployments across U.S. federal, state, and local agencies.

The AI Factory for Government provides a blueprint for building sovereign AI environments, bringing together FIPS- and STIG-compliant infrastructure components to help agencies run AI workloads within strict security frameworks.

According to a July 2025 GAO report, 10 of 12 federal agencies cite regulatory barriers (such as data privacy and policy compliance) as major inhibitors to AI adoption. Mirantis and NVIDIA’s collaboration addresses these pain points head-on.

With k0rdent AI, agencies can:

  • Automate provisioning and lifecycle management across compute, network, and GPU resources
  • Enforce federal security standards through built-in authentication, observability, and auditing
  • Deploy validated, compliant configurations in air-gapped or disconnected environments
  • Run inference, generative, and high-performance computing workloads under a unified control plane

As Kevin Kamel, VP of Product Management at Mirantis, stated: “This integration empowers organizations to confidently accelerate their AI initiatives within governed environments.”

A Proven Foundation for Secure AI Infrastructure

Mirantis’s track record of serving defense and federal clients under STIG and FedRAMP-aligned frameworks provides a strong foundation for this partnership. Its full-stack lifecycle management capabilities (spanning Linux, Windows, and GPU workloads) make it a trusted provider for air-gapped and classified environments.

These capabilities align with theCUBE Research’s findings that governance and policy enforcement are emerging as top differentiators in enterprise AI infrastructure strategy. Mirantis’s focus on “AI from Metal-to-Model” gives customers end-to-end visibility, from hardware orchestration through model deployment, bridging the traditional gap between DevOps and MLOps.

Looking Ahead

Mirantis’s integration with NVIDIA BlueField and participation in the AI Factory for Government initiative signal a broader movement toward secure, policy-aware AI infrastructure. As organizations navigate the tradeoffs between openness, compliance, and performance, solutions like k0rdent AI are reshaping the architecture of enterprise and sovereign AI.

Future updates are expected to include:

  • Native support for BlueField-4 and Hopper/Blackwell GPUs
  • Expanded automation via AI-driven network telemetry and risk-based orchestration
  • Deeper alignment with federal data classification standards to enable next-generation secure AI environments

By merging Mirantis’s Kubernetes-native automation with NVIDIA’s hardware acceleration, the two companies aren’t just optimizing AI performance. They’re laying the groundwork for governed, composable, and resilient AI ecosystems across industries.

Key Takeaways

  • Mirantis integrates with NVIDIA BlueField DPUs, delivering secure, multi-tenant Kubernetes for AI infrastructure.
  • k0rdent AI provides declarative orchestration, policy enforcement, and automation from provisioning to lifecycle management.
  • Collaboration with NVIDIA AI Factory for Government accelerates compliant AI adoption across federal and state agencies.
  • Data sovereignty and network isolation ensure AI workloads remain secure, scalable, and compliant.
  • BlueField-4 readiness prepares Mirantis for the next wave of high-performance, policy-driven AI infrastructure.

Author

  • Paul Nashawaty

    Paul Nashawaty, Practice Leader and Lead Principal Analyst, specializes in application modernization across build, release and operations. With a wealth of expertise in digital transformation initiatives spanning front-end and back-end systems, he also possesses comprehensive knowledge of the underlying infrastructure ecosystem crucial for supporting modernization endeavors. With over 25 years of experience, Paul has a proven track record in implementing effective go-to-market strategies, including the identification of new market channels, the growth and cultivation of partner ecosystems, and the successful execution of strategic plans resulting in positive business outcomes for his clients.

    View all posts