Nutanix unveils AI 2.8 & Kubernetes platform update

Nutanix has released Nutanix Enterprise AI 2.8, unveiled partner programmes tied to artificial intelligence services and said Nutanix Kubernetes Platform 2.19 is nearing general availability.

The update extends Nutanix Cloud Platform with tools to run AI workloads alongside existing applications and data across virtual machines and containers. Many enterprises still split critical systems between virtualised and containerised environments, which can complicate AI deployment and governance, Nutanix said.

Nutanix Enterprise AI 2.8 is available now and adds centralised controls for AI inference and agentic AI. It also makes the Model Context Protocol gateway in Nutanix Agent Gateway generally available, to govern how software agents connect to applications and data.

The release also expands Nutanix Private Inference with support for fine-tuning smaller language models using Low-Rank Adaptation, multi-GPU inference through tensor parallelism, batch inference and speculative decoding to reduce latency.

Security is another focus. The new version includes fine-grained identity and access management, custom roles, model sharing controls and support for air-gapped NVIDIA NIM deployment, as companies seek to limit risks linked to autonomous AI systems.

Alongside the AI software update, Nutanix outlined the next stage of its Kubernetes product. Nutanix Kubernetes Platform 2.19 is expected to add container management across bare metal and virtualised environments, along with an AI applications catalogue for software including Kubeflow, Milvus and Slurm.

NKP Metal is designed to automate operating system, firmware and container deployment on bare-metal Kubernetes, while NKP on AHV remains the virtualised deployment option. The platform has also received Cloud Native Computing Foundation certification for Kubernetes AI conformance, according to Nutanix.

A broader infrastructure argument underpins the launch. Nutanix is promoting what it calls a dual-native architecture, which lets virtual machines and containers run side by side under a single control plane. The aim is to help customers place AI tools closer to the data and applications they already operate.

That pitch comes as enterprises re-examine infrastructure strategies around AI and virtualisation. Nutanix cited market disruption among service providers and partners, linking the latest releases to efforts to win more channel business.

Partner push

To support that effort, Nutanix has launched the Powered by Nutanix: Verified Services programme and made SP Central generally available. SP Central is a multitenant control plane intended to let service providers build infrastructure, application, cloud-native and AI services from one management layer.

The Verified Services programme is designed to help partners develop recurring services practices across hybrid cloud infrastructure, Kubernetes and virtual machine migration. Nutanix is positioning it as a way for partners to expand beyond one-off product sales.

"Organisations across APJ are under pressure to turn AI ambition into business outcomes, while continuing to support the applications and infrastructure they already rely on. As agentic AI adoption accelerates, the priority across the region is finding a simpler way to run AI at scale, with the governance, security and consistency required for production environments. Customers want the flexibility to adopt AI on their own terms, without embarking on large-scale rearchitecting projects. At the same time, there are significant opportunities for partners and service providers to develop new AI and cloud services that help customers accelerate their modernisation efforts," said Jay Tuseth, Vice President and General Manager - APJ, Nutanix.

Nutanix also pointed to storage and hardware validation as part of its AI proposition. Nutanix Unified Storage has achieved NVIDIA-Certified Storage validation at the enterprise level, which should help customers avoid bottlenecks between data systems and graphics processors in large AI deployments, the company said.

Industry analysts see Nutanix trying to differentiate through consistency across old and new application environments. One cited commentary linked that strategy to customer demand for common policies across legacy, refactored and cloud-native systems.

A partner involved in the service provider push described the approach in terms of regulated markets. "At Continent 8 Technologies, we've worked closely alongside Nutanix to help shape the next generation of cloud services for regulated industries. Our continued collaboration with Nutanix, including the development of Service Provider Central, is built around helping our customers adopt modern cloud and AI capabilities without adding unnecessary complexity. Together, we're enabling organisations in highly regulated industries to innovate faster and focus on their core business, while maintaining the security, compliance, and operational control they require," said O'Connor.