Over the last decade, cloud-native technologies have reshaped the way we think about deploying and managing workloads. Containers, microservices, GitOps, and serverless computing have become the dominant vocabulary for building and scaling modern applications. As this cloud-native wave matured, some started to ask: does virtualization – the foundation of cloud computing for the past two decades – still matter?
This is not just a technical question. It reflects deeper shifts in how organizations build and operate their infrastructure, how teams collaborate, and how businesses align IT strategy with their broader goals.
Virtualization: the invisible layer
Virtualization is sometimes perceived as “old school,” a legacy technology from the era before Kubernetes. Yet it remains the invisible backbone of most compute infrastructures. Hyperscalers still run enormous fleets of virtual machines underneath their container platforms. Serverless computing lives also on the shoulders of virtual machines. Enterprises continue to rely on VMs for critical workloads because of the strong isolation, predictable performance, and mature management capabilities they offer.
The reality is that virtualization has not disappeared; it has simply become less visible. Instead of being the star of the show, it has turned into a foundational utility – like the power grid – enabling the layers above it.
Virtualization and cloud-native: friends, not foes
Rather than being competitors, virtualization and cloud-native technologies often complement each other.
Some examples:
- Security and isolation: VMs provide strong workload boundaries for untrusted or multi-tenant environments, particularly at the edge or in regulated sectors.
- Hardware lifecycle management: Virtualization simplifies hardware refreshes, migrations, and mixed-generation clusters.
- Legacy workload support: Not every application is ready to be containerized. Running these workloads alongside containerized ones in the same environment is often a transitional necessity.
- Hybrid and multi-cloud strategies: VMs provide a common denominator across public, private, and edge environments, making it easier to build consistent hybrid deployments.
Rather than seeing virtualization as a relic, we can think of it as the substrate for innovation – a stable foundation on which cloud-native platforms evolve.
From infrastructure-centric to developer-centric
The cultural shift driven by DevOps and GitOps has been one of empowerment: giving developers more autonomy and reducing the cognitive load of dealing with infrastructure.
This cultural evolution is one reason why the virtualization layer receded from view – developers don’t need to think about it anymore.
However, platform engineers, SREs, and infrastructure teams still have to care. Efficient virtualized infrastructure means more predictable performance for clusters, better density, and lower energy consumption. In the age of sustainable computing, the operational efficiency of the underlying virtualization layer is coming back into focus.
The next frontier: convergence at the edge
One of the fastest-growing frontiers for cloud-native adoption is the edge: telco networks, industrial IoT, and distributed retail. These environments often run a mix of VMs and containers.
For example, CNFs (cloud-native network functions) may need to coexist with older VNFs (virtualized network functions) on the same nodes. Resource-constrained environments demand hypervisors that are lightweight, efficient, and adaptable.
At the edge, the collaboration between virtualization and cloud-native is not just pragmatic – it is essential.
Key takeaway: the value of choice
The conversation about whether to “abandon virtualization” in favor of cloud-native is often framed as a zero-sum game. Reality is far more nuanced. Organizations benefit from having layers of abstraction that address different needs:
- Virtualization ensures stable, predictable, and secure compute resources.
- Cloud-native platforms accelerate application development, scaling, and resilience.
Far from being outdated, virtualization continues to evolve – embracing modern hardware (GPUs, accelerators), improving performance and sustainability, and exposing APIs that integrate smoothly with cloud-native tooling.
In this sense, the question is not whether virtualization belongs in the future, but how we ensure that it integrates seamlessly with the cloud-native ecosystem so that teams can choose the right tool for the right workload.
KubeCon + CloudNativeCon North America 2025 is taking place in Atlanta, Georgia, from November 10 to 13. Register now.
##
ABOUT THE AUTHOR
Nathana�l Hannebert, DevOps engineer, Vates
Passionate about IT for over fifteen years, Nathana�l has worked as a system and network administrator, then as a DevOps engineer. Since 2024, he has been working at Vates, a French company specializing in open-source virtualization.
In parallel, he has developed several Android and iOS applications as a contributor, maintainer, either as a freelancer or a volunteer (using Flutter). Committed to privacy issues, he hosts online services for his community as part of the CHATONS movement and showcases them on isidorus.fr.
