chiprook
← AI
AIOctober 21, 2026, 22:11

Nutanix and Supermicro unveil private inference system with 8 B300 GPUs per node

At NVIDIA GTC Berlin, Nutanix announced a jointly engineered private inference system with Supermicro featuring eight NVIDIA B300 Blackwell Ultra GPUs per node, dedicated management and storage clusters, and Spectrum-X networking. It is set to ship in early 2027 through Supermicro and its channel partners, with no pricing disclosed.

Nutanix and Supermicro unveil private inference system with 8 B300 GPUs per node
#Nutanix#Supermicro#Nvidia
Read next
Business

Nutanix partners with Nvidia, Supermicro and Netris on agentic AI

Business

Dell Private Cloud splits compute and storage for AI workloads

Business

Nutanix bets $20m on GPUs to run 80% of workloads on open-weight models

Hardware

NetApp Novus splits pNFS metadata from AFF A90 data nodes for 100 TB/s AI factory target