Delivers performance, governance, and agility required by AI workloads
WASHINGTON–(BUSINESS WIRE)–#AI—NVIDIA GTC – Mirantis, delivering Kubernetes-native infrastructure for AI, today announced that it is enabling organizations to build next-generation infrastructure with NVIDIA BlueField Data Processing Units (DPUs) that meet the performance, governance, and agility required by AI workloads.
Mirantis is one of the first design and integration partners to deliver secure multi-tenant networking for AI with NVIDIA BlueField. By combining the declarative, Kubernetes-native capabilities of Mirantis k0rdent AI with the high-performance data processing and isolation features of NVIDIA BlueField, organizations can build infrastructure that is not only optimized for AI but also governed, flexible, and cost-effective.
“AI workloads demand precision through specialized hardware, tightly-integrated networking, secure multi-tenancy, and real-time provisioning, all orchestrated efficiently and at scale,” said Shaun O’Meara, chief technology officer, Mirantis. “With the integration of Mirantis k0rdent AI and NVIDIA BlueField, we’re delivering multi-tenant Kubernetes with strict isolation at all layers that provides support for complex networking and isolation, while still providing tenants full access to their Kubernetes cluster.”
Delivering infrastructure for AI high-performance computing brings unique challenges requiring specialized fine-tuning and setup. These include allocating and slicing GPUs, enabling RDMA networking, meeting advanced scheduling needs, tuning for performance, and efficiently scaling platforms like Kubernetes. The complexity increases when solutions must support multiple, dynamic workloads shared across different users.
Mirantis k0rdent AI and NVIDIA BlueField deliver:
- Multi-tenancy support for data security and resource sharing;
- Data sovereignty for AI workloads that are driven by confidential data that often contain proprietary models;
- Ability to share GPUs and other key compute resources;
- Strict network isolation at the network interface level with tenants that only have access to the interfaces assigned to them, and those interfaces only have access to the VLANs assigned to the tenant;
- High-speed access to data storage;
Mirantis is also preparing for support of the newly announced NVIDIA BlueField-4, which offers six times increased compute powers and doubled network throughput compared to BlueField-3. Mirantis expects to simplify underlying switch configurations and provide real-time risk detection.
The integration of Mirantis k0rdent AI and NVIDIA BlueField provides flexible, repeatable patterns for service delivery to end users via templates, so infrastructure can be deployed consistently and configuration drift is prevented. This overall approach allows for the automation of infrastructure provisioning as a service.
For more information, read the blog post, Mirantis and NVIDIA BlueField Next-Generation and visit us at NVIDIA GTC, Washington, DC, October 27–29, 2025, booth 634.
About Mirantis
Mirantis delivers the fastest path to enterprise AI at scale, with full-stack AI infrastructure technology that removes GPU infrastructure complexity and streamlines operations across the AI lifecycle, from Metal-to-Model. Today, all infrastructure is AI infrastructure, and Mirantis provides the end-to-end automation, enterprise security and governance, and deep expertise in Kubernetes orchestration that organizations need to reduce time to market and efficiently scale cloud native, virtualized, and GPU-powered applications across any environment – on-premises, public cloud, hybrid, or edge.
Mirantis serves many of the world’s leading enterprises, including Adobe, Ericsson, Inmarsat, MetLife, PayPal, and Societe Generale. Learn more at www.mirantis.com.
Contacts
Joseph Eckert for Mirantis
[email protected]
