Nutanix dual-native AI architecture bridges virtual machines and containerized workloads

This release addresses the technical friction of running modern AI agents alongside legacy virtualized infrastructure, a common bottleneck for streaming platforms managing hybrid cloud environments. By unifying the control plane for VMs and containers, Nutanix reduces the need for data migration or costly re-architecting of existing workloads. In the broader ecosystem, this positioning directly challenges infrastructure stacks that lock users into specific hardware or siloed management tools. As streaming providers increasingly integrate agentic AI for metadata and personalization, the ability to scale inference without adding architectural complexity becomes a competitive necessity. Watch for the general availability of NKP 2.19 to see if the bare-metal automation gains traction among high-throughput video service providers.

Nutanix has been building its Enterprise AI portfolio aggressively over the past year, positioning itself against hyperscalers and other hybrid-cloud vendors for on-premises AI workloads. In May 2025, Nutanix announced a strategic partnership with NVIDIA to deliver AI factories for enterprise customers, combining Nutanix's hyperconverged infrastructure with NVIDIA's accelerated computing to simplify deployment of large language models and inference pipelines. That partnership directly supports the Enterprise AI 2.8 release by providing the GPU-accelerated foundation on which the new Agent Gateway and dual-native management layer operate. Nutanix also expanded its collaboration with Microsoft Azure to offer Nutanix Cloud Clusters on Azure in early 2025, giving customers a unified control plane across on-premises and public cloud environments, a capability that aligns with the hybrid workload management goals of the latest release.

On the business and competitive front, Nutanix faces pressure from VMware's post-Broadcom acquisition strategy and from Red Hat OpenShift's growing footprint in containerized AI deployments. Broadcom completed its acquisition of VMware in November 2023 and subsequently restructured VMware's licensing model, prompting many enterprises to evaluate alternatives for virtualized workloads. Nutanix has positioned itself as a migration destination for displaced VMware customers, and the dual-native architecture in Enterprise AI 2.8 strengthens that pitch by offering a single platform for both legacy VMs and modern containerized AI agents. Nutanix reported fiscal year 2025 revenue of approximately $2.4 billion, with subscription and support revenue growing 14% year over year, signaling sustained demand for its hybrid cloud platform even as the market consolidates around fewer vendors.

From a technical standpoint, Nutanix Kubernetes Platform 2.19's bare-metal automation targets latency-sensitive workloads that streaming platforms and content delivery operators require for real-time inference. Nutanix introduced support for GPU-direct storage and NVIDIA A100 and H100 GPUs in its platform during 2024, reducing data movement overhead for model serving at scale. Independent benchmarking from the Tolly Group found that Nutanix's hyperconverged infrastructure delivered up to 40% lower total cost of ownership compared to traditional three-tier architectures for AI inference workloads, a finding that supports the economic case for consolidating VM and container management under a single control plane. For streaming operators evaluating agentic AI for metadata enrichment, content recommendation, and automated quality control, the ability to run inference endpoints alongside existing virtualized encoding and packaging pipelines without re-architecting represents a meaningful operational advantage.