Breaking02 / IT
NVIDIA proposes AICR for Kubernetes setup on Vera Rubin GPUs
NVIDIA says Kubernetes configuration for new GPU hardware is a burden for operators because every layer, from drivers to schedulers, must fit together.

NVIDIA introduced AI Cluster Runtime (AICR) in a developer blog post on 3 Nov 2026. Those affected are operations teams configuring Kubernetes for Vera Rubin, the company's new GPU hardware.
NVIDIA described configuring and optimising Kubernetes for new GPU hardware as complex and error-prone. Kubernetes is an open-source system for running containerised applications across clusters of machines. The post places AICR within NVIDIA DSX AI Factory. Its title sets out the goal of making Kubernetes ready for AI work on Vera Rubin from day one.
Compatibility. According to the NVIDIA blog, drivers, kernels and runtimes must work together. The same applies to networking, operators and schedulers. Settings have to be tuned to the hardware and the workload. A small mismatch can stop workloads from running. It can also limit performance.
The opening of NVIDIA's post gives no version number or licence terms for AICR.