Mirantis revealed that Nebul, a cloud service provider based in the Netherlands, has implemented the open-source k0rdent to support on-demand Artificial Intelligence (AI) inference workloads. This move leverages NVIDIA technologies to enhance infrastructure unification and streamline operations while prioritizing privacy and compliance.

Nebul emphasizes privacy and sovereignty across Europe and is recognized for its capabilities in high-performance computing (HPC), AI, and machine learning. The company, a partner of NVIDIA, is utilizing k0rdent for Kubernetes-native multi-cluster management integrated with NVIDIA GPU Operator and Gcore Everywhere Inference.

The infrastructure enables low-latency, high-performance AI inference with automated resource provisioning to meet demand effectively. Alex Freedland, co-founder and CEO of Mirantis, stated, “We believe open source is the enabler for infrastructure to support AI.” Freedland pointed out Nebul's potential in addressing challenges related to delivering AI workloads reliably at scale.

Nebul’s CEO, Arnold Juffer, highlighted the operational transition towards unifying diverse infrastructures and adopting open-source solutions to accelerate the provision of Inference-as-a-Service for enterprise clients. This transition allows clients to deploy trained AI models securely and in compliance with regulations.

Gcore’s product director Seva Vayner noted that successful AI inference at scale requires infrastructure that adapits to customer needs while ensuring optimal performance and cost efficiency.

Mirantis and Gcore announced a collaborative agreement regarding AI infrastructure solutions earlier this year. Attendees at KubeCon in London can view a demo of k0rdent and its capabilities at the Mirantis booth.