KubeVirt Graduates from Cloud Native Computing Foundation
The Cloud Native Computing Foundation (CNCF) has officially announced the graduation of KubeVirt, an open-source project that has been instrumental in extending Kubernetes to run and manage virtual machines (VMs) alongside containers. This achievement comes after ten years of development, marking a significant milestone not only for KubeVirt itself but for organizations looking to unify their cloud-native infrastructure.
KubeVirt was founded by Red Hat in 2016 when the management of VMs and containers was still largely handled as two separate entities with distinct tools, workflows, and expertise. The project was designed to address this fragmentation, allowing teams to manage VMs under the same control plane as their container workloads. It does so by adding VM management capabilities to existing Kubernetes clusters through custom resources and controllers. This means organizations can now leverage the same declarative APIs they use for managing containers to create, schedule, and control VMs.
As a result, KubeVirt has enabled a smoother and less disruptive transition towards modernization by allowing teams to maintain their legacy applications running on VMs while progressively adopting container technologies – all without needing to re-architect their applications outright.
The last decade has seen KubeVirt grow to be a preferred solution among industry giants, including ByteDance, Cloudflare, CoreWeave, and NVIDIA. In a recent survey, 86% of Kubernetes adopters expressed awareness of KubeVirt, and 26% reported utilizing it in production environments. These implementations showcase the flexibility and efficiency that KubeVirt offers, supporting various workloads ranging from cloud gaming to AI cloud infrastructures.
In 2026, KubeVirt has achieved several key milestones. For instance, the first-ever KubeVirt Summit was held at KubeCon + CloudNativeCon Europe, coinciding with the release of KubeVirt v1.8. This version introduced critical features such as a Hypervisor Abstraction Layer, support for confidential VMs with Intel TDX attestation, and enhancements that optimize performance for AI and high-performance computing workloads.
As organizations continue transitioning to cloud-native infrastructure, the ability to run both VMs and containers on a single platform is becoming indispensable. As Chris Aniszczyk, CTO of CNCF, states, “KubeVirt's graduation solidifies it as a Kubernetes-native way to introduce VMs into the platform, providing teams with the reliability required to modernize at their pace.”
KubeVirt's roadmap includes exciting developments aligned with its Kubernetes-native design philosophy, focusing on leveraging existing cluster capabilities rather than developing virtualization-specific solutions. Planned enhancements include:
- - Device and AI Hardware Support: Expanding GPU claims through Kubernetes Dynamic Resource Allocation (DRA) and building on this model for network devices and CPUs to allow more versatile resource requests for VMs.
- - Confidential VMs Support: Expanding operations beyond just Intel TDX, also covering AMD SEV and IBM Secure Execution, as well as exploring additional integrations with attestation services.
- - Extensible Framework: Initiatives to create structured plugins and a pluggable virtualization stack, enabling compatibility with alternatives to KVM over time.
This graduation is not only a validation of over a decade of hard work but also underscores the commitment to providing a robust, cloud-native approach to virtualization that can meet the varying demands of modern workloads. KubeVirt's success story is about bringing legacy systems and cloud-native technologies together without conflict, creating a powerful and simple operational model for organizations navigating their digital transformation journeys.
For more detailed information about KubeVirt, feel free to visit its official site at
kubevirt.io.