The jointly engineered platform combines Supermicro servers, NVIDIA Blackwell Ultra GPUs and Nutanix’s full-stack AI software to simplify deployment, strengthen data control and scale production AI workloads.
Nutanix and Supermicro are expanding their collaboration to help enterprises move AI from experimentation into production while gaining greater control over infrastructure costs. The companies have announced plans for a jointly engineered private AI inference solution combining Supermicro server systems with the Nutanix Cloud Platform, NVIDIA GPUs and NVIDIA Spectrum-X networking.
The offering, powered by Nutanix Agentic AI, is designed to reduce the complexity of deploying and managing enterprise AI infrastructure. It aims to provide organisations with a pre-validated, rack-scale system that reduces custom integration requirements, accelerates deployment and enables more predictable access to high-density GPU computing.
The architecture brings together infrastructure provisioning, model execution and agentic workload governance. Nutanix Kubernetes Platform and Nutanix Enterprise AI will support bare-metal provisioning, model-as-a-service delivery and a secure Agent Gateway. Nutanix Unified Storage (NUS) will provide high-performance storage for AI workloads, including context caching and model serving.
The hardware configuration includes Supermicro systems housing eight NVIDIA Blackwell Ultra GPUs per node. A dedicated Nutanix management cluster, based on the NX-8170-G10, is designed to protect control-plane operations, while an integrated NUS cluster using the NX-8150-G10 supports high-bandwidth data ingestion, checkpointing and cache warming.
“By combining validated server systems with Nutanix’s full-stack AI software, we aim to help organisations across the region scale AI workloads more efficiently, reduce deployment complexity and build a more predictable foundation for long-term AI adoption.”
— Sush Kajaria, EMEA Alliance Sales Director, Nutanix
NVIDIA Spectrum-X networking, incorporating Spectrum-4 Ethernet switches, ConnectX-7/8 SuperNICs and BlueField-3 DPUs, will provide high-performance GPU-to-GPU connectivity and infrastructure offloading.
“Moving AI from experimentation into production remains a challenge for many enterprises, with infrastructure complexity and hardware procurement often slowing progress,” said Thomas Cornely, Executive Vice President, Product Management, Nutanix. He said the collaboration aims to help enterprises deploy private inference and better control cost per token as AI adoption expands.
For enterprises across the Middle East and Africa, the proposition is particularly relevant as AI initiatives move towards production and organisations seek measurable business value from their investments.
“By combining validated server systems with Nutanix’s full-stack AI software, we aim to help organisations across the region scale AI workloads more efficiently, reduce deployment complexity and build a more predictable foundation for long-term AI adoption,” said Sush Kajaria, EMEA Alliance Sales Director, Nutanix. He added that the collaboration is designed to give enterprises greater control over their data, security and AI operating costs.
The systems powered by Nutanix Agentic AI are expected to become available through Supermicro and its channel partners in early 2027.
