At KubeCon + CloudNativeCon Atlanta, Mirantis unveiled Mirantis k0rdent Enterprise with integrated virtualization and unified lifecycle management for cloud, on-premises, and edge workloads. The platform consolidates virtual machines, containers, and AI applications under one interface, lowering infrastructure costs and simplifying operations through automation. Its Distributed Resource Balancer automatically reallocates CPU and memory capacity to maintain stable, consistent performance, while GPU passthrough support maximizes acceleration for machine learning and AI workloads.
Table of Contents: What awaits you in this article
Mirantis k0rdent Enterprise integrates virtualization into open-source Kubernetes management
Mirantis k0rdent Enterprise expands the open-source k0rdent Kubernetes management platform by integrating extensive virtualization capabilities. Platform engineering teams leverage declarative automation, GitOps workflows, and validated Mirantis ecosystem templates to define consistent, strictly policy-driven infrastructures. By incorporating CNCF sandbox distribution k0s, solution accelerates deployment and operation of distributed container and virtual machine workloads. This unified approach simplifies lifecycle management, reduces manual configuration errors, enhances reliability across cloud, on-premises, edge production environments.
Automated Resource Balancing Ensures Consistent Performance Without Manual Planning
The Distributed Resource Balancer (DRB) continuously monitors server clusters to detect unused CPU and memory resources across nodes. By automatically redistributing workloads, it maintains consistent application performance without manual intervention or elaborate capacity planning. Peak usage imbalances are seamlessly absorbed as the DRB shifts compute tasks toward underutilized servers in real time. This intelligent orchestration optimizes utilization operations, allowing teams to focus on development than infrastructure management while reducing overhead.
Mirantis k0rdent Enterprise Optimizes AI With Preserved NUMA Topology
In Mirantis k0rdent Enterprise aligns with Mirantiss AI reference architecture to support machine learning workflows. It retains NUMA topology for optimal memory locality across CPU sockets, while GPU passthrough ensures direct access to accelerators, maximizing computational throughput. Hugepages reduce memory fragmentation and overhead, and CPU pinning reserves specific cores for critical processes. This holistic approach delivers consistent low-latency performance, efficient, automated resource utilization, and predictable scaling for AI training and inference operations.
Unified platform scales virtualization across cloud, on-premises, edge environments
By seamlessly extending virtualization across public clouds, on-premises facilities, and edge environments, this platform enables organizations to deploy workloads where they are most effective. Its unified management interface offers granular visibility into both distributed and centralized application infrastructures, ensuring operational oversight. Real-time cost transparency empowers teams to identify optimization opportunities, while integrated automation minimizes manual intervention. Consequently, enterprises achieve significant reductions in overhead expenses, eliminate data silos, and streamline delivery.
Mirantis k0rdent Enterprise integrates KubeVirt 1.6 and Ceph Squid
Mirantis k0rdent Enterprise integrates the latest KubeVirt 1.6 and Ceph 19.2.3 Squid, providing a powerful software-defined storage solution that meets the demands of modern data centers. By leveraging advanced virtualization capabilities, it streamlines resource provisioning and management across heterogeneous environments. Additionally, the new Mirantis Kubernetes Engine for k0rdent (MKE 4k) enables concurrent execution of virtual machines alongside containers. This approach delivers a fully open-source virtualization alternative to proprietary hypervisor technologies.
Mirantis k0rdent & MKE unite containers, VMs, AI workloads
Mirantis k0rdent Enterprise and MKE 4k integrate containerized workloads, virtual machines, artificial intelligence applications in a unified control plane. They deliver automated resource allocation across compute and memory, apply GPU passthrough and NUMA-aware optimizations for AI tasks, and provide GitOps-based lifecycle management. Organizations benefit from reduced operational overhead, minimized infrastructure silos, predictable performance. Hybrid cloud, on-premises and edge environments gain simplified orchestration, cost transparency and scalable architecture for workload demands.
