Nutanix Transforms: From Hyperconvergence to a Full Stack Platform for Hybrid Cloud and AI
There is one word that is hardly ever heard at .NEXT 2026 in Chicago: hyperconvergence. This absence is significant because Nutanix is the company that invented that concept just over ten years ago. Yet, during three days of conference, from April 7 to 9, the word appeared only to recount the past. The present and future speak of something else: a platform for hybrid cloud, Kubernetes on bare metal, infrastructure for agent-based AI, tools for transitioning from VMware.
A slide from the keynote summarized the positioning better than any statement. On three axes, virtual machines, Kubernetes, and AI, and the ability to operate anywhere (on-premise, edge, public cloud), Nutanix presented itself as the only vendor with full coverage. Broadcom remains strong on VMs but with the lock-in constraint of VCF and a limited cloud presence. IBM/Red Hat has Kubernetes and AI but arrived late to virtualization. The hyperscalers cover containers and AI but with the reverse constraint: cloud lock-in. This is a simplified representation, of course, but it clearly captures Nutanix's ambition: to be the cross-platform that none of the others can be.
Reality, as often happens in enterprise IT, is more nuanced than the slides. With all the competitors mentioned, there are also forms of collaboration: Nutanix runs on hyperscalers with NC2, integrates with partner storage such as Dell and NetApp, and builds joint solutions with Cisco. It’s the model of coopetition, where companies compete and collaborate simultaneously, a dynamic that is becoming structural in the infrastructure market.
From appliance to platform: a sought change, not a suffered one
To understand how deep the change is, it’s worth listening to those who have experienced it from the inside. Christian Turcati, Senior Director Presales System Engineering for Southern Europe at Nutanix, in our video interview recounts a trajectory that is not made of breaks but of continuity: "Nutanix was founded with the idea of simplifying the challenges that could limit the development of technological solutions in companies". In other words, hyperconvergence was not the end but the means: a format that the market could quickly understand and adopt, because hardware was still the language used to sell IT infrastructure. However, the vision was already of an entirely software-defined platform.
The transition happened gradually. First, the separation between software and hardware, with the opening to third-party servers. Then the extension to the public cloud with NC2, which brought the Nutanix platform to AWS, Azure, and Google Cloud. Finally, in the last year and a half, an impressive acceleration on hardware and storage partnerships: Dell PowerFlex, Everpure (formerly Pure Storage), and now NetApp with the integration of ONTAP into the platform, announced precisely at .NEXT.
And then there's the VMware factor. Since Broadcom completed the acquisition at the end of 2023, the virtualization market has been reshuffled. What strikes about Nutanix is not so much having intercepted the demand for an alternative, but the speed with which it has put the tools in place to respond: in less than two years it has built a migration path that eliminates the need to copy the data of virtual machines, created a multi-tenant platform for service providers with Service Provider Central, closed storage integration partnerships with Dell, Everpure, and NetApp, and launched a commercial program with discounted subscriptions to facilitate the transition. This execution capability speaks volumes about the maturity of the underlying platform: if you can add features at this pace, it means that the software architecture allows it. The technical and commercial details are in the dedicated news.
Dan Ciruli, VP and GM of Cloud Native at Nutanix, frames the migration from VMware in a broader context: "We are not just telling customers that we have a way to run the infrastructure they’ve been using for ten years. We want to ensure they are in a place where their infrastructure is ready for the next ten years, with Kubernetes, with AI". The message is that migration is not just a way out of VMware but an opportunity to move to a platform that also covers workloads that VMware has never managed natively.
Containers, agent-based AI, and the role of neoclouds
On the technical announcements front, what best describes where Nutanix is headed is NKP Metal: Kubernetes directly on physical hardware, without going through a hypervisor. Kubernetes on bare metal is not a new concept, but no other vendor today offers it alongside classical virtualization under a single operating model. Nutanix calls this approach "dual-native" architecture: those managing the infrastructure can handle containers and VMs with the same console, the same policies, and the same data services, choosing the most suitable format for each workload. As Ciruli says in our interview: "Virtual machines will never disappear. If you work in IT today, you will need to manage both new container applications and VM applications, forever". NKP Metal is in early access and will be generally available in the second half of 2026.
The other strategic axis is artificial intelligence. Nutanix Agentic AI, previewed at NVIDIA GTC and presented at .NEXT with the complete architecture, is a full-stack platform to run autonomous agents on enterprise infrastructure. Debo Dutta, Chief AI Officer at Nutanix, in our interview, uses an effective image to explain the company's role: "We want to be the best 'plumbers' of AI. The factory produces tokens; a good 'plumbing' system allows tokens to be produced more efficiently and quickly". In other words, the infrastructure is not the protagonist but what makes everything else possible: security, isolation, governance of access to data and models, cost management.
In our interview, Dutta focuses extensively on the challenge of security. Agents like OpenClaw, the open-source framework for long-lived autonomous agents, run for hours without supervision, access sensitive data, and call external tools through protocols like MCP. Keeping them in check requires sandboxes, micro-segmentation of the network, and granular governance of each individual access: a problem of infrastructure, not an application, and it is precisely the ground on which Nutanix believes it can make a difference. This theme becomes even more relevant when the infrastructure is shared among multiple clients, as in the case of neoclouds, the next-generation cloud providers specializing in GPUs. Here, too, Nutanix's execution speed is evident: at .NEXT it presented an already structured offering for these operators, with a strong separation between customers hosted on the same physical infrastructure and consumption measurement tools designed for those who need to bill consumption-based AI services.
Completing the picture is a profound update of the Nutanix Cloud Platform, touching on storage, data protection, operational management, and sovereignty. Each single component (NUS 5.3 for storage, Data Lens 2.0 for on-premise anti-ransomware protection, NCM 2.0 for multi-site management) would deserve a separate analysis, and for this reason, we refer to the dedicated news on the platform. What is worth noting here is the overall volume of new features: if we sum up the platform updates, the integrations with five server manufacturers, the storage partnerships with Dell, Everpure, and NetApp, and the expansion towards sovereign clouds on AWS and Google Cloud, the result is what Nutanix defines as the broadest ecosystem in its history. And it is an ecosystem built largely in the last eighteen months, confirming the underlying thesis: the execution capability is, today, as distinctive a feature of Nutanix as the platform itself.
If there is a thread holding together all that Nutanix showcased in Chicago, it is not just the consistency of a vision but the speed with which that vision has translated into products. Hyperconvergence opened the door. Software-defined allowed a break from proprietary hardware. Hybrid cloud broadened the scope. But it is in the last two years that the pace has changed: the number of closed partnerships, the speed with which migration tools from VMware were built, and the launch of a complete AI platform tell the story of a company that does not just have a strategy but is able to implement it. The challenge now is to demonstrate that this breadth holds up at scale, in production and over time.