Anindo Sengupta & Dan Ciruli, Nutanix | Nutanix .NEXT 2026
summary
Nutanix positions software between chip infrastructure and models, shipping an NKP AI path that bundles Kubeflow, KServe and vector DB tooling into the platform, plus SP Central to run multi-tenant GPU clouds and sell managed compute. A new Nutanix Enterprise AI gateway enforces model and tool governance and routes requests to cheaper small models or larger foundation models while inference-engineering (quantization, KV cache) cuts token cost. The talk stresses token economics, predicts AI FinOps, and demos backpack-sized Kubernetes clusters for far-edge deployments.
AI-written summary · about
