The integration between Nutanix and Nvidia is being further deepened. The latest Nutanix Enterprise AI (NAI) solution works more closely than ever with Nvidia AI Enterprise. This makes it easier and faster to roll out agentic AI workloads.
The ecosystem surrounding agentic AI is becoming increasingly clear. At its core is the idea of AI systems that bridge multiple IT applications and environments and take over certain tasks from humans. This is currently a daunting task, but Nutanix wants to increase its feasibility. There is also talk of an “agent cycle,” in which the output of models is fed back, checked, and further refined.
Secure endpoints and Nvidia synchronization
The management layer of agentic AI is hotly contested. NAI needs to become more attractive to users thanks to shared LLM endpoints. Existing model endpoints can now be reused. This prevents the costly hardware and storage from being used much more than is strictly necessary.
In other areas, Nutanix uses Nvidia’s solutions for specific AI problems. For example, Nvidia Blueprints can serve as building blocks for AI applications and microservices. Another common concern is AI security. Within organizations, this means that LLMs must comply with company policy. Failure to do so can have major compliance, privacy, and/or security consequences. That’s why Nutanix Enterprise AI uses Nvidia’s NeMo Guardrails, which filter out content that is not approved for AI use. Even if a malicious actor manages to inject prompts, the AI model would not be able to assist with data extraction.
AI insights
The goal of using AI in business is often twofold: one is about automation, i.e., simplifying or accelerating existing work, and the other is about innovation. In concrete terms, “innovation” means gathering knowledge that would not be possible without AI technology. Think of immense mountains of data that were previously unfiltered and therefore could not be meaningfully interpreted by a human being.
With GenAI, this should be possible, with Nvidia’s AI Data Platform serving as a foundation. The Nutanix Cloud Infrastructure solution provides a private cloud layer for this purpose, converting data into insights. This includes an optimized stack that provides GPUs with relevant data as quickly as possible, while Nutanix Unified Storage makes as much data as possible available for this purpose. This means that deployment is possible via HCI, bare-metal, and cloud Infrastructure-as-a-Service.