Nutanix has announced that it will be part of the next-gen NVIDIA Rubin ecosystem by offering new ways for enterprises to build and run AI infrastructure more simply.
On top of its Acropolis Operating System and AHV hypervisor, Nutanix is layering services like Nutanix Kubernetes Platform, Nutanix Enterprise AI, Nutanix Unified Storage, and Nutanix Database, and even working with Canonical to bundle Ubuntu Pro for bare metal NKP deployments, so customers get a more complete, end-to-end stack that’s optimised for NVIDIA compatibility.
What they’re really selling is the idea of a turnkey system for operationalising AI factories built on NVIDIA AI infrastructure and NVIDIA AI Enterprise software, including NIM microservices, so organisations can move from dock delivery to actually producing tokens as fast as possible with a single OS vendor tightly integrated into the NVIDIA ecosystem.
Looking ahead, Nutanix says it’s already collaborating with NVIDIA to prepare for the upcoming Rubin platforms, with planned support for Vera Arm-based CPUs, Rubin GPUs across bare metal and virtualised environments, as well as upcoming software features like NVIDIA Inference Context Memory with BlueField-4, while also reaffirming its commitment to open source by supporting NVIDIA’s open-permissible models and NIM microservices.
On top of that, support for NVIDIA Spectrum-X Ethernet Photonics switching is meant to push acceleration beyond compute into the datacentre fabric itself, and when you zoom out, the core message is simple – generative AI runs on tokens, and Nutanix and NVIDIA want to be the infrastructure layer that makes those tokens faster, more efficient and easier for enterprises to generate at scale.
