The evolution towards Agentic AI models is causing the main challenge for companies to no longer be the model to choose or the creation of individual agents, but the complexity of managing the infrastructure necessary to securely run thousands of agents at scale.
Infrastructure and platform teams need more effective solutions to facilitate shared access to resources while maximizing performance, security, and compliance with sovereignty requirements. Additionally, data scientists and AI solution developers demand easier access to tools and services that allow them to run and tune models, create agents, and securely connect them to enterprise data.
That’s why Nutanix has announced Nutanix Agentic AI, a new full stack software solution designed specifically to help businesses accelerate the adoption of agentic AI by reducing operational complexity and integrating with Nvidia AI Enterprise.
“Unlike model-training AI infrastructure optimized to run a single large process, agentic AI infrastructure must manage thousands of AI services and agents, multiple concurrent users, and constantly evolving environments. Therefore, Nutanix Agentic AI extends the capabilities of our AHV hypervisor, Flow Virtual Networking, Nutanix Kubernetes Platform, and Nutanix Enterprise AI to deliver a cloud operating model applied to enterprise AI factories. So infrastructure and platform teams can build, operate and govern these factories more easily, while developers have the performance and AI services they need,” explains Thomas Cornely, Executive Vice President of Product Management at Nutanix.
This new solution integrates with NVIDIA AI Enterprise, at the Agent Builder layer, and orchestrates the ecosystem of NVIDIA-certified AI factories in supported configurations. This way, companies can build, run, and secure agentic AI applications in dynamic, multi-tenant environments, combining advanced infrastructure security and orchestration capabilities with platform services (PaaS) and models as a service (MaaS) for data scientists and developers. Nutanix and NVIDIA are also collaborating to lay the foundation for autonomous agents in the enterprise, by integrating with the NVIDIA Agent Toolkit, which includes the open source NVIDIA OpenShell runtime environment.
Optimizing performance, security and cost in AI applications and factories
Nutanix Agentic AI reduces operational complexity, optimizes performance and security, and is designed to deliver lower, more predictable costs per token. This solution includes:
- Advanced AI Gateway and Model-as-a-Service (MaaS). The latest version of Nutanix Enterprise AI (NAI 2.6) incorporates an AI Gateway service that allows unified policies to be applied to hosted LLMs, both in the cloud and in private environments. Additionally, it adds support for the Model Context Protocol (MCP), extending its Model as a Service (MaaS) capabilities so agents can securely connect to enterprise tools and data sources. NAI also now includes support for the NVIDIA Nemotron family of open source AI models, datasets, and training tools, designed to help developers create agentic AI systems capable of reasoning, securely accessing tools, and completing complex tasks autonomously.
- Open Kubernetes platform with extensive AI catalog. Nutanix simplifies the path to agentic AI by extending its CNCF standards-compliant Nutanix Kubernetes Platform (NKP) with an extensive catalog of open source tools for AI developers. Thanks to its integration with NVIDIA AI Enterprise, developers can immediately deploy NVIDIA NIMs, including Nemotron models, and accelerate the development of high-performance AI applications.
- Infrastructure optimization and security. The Nutanix AHV hypervisor can automate the allocation of physical resources to virtual machines on servers with high GPU density. For its part, Nutanix Flow Virtual Networking allows you to offload the network data plane to NVIDIA BlueField, offering high-performance networks while reducing host CPU and memory consumption. These capabilities bring all the benefits of virtual machines to agentic AI workloads, ensuring maximum performance, security and resource efficiency, and helping to reduce the cost per token.
- Fundamental data services for AI. Agentic AI applications require advanced data services. As a solution based on NVIDIA’s AI data platform, Nutanix Unified Storage delivers linearly scalable read/write performance for thousands of GPU clients. By incorporating a high-capacity layer for KV Cache offloading and support for S3 over RDMA and NFS over RDMA, Nutanix provides a scalable, low-latency data fabric that maximizes GPU efficiency across all enterprise AI workloads.
The Nutanix Agentic AI solution helps accelerate the adoption of agentic AI through NVIDIA-certified AI factories. Thus, companies can deploy these factories on Cisco, Dell or Supermicro infrastructure, jointly validated by Nutanix and NVIDIA.
Nutanix Agentic AI reduces operational complexity, optimizes performance and security, and delivers lower, more predictable costs per token
“Agentic AI requires a high-performance infrastructure capable of securely managing thousands of agents at enterprise scale. Nutanix’s integration of the NVIDIA Agent Toolkit and Nemotron’s open models provides enterprises with a foundation to create and manage efficient AI factories. We are collaborating to scale autonomous AI in the enterprise as AI agents continue to evolve.” Justin Boitano, Vice President, Enterprise AI Platforms, NVIDIA.
